Content uploaded by Christina L. Scanlon
Author content
All content in this area was uploaded by Christina L. Scanlon on Sep 16, 2021
Content may be subject to copyright.
Original article
COVID-19 Employment Status, Dyadic Family Relationships, and
Child Psychological Well-Being
Ming-Te Wang, Ed.D.
a
,
*
, Daphne A. Henry, Ph.D.
b
, Juan Del Toro, Ph.D.
a
,
1
,
Christina L. Scanlon, Ph.D.
a
,
1
, and Jacqueline D. Schall, M.S.
a
,
1
a
Learning Research & Development Center, Department of Psychology, School of Education, University of Pittsburgh, Pittsburgh, Pennsylvania
b
Department of Counseling, Developmental & Educational Psychology, Boston College, Chestnut Hill, Massachusetts
Article history: Received March 24, 2021; Accepted July 14, 2021
Keywords: COVID-19; Employment status; Family relationship; Psychological well-being; Socioeconomic status
ABSTRACT
Purpose: COVID-19 has led to soaring unemployment rates and the widespread adoption of
working-from-home (WFH) arrangements that have disrupted family relationships and adolescent
psychological well-being. This longitudinal study investigated how parental employment status
(i.e., job loss and WFH) influenced adolescents’daily affect indirectly through family functioning
(i.e., parent-adolescent conflict and parental warmth) and whether these links varied by family’s
socioeconomic status.
Methods: Daily-diary approaches were used to collect dyadic parent-adolescent data from a
nationwide American sample (6,524 daily assessments from 447 parent-adolescent dyads; 45%
black, 36% white, 10% Latinx, 7% Asian American, 2% Native American) over the course of 15
consecutive days at the onset of the COVID-19 pandemic.
Results: Parents who experienced job loss demonstrated increases in parent-child conflict, which
in turn predicted decreases in child positive affect and increases in child negative affect.
Furthermore, parents’WFH status predicted increases in parental warmth, which in turn predicted
increases in child positive affect and decreases in child negative affect. Parents of low-income
families were more likely to experience job loss (24% vs. 13%) and less likely to WFH (44% vs.
73%) than middle-high income parents.
Conclusions: Adolescents from families facing economic hardship and employment shifts during
COVID-19 experienced changes in parent-child relational dynamics that influenced their emotional
well-being. Recognizing these shifts in family ecology is critical to health providers’ability to
screen for mental health, assess existing family supports, and provide timely, targeted information
about stress management and contending with family conflict.
Ó2021 Society for Adolescent Health and Medicine. All rights reserved.
IMPLICATIONS AND
CONTRIBUTION
COVID-19 has upended
the economic landscape
in the United States
through widespread job
loss and rapid transitions
to working from home.
These results suggest that
working-from-home ar-
rangements help parents
foster warm relationships
with their children that in
turn reduce negative
affect and promote posi-
tive affect.
This work was supported by Spencer Foundation (#201600067).
Conflicts of interest: The authors have no conflicts of interest to disclose.
Data availability statement/Data deposition: The data sets generated and/or
analyzed during the present study are not publicly available, but they are
available from the corresponding author on reasonable request.
Contributors statement: Wangconceptualized and designedthe study,drafted the
introduction, method, and result sections, and reviewed/revised the manuscript.
Henry drafted the discussionsection and reviewed/revised the manuscript. Del Toro
conducted data analysis and critically reviewed the manuscript for important
intellectual content. Scanlon and Schall coordinated and supervised data collection
instruments, collected data, drafted the literature review, and reviewed/revised the
manuscript. All authors approvedthe final manuscriptas submitted and agree to be
accountable for all aspects of the work.
*Address correspondence to: Ming-Te Wang, Ed.D., University of Pittsburgh,
5948 Wesley W. Posvar Hall, 230 South Bouquet St., Pittsburgh, PA 15260.
E-mail address: mtwang@pitt.edu (M.-T. Wang).
1
These authors contributed equally and share the last authorship.
www.jahonline.org
1054-139X/Ó2021 Society for Adolescent Health and Medicine. All rights reserved.
https://doi.org/10.1016/j.jadohealth.2021.07.016
Journal of Adolescent Health xxx (2021) 1e8
COVID-19 has resulted in shifts in parental employment that
have negatively impacted family relationships [1] and individual
psychological well-being [2], especially for those contending with
poverty or economic uncertainty [3,4]. While many American
parents have experienced job loss during COVID-19, others have
shifted to temporary work-from-home (WFH) arrangements.
Researchers have examined predictors and implications of these
employment changes [5]; however, it remains unclear whether
and how parent-adolescent relationships have been impacted by
job loss and WFH during the pandemic. Furthermore, we do not
understand how these relationships may influence adolescents’
psychological well-being under circumstances where youth are
facing unprecedented and unforeseen threats to their mental
health [6]. As such, this longitudinal study examines the relation
between COVID-19 employment status, family relationships, and
adolescent affect.
To build on existing work addressing family ecological
vulnerabilities [7], we define resilience as “the capacity of a
dynamic system to adapt successfully to disturbances that
threaten system function, viability, or development”[8]. We assert
that pandemic-related changes in parents’employment status
shape family inter-relations that in turn influence individual
family members’psychological resilience and well-being. Specif-
ically, we hypothesized that job loss and WFH status influence
parent-child relationships, including parent-adolescent conflict
and parental warmth. In turn, we expected these changes in family
relationshipstobelinkedtochildren’s daily affect.
COVID-19 has upended the economic landscape in the United
States through widespread job loss and rapid transitions to WFH
arrangements. WFH requires parents to simultaneously be a
productive employee, dedicated parent, and, in the case that
their children participate in virtual learning, teacher’s assistant.
Changes to COVID-19 employment circumstances have also
occurred in conjunction with social disruption as a result of
public health measures. Accordingly, job loss and changing work
schedules may stress family routines and put significant strain on
family relationships [2,4,9]. During these times of unrest, parents
are more likely to exhibit low warmth and experience height-
ened conflict with their children [1,10]. Such contentious parent-
child relationships place adolescents at risk for elevated negative
affect [11], whereas parental warmth may buffer against negative
affect while also promoting positive affect [12].
While the extant literature is clear on the consequences of job
loss on family stