Exploring the Relationship Between Disability and Children’s Household Chores in Selected Low- and Middle-Income Countries

Abstract

Studies on disability and unpaid domestic and care work (UDCW) in low- and middle-income countries are rare. This paper contributes to a nascent body of literature by exploring how children’s unpaid workloads at home are affected by the disability status of their primary caregiver and of younger children within the household in 34 low- and middle-income countries and areas. We constructed ordinary least-squares regression models for the pooled data using country-fixed effects and for each country and area separately. Random effects models are included as a test of robustness. Our analysis of the pooled data revealed that when a household included a child aged 2–4 with a disability, the time dedicated to UDCW by children aged 5–17 rose by approximately 10 percent, on average. While we did not find an association between the disability status of the caregiver and the time children dedicated to UDCW in the pooled regressions, in the country-specific regressions, having a caregiver with a disability was associated with changes in the time spent on UDCW in certain countries, signaling the heterogeneity of country experience. As such, there is reason to devote additional attention to understanding the implications of disability on the amount of time children spend on UDCW and to consider what types of support might be needed in households with diverse disability profiles.

1. Introduction

Worldwide, nearly 240 million children have some form of disability [1]. Depending on the type and severity of their disability, many of these children can contribute to household chores [2,3,4]. Such children, however, often require additional care to ensure their basic needs are met, and this role typically falls to their mothers [5,6,7,8,9,10]. A disabled child’s siblings may also contribute to their care or take on additional household work while their mother tends to the child. Furthermore, a child may take on added care and household responsibilities if their mother has a disability, especially where formal and informal support is lacking. However, empirical evidence on the impact of having household members with a disability on children’s engagement in unpaid domestic and care work (UDCW) is relatively limited.

Much of the existing research focuses on high-income settings, particularly the United Kingdom, United States and Australia [11], and on children’s care of a parent with an illness or disability. In the UK, for instance, young carers have been found to take on a wider range of household chores and to spend much longer hours on these tasks than other children [12,13]. A sizeable literature also documents the large number of children, particularly girls, caring for parents with HIV/AIDs and other diseases in Sub-Saharan Africa [11,14,15,16,17]; some research extends this focus to the care of parents with HIV/AIDs in other contexts such as India [18] and to the care of parents with disabilities, mental illness and other health conditions [19]. Research has also found children’s heightened engagement in UDCW in households when a sibling has a disability, but this is very limited in scope [20,21,22,23]. Across contexts, young carers have been found to face disadvantages relating to education, health, well-being, social opportunities and employment [15,24,25,26]. These constraints are likely to be more acute in low- and middle-income settings where protective factors may be available only to limited populations [15,17].

It is unclear at what point along a continuum of engagement children’s involvement in UDCW becomes detrimental. Amarante et al. (2024) summarize the potential positive and negative outcomes: “Performing caring activities at early ages can help to develop children’s sense of responsibility and it can be a rewarding and meaningful experience. But being a caregiver may also entail important consequences for children, both in their present and future lives. Performing permanent and intensive caregiving activities can crowd out other enriching activities for children. It may affect children’s ability to attend classes, spend time studying or engage in other age-appropriate activities that contribute to the development of healthy and positive adult personalities… [and] may contribute to the formation of preferences that determine gender roles in adulthood” [27] (p. 2600).

Studies of child disability and time spent on UDCW in low- and middle-income countries are rare, with most analyses focusing on whether children with disabilities are more likely to be engaged in child labor (i.e., economic activity and/or unpaid household services at or above age-specific hourly thresholds) [1,28]. Yet, there is growing recognition that even below these hourly thresholds, time spent on UDCW can impact children’s current well-being and transition into adulthood. For example, data from the Young Lives study for Ethiopia, India, Peru and Vietnam show that child involvement in domestic chores can hinder cognitive and non-cognitive achievements when it crowds out school and/or study time [29,30]. In fact, Keane et al. find “no clear evidence that the adverse effects of either work or chores increase with the level of hours” [30] (p. 457). The important conclusion is that it is essential to consider the impacts of engagement in chores at lower thresholds and what aspects of that engagement may either support or undermine a child’s well-being.

Moreover, time allocation is not gender-neutral. Like adult women, girls spend more time, on average, engaged in UDCW than boys, with gender disparities increasing as children age [31]. These differentials can increase the opportunity cost of sending girls to school and stymie young women’s entry into the labor market [32]. Indeed, across contexts, including Colombia, Mexico, Egypt and Ethiopia, excessive engagement in domestic chores has been found to interfere disproportionately with girls’ education [27,32,33,34,35].

The paucity of quantitative information on how the presence of a household member with a disability affects children’s engagement in household chores motivates the analysis in this article. Our findings contribute to a nascent body of literature by exploring cross-nationally how children’s unpaid workloads at home are affected by the disability status of their primary caregiver and of younger children within the household in low- and middle-income countries and areas. In doing so, its findings can be used to inform programs designed to support the caregivers of children with disabilities as well as family-friendly policies aimed at promoting gender equality.

In analyzing the data, we tested two hypotheses: first, that children aged 5–17 who lived with a young child (aged 2–4) with a disability spent more time on unpaid household services than children in households without a young child with a disability and, second, that children aged 5–17 spent more time on unpaid household services if their mother/primary caregiver had a disability than peers whose mothers/primary caregivers did not have a disability. In both instances, we expected that increases in time spent on unpaid work would be greater for girls than boys owing to gender norms [30,36].

2. Materials and Methods

2.1. Data Sources

This study used data from the UNICEF-supported Multiple Indicator Cluster Surveys (MICS) conducted in 34 countries and areas between 2017 and 2022 to explore how living with a family member with a disability affects children’s unpaid workloads [37]. These countries are spread across Africa, Asia, Europe and the Pacific, representing a diverse range of geographical regions, including North Africa, West Africa, Central Africa, South Asia, Eastern Europe, the Middle East and the Pacific Islands.

The MICS is an international household survey program that provides one of the largest sources of internationally comparable population-level data on women and children. The sixth round of MICS (2017–2022) collected data on the disability status of children and adults within the household. In 2016, the Child Functioning Module (CFM) and the Washington Group Short Set on Functioning (WG-SS) became part of the MICS; they are respectively used to collect data on children aged 2–17 and adult women and men aged 18–49. The CFM comprises two questionnaires: one with 16 questions for children aged 2–4 and another with 24 questions for children aged 5–17 [1]. The questions are meant to be addressed to the child’s mother or primary caregiver in cases where the mother is not alive or not living in the household. They are designed to identify difficulties in several domains of functioning (8 domains for children aged 2–4 and 12 domains for children aged 5–17). For children aged 2–4, the domains assessed are seeing, hearing, walking, fine motor, communication, learning, playing and controlling behavior. For children aged 5–17, the functional domains are seeing, hearing, walking, self-care, communication, learning, remembering, concentrating, accepting change, controlling behavior, making friends and affect (anxiety and depression). To better reflect the degree of functional difficulty, each area is assessed against a rating scale (“no difficulty”, “some difficulty”, “a lot of difficulty” and “cannot do at all”). The WG-SS covers difficulties in six domains of functioning (i.e., seeing, hearing, walking or climbing stairs, remembering or concentrating, self-care and communication) [38].

The surveys also gathered data on the time children aged 5–17 spent on unpaid household services in the past seven days, including cooking, cleaning, shopping and taking care of other children or sick or elderly family members.

2.2. Measures

We calculated, as the dependent variable, the log number of hours children aged 5–17 spent in the past seven days on UDCW, as reported by their mother or primary caregiver. These are tasks “carried out by and for household members, including cleaning the house, looking after siblings, washing dishes and shopping” and more formally categorized as “unpaid household services” [34] (p. 1). Because a small subset of children in certain countries exhibited a total number of hours exceeding plausible limits, we winsorized the upper 0.5% of the distribution to reduce the influence of these extreme values. Winsorization is a method of averaging that involves replacing extreme values in the distribution with less extreme ones; in this case, values in the top 0.5% were replaced with the value at the 99.5th percentile. Further, our use of the log transformation of hours worked in regression analysis alongside robust estimation methods reduced the impact of outliers.

Independent variables included the disability status of young children aged 2–4 and mothers/primary caregivers aged 18–49. For children aged 2–4, we constructed one binary measure of disability. Children were defined as having a disability if they reportedly kicked, bit, or hit other children or adults a lot more than other children of the same age and/or had “a lot of difficulty” with or “cannot do at all” one or more specified functions, including seeing, even if using glasses; hearing, even if using a hearing aid; walking, even if using equipment or assistance; understanding or being understood when speaking; picking up small objects with their hands; learning things; and playing [1] (p. 20).

Mothers/primary caregivers were defined as having any disability if they were reported to have any functional difficulty in one or more of the six functional domains (i.e., seeing, hearing, walking, remembering/concentrating, self-care and communication). The measurement of adult disability in MICS, based on the WG-SS, is not without limitations. Its purpose is to permit the disaggregation of key well-being indicators by disability status rather than to provide an estimate of the entire population of persons with disabilities. The WG-SS, for instance, does not cover psychosocial difficulties. It follows that our measure of caregiver disability does not capture the full spectrum of persons with disabilities. We nonetheless believe that the inclusion of this disability marker is important in order to begin to understand better how caregiver disability affects children’s time allocation and how this may vary across contexts.

Because the questionnaire containing adult disability questions was administered only to adults aged 18–49, our data did not contain any information about the disability status of any older household members. As a result, in households where an older person had a disability, children may have assumed added unpaid care responsibilities to compensate, an aspect our analysis overlooked. Equally, older adults without any functional difficulties may have substituted for the unpaid care work of children. However, the distributions of older people with a disability were unlikely to be correlated with the disability status of other household members, so we expected it to not bias our estimates unduly.

Control variables included several covariates pertaining to the child, the child’s mother/primary caregiver and the household that may directly influence the amount of time children spent performing household chores. Among these were six measures of the disability status of the children aged 5–17 whose time allocations to UDCW we were measuring. First, we created a binary variable measuring whether a child had a disability, defined as reportedly seeming very anxious, nervous or worried, very sad or depressed on a daily basis and/or had “a lot of difficulty” or “cannot do at all” with at least one specified function, including seeing, even if using glasses or contact lenses; hearing, even if using a hearing aid; walking on level ground, even if using equipment or assistance; performing self-care activities, such as feeding or dressing themselves; being understood when speaking to people inside or outside their household; learning things; remembering things; concentrating on an activity they enjoyed; accepting changes in their routine; controlling their behavior; and making friends. We also created four additional categorical variables to distinctly measure severe functional difficulties (i.e., those with more than one functional difficulty) in each of the following domains: learning, communication, walking and self-care. The sixth variable measured severe disability across all four categories.

2.3. Statistical Analysis

To test our hypotheses, we undertook multivariate regression analysis using ordinary least-squares regression models to identify the extent to which the presence of a primary caregiver and/or a child aged 2–4 with a disability in a household affected the amount of time that girls and boys aged 5–17, who may also have a disability, spent on unpaid household services. Our main specification for the pooled data takes the form (Equation (1)):

$$\begin{array}{c}{ln(hours\_household chores)}_{child}\hfill \\ ={\beta }_{0 }+{\mathit{\chi }}_0+{\mathit{\chi }}_{child}{\beta }_{1 }+{\mathit{\chi }}_{caregiver}{\beta }_2+{\mathit{\chi }}_{household}{\beta }_{3 }\hfill \\ +{\beta }_{4 }( 2–4 disability)+C_{\gamma }+ {ϵ}_{child }\hfill \end{array}$$

(1)

where:

${\mathit{\chi }}_{child}$ represents a vector of child-level characteristics: gender, age (and age squared), birth order (and birth order squared), school attendance, disability status (a binary variable indicating if the child has any disability) and specific types of severe disability (the presence of a severe learning, self-care, walking or communications disability and all four severe disabilities together);

${\mathit{\chi }}_{caregiver}$ represents a vector of caregiver characteristics: education level, co-residence with the child and disability status.

${\mathit{\chi }}_{household}$ represents a vector of household characteristics: place of residence (urban or rural), wealth quintile, the number of women of reproductive age, the number of older women, the number of children under five, the number of girls and boys aged 5–17, the gender and age of the household head and the number of women aged 18–49 with a disability in the household;

$2–4 disability$ is a variable indicating the presence of one or more children aged 2–4 with a disability; and

$C_{\gamma }$ represents a vector of country fixed effects, where each country (excluding the baseline) is indicated by a binary variable.

To this specification, we interacted in turn, key variables related to the disability status of the caregiver and children aged 2–4 years old with the sex of the child aged 5–17, whether they were attending school and their disability status, in order to uncover any additional patterns underlying these relationships. We interacted too the disability status of the child aged 5–17 with their sex and school attendance.

Our main specification for the country-level regressions is identical to that of the pooled regression, with the exception of the country fixed effects. Appendix A.1 contains further details of how each independent variable is constructed and the key interactions that were included.

We undertook our analyses for each country and area separately and as a pooled dataset. Population-based sampling weights were assigned to the pooled data, and these models were run with and without country-fixed effects. Standard errors across all models were clustered at the level of primary sampling units to accommodate the MICS design.

Aggregate results were presented in two ways. First, we assigned each country an equal weight, and, second, we used pooled data combined with population-based sampling weights. To calculate sampled weights in the pooled data, we merged country-level data into a pooled dataset together with the country’s sample weights and sample design variables (primary sample units and strata). In this pooled dataset, weights were maintained for country-level analyses. First, a single-stratum indicator was generated to reflect the complete multiframe design and ensure that each stratum and primary sample unit assumed a unique value across countries; next, population scaling was carried out to recalibrate the country-specific weights, adjusting for country population size relative to sample sizes [39]. The population weights were computed based on each country’s relative share of 5–17-year-olds according to World Population Prospects for 2019, the median year among the surveys included in the analysis.

3. Results

The descriptive results (

Table 1. Key descriptive statistics for the variables included in the regression in the pooled dataset ¹.

Variable	Mean (Confidence Interval)
Child aged 5–17
Log of hours spent weekly on household chores	1.36 (1.35, 1.38)
Sex
Female (%)	60.07 (59.36, 60.77)
Male (%)	39.93 (39.23, 40.64)
Mean age	11.67 (11.62, 11.72)
Birth order	2.03 (2.00, 2.06)
Child is out of school (%)	10.44 (9.96, 10.94)
Has any disability (%)	13.46 (12.87, 14.08)
No severe disability (%) (if has any disability)	99.65 (99.55, 99.72)
Severe learning disability (%) (if has any disability)	0.06 (0.03, 0.11)
Severe self-care disability (%) (if has any disability)	0.05 (0.02, 0.13)
Severe walking disability (%) (if has any disability)	0.19 (0.15, 0.25)
Severe communication disability (%) (if has any disability)	0.04 (0.02, 0.07)
All four severe disabilities (%)	0.01 (0.01, 0.03)
Child aged 2–4
Has any disability (%)	3.05 (2.78, 3.35)
Mother or primary caregiver
Mother lives away from household (%)	5.18 (4.82, 5.57)
Mother has more than a primary education (%)	37.24 (36.40, 38.09)
Has any disability (%)	7.06 (6.57, 7.58)
Household
Lives in rural area (%)	66.08 (65.10, 67.04)
Socio-economic status (Wealth quintile)
Poorest	21.62 (20.82, 22.45)
Quintile 2	21.91 (21.21, 22.62)
Quintile 3	20.54 (19.92, 21.18)
Quintile 4	19.43 (18.78, 20.10)
Richest	16.50 (15.75, 17.27)
Household composition
Number of women aged 18–49	1.73 (1.71, 1.76)
Number of women aged 50+	0.36 (0.34, 0.37)
Number of males aged 5–17	1.57 (1.53, 1.61)
Number of females aged 5–17	1.72 (1.69, 1.76)
Number of under-fives	0.90 (0.88, 0.92)
Household head is female	14.59 (14.05, 15.14)
Mean age of household head	45.10 (44.91, 45.29)

¹ n = 85,392. Data are proportions or means with 95% confidence intervals.

) are presented for the pooled data. We conducted multivariate analyses separately for the pooled data with country-fixed effects, for random effects (as a robustness test) and for each country. We will discuss the results of the pooled and country-level analyses in turn for the main specification (model 1) and, additionally, the specification containing the interaction between the disability status of the caregiver and the disability status of children aged 2–4 years old (model 2). The reported coefficients are average marginal effects, indicating the impact of a one-unit change in continuous variables or a category change in categorical variables. Since the dependent variable is the natural log of hours worked, each coefficient is exponentiated to estimate the approximate percentage change in hours worked, holding all else constant.

3.1. Pooled Regressions

Our reporting of the pooled analyses focuses on country-fixed effects, with the random effects model computed as a robustness test (Appendix A.2). Our analysis supported our hypothesis that the presence of a child aged 2–4 with a disability in a household is associated with an increase in the amount of time children aged 5–17 spent on unpaid household services per week (

Table 2. Average marginal effects for covariates in the pooled model, country-fixed effects ^1,2,3,4.

	Model 1	Model 2
Child aged 5–17
Female	0.358 *** (0.016)	0.357 *** (0.016)
Age (in years)	0.061 *** (0.002)	0.061 *** (0.002)
Birth order (continuous)	−0.012 (0.008)	−0.012 (0.008)
Does not attend school	0.174 *** (0.023)	0.174 *** (0.023)
Has any disability	0.012 (0.020)	0.012 (0.020)
Severe learning disability	−0.058 (0.183)	−0.054 (0.183)
Severe self-care disability	−0.532 (0.286)	−0.534 (0.286)
Severe walking disability	0.172 (0.098)	0.172 (0.098)
Severe communications disability	−0.173 (0.246)	−0.179 (0.243)
All four severe disabilities	−0.333 (0.176)	−0.334 (0.177)
Children aged 2–4
Has any disability	0.087 * (0.043)	0.105 * (0.043)
Mother or primary caregiver
Has secondary education or higher	−0.062 *** (0.015)	−0.062 *** (0.015)
Lives in different household from child aged 5–17	0.040 (0.037)	0.040 (0.037)
Has any disability	−0.023 (0.033)	−0.017 (0.033)
Mother or primary caregiver and child aged 2–4
Caregiver and young child both have a disability		−0.180 (0.104)
Household
Rural	0.042 * (0.018)	0.042 * (0.018)
Wealth quintile 2	−0.019 (0.020)	−0.020 (0.020)
Wealth quintile 3	−0.078 *** (0.019)	−0.078 *** (0.019)
Wealth quintile 4	−0.086 *** (0.021)	−0.085 *** (0.021)
Wealth quintile 5	−0.147 *** (0.024)	−0.147 *** (0.024)
Number of women aged 18–49	−0.022 ** (0.008)	−0.022 ** (0.008)
Number of women aged 50+	0.007 (0.014)	0.007 (0.014)
Number of under-fives	0.047 *** (0.008)	0.046 *** (0.008)
Number of girls aged 5–17	0.001 (0.010)	0.001 (0.010)
Number of boys aged 5–17	0.02 1** (0.006)	0.021 ** (0.006)
Household head is female	0.013 (0.017)	0.013 (0.017)
Age of household head (in years)	−0.000 (0.001)	−0.001 (0.001)
Number of women in household with a disability	−0.014 (0.022)	−0.014 (0.022)
Coefficient of determination (R2)	0.1696	0.1698
Number of observations (regression)	85,392	85,392
Number of observations (marginal effects)	80,024	80,024

¹ The dependent variable is the log of hours spent on household chores. ² Standard Errors in parenthesis. ³ Country fixed effects are not displayed for the sake of clarity. ⁴ Significance probabilities: * p < 0.05, ** p < 0.01, *** p < 0.001.

). Specifically, the presence of a 2–4-year-old child with a disability correlated with an increase of approximately 10% in hours worked. However, the analysis did not find support for our second hypothesis. The disability status of the primary caregiver did not predict the time children aged 5–17 had allocated to UDCW.

Several covariates were observed to be consistent predictors of the amount of time 5–17-year-olds allocated to UDCW. Being female was the most influential predictor, leading to an increase of around 43% in hours worked. However, neither interaction between the sex of the child providing care and the disability status of either the caregiver or any young children was statistically significant. Other significant predictors included being out of school, the child’s age, residence in a rural setting and the household counts of children under five and male children aged 5–17. In contrast, protective factors included household wealth, the number of adult women aged 18–49 and the educational level of the child’s mother, notably if she had attained more than primary education. The disability status of the child aged 5–17 under consideration was not a statistically significant predictor of their own time use.

Finally, we introduced various interactions in the models to better understand whether and how the disability status of the various household members intersected with the sex of the child and the child’s education. These interactions yielded statistically insignificant outcomes. The interaction between the disability status of a 2–4-year-old and their caregiver, however, was associated with a reduction in the amount of time spent by children aged 5–17 on household chores by 16.5% that was statistically significant at the 10% level. The observed relationships are robust to the exclusion of country-fixed effects.

3.2. Country Regressions

Our analysis of the country-level data points to the heterogeneity of country experience, albeit for many countries the results corresponding to the disability indicators were not statistically significant. For each disability marker with the exception of a 5–17 year old child having all four severe disabilities, some countries showed a substantial positive impact on the time allocated to UDCW, while others indicated a pronounced negative effect (

Table 3. Summary of sign of marginal effects for covariates in 34 countries or areas ¹.

	Caregiver Disability	2–4-Year-Old Disability	Child Aged 5–17
			Disability	Severe Learning Disability	Severe Self-Care Disability	Severe Walking Disability	Severe Communication Disability	All Severe Disabilities
Total Results	4	5	8	7	4	12	12	3
Positive Results	2	2	4	3	1	5	5	0
Negative Results	2	3	4	4	3	7	7	3

¹ The table summarizes the results of country-level regressions corresponding to model 1 in Table 2.

, Appendix A.3).

For example, the presence of a child aged 2–4 with a disability in the household was associated with an increase in the amount of time children aged 5–17 spent on unpaid household services of upwards of 80% in Kiribati and Republic of North Macedonia (Roma), whereas it was associated with a decrease in the amount of time children spent on household chores of up to 59% in Tuvalu. Having a caregiver with a disability corresponded with an increase in the amount of time 5–17-year-olds of up to 191% in Montenegro (Roma) and with a decrease of up to 65% in Samoa.

Finally, as observed in the analysis of the pooled data, certain consistent drivers emerged regarding the time 5–17-year-olds dedicated to UDCW. These included being female, being out of school, age, and, to a lesser extent, the presence of under-fives in the household. Conversely, factors associated with reductions in time spent on UDCW—or protective factors—included the number of adolescent girls and women aged 15–49 in the household, household wealth, whether the mother has an education level beyond primary school and the presence of all four severe disabilities for a child aged 5–17.

4. Discussion

Our analysis found support for one of the two hypotheses informing this study. Notably, the analysis of the pooled regression results revealed that when a household includes a child aged 2–4 with a disability, the time dedicated to unpaid household services rose. The country-fixed effects model predicted an increase of approximately 10% in the amount of time children allocated to UDCW. While we did not find an association between the disability status of the caregiver and the time children aged 5–17 dedicate to UDCW in the pooled regressions, in the country-specific regressions, having a caregiver with a disability was associated with changes in the time spent on household services in certain countries. As discussed above, our measures of adult disability may be biased, as they do not include the full range of differences that may have an impact on functioning. However, they also suggest that caregiver disability may be important in some contexts and hint at the heterogeneity of country experience.

When outlining our hypotheses, we posited that increases in time spent on unpaid work owing to the disability of another household member would be greater for girls than for boys, owing to gender norms. While sex was indeed an important marker of engagement in household chores—with girls tending to devote upwards of 40% more time to this activity than boys—there was no additional effect associated with being female and living in a household with either a disabled caregiver or young child. In other words, girls and boys alike tended to perform more household chores if they lived in a household with a disabled young child (and in certain countries, with a disabled caregiver) than they would otherwise.

The interaction between the disability status of a 2–4-year-old and their caregiver on time spent by children aged 5–17 on UDCW was negative, though only statistically significant at the 10% level. In such households, children may be reallocating their time to other forms of work, such as paid labor, to compensate for constraints to the household, or these households may be accessing more external assistance. The distinct policy implications of these two scenarios render this an important area for future inquiry.

Our study also highlights several avenues for future research. First, the MICS datasets used in this analysis provided only a partial profile of the disability status of household members. Within a household, the data collected provided information regarding the disability status for all women aged 15–49, a subsample of men aged 15–49, one randomly selected child aged 5–17 and all children aged 2–4 years old. The disability status of additional children in the household aged 5–17 and of any older household members is not known. Moreover, the existing information on adult disability is, as discussed above, incomplete as the measure in the MICS dataset is not intended to provide a full picture of functional difficulties, further explained by Hanass-Hancock et al. [40]. Household surveys that provide—at least for a subsample of households—a comprehensive disability profile for all household members, alongside data on children’s time use, would enable a more nuanced and comprehensive understanding of how disability affects intrahousehold time allocations.

Second, our analysis did not probe the impact of engagement in UDCW on the well-being of children who are spending additional time on household chores owing to the presence of a disabled member. This is an important area for future research as evidence is still accumulating as to the level and types of engagement in household chores that are considered harmful for children and why. The study of young carers in low- and middle-income countries, in particular, is under-researched.

Advances in collecting data on children’s time use cross-nationally—for example, the inclusion of time diaries for children as an optional module in the seventh round of MICS—should enable a richer and more holistic understanding of how time spent on UDCW affects the time children spend on schooling, leisure and in other domains, and the consequences for their well-being. It would also reveal whether children in households with both a caregiver and a young child with a disability are engaging in less unpaid household services because they are undertaking more paid work or for other reasons.

Third, the diversity of country experiences our analysis has highlighted underscores the need for the additional study of what factors within countries are prompting the greater and lesser engagement of children in UDCW in households with a disabled young child. Fourth and finally, there is limited information on the practical implications of the finding that children may be taking on additional household work in households in which another member has a disability. Various potential avenues warrant exploration—for example, social protection and human assistance policies geared toward easing the care workloads in households containing young children with disabilities [41]; labour market policies that provide paid leave to care for children with disabilities [31] (p. 76); disability-friendly schools and childcare centers [42,43]. This article takes the important step of recognizing the additional chores that children appear to be taking on in households containing young children with a disability—further work is needed to establish why this is occurring and consequently, the appropriate policy response.

5. Conclusions

To our knowledge, our study is the first cross-national analysis to examine how children’s unpaid workloads at home are affected by the disability status of their primary caregiver and younger children within the household in low- and middle-income countries. While we expect the robustness of these findings to improve as more countries collect data, our results indicate that children take on significant responsibilities when they have younger siblings with disabilities and, in certain countries, when their caregiver has a disability.

The amount of time that most children aged 5–17 spend on unpaid household services is relatively low, well beneath the threshold that constitutes child labor for children aged 5–14. However, emerging research suggests that any engagement may be harmful if it impinges on human capital development, making this an important area of inquiry. Moreover, inequalities in the distribution of UDCW associated with markers such as sex (e.g., a higher engagement by girls than boys) and the disability status of household members are themselves unjust and may, therefore, warrant policy attention.

As such, there is reason to devote additional attention to understanding the implications of disability on the amount of time children spend on UDCW and to consider what types of support might be needed in households with diverse disability profiles. Accordingly, family-friendly policies and programs designed to support the caregivers of children with disabilities should recognize that the siblings of such children, particularly girls, are often caregivers themselves in need of assistance. For example, respite services designed to mitigate caregiver burdens can also be directed toward siblings to ensure that the additional care and chore burdens they take on do not impinge on the time they spend going to school, studying, or playing—activities that all children independent of their disability status need to develop and thrive.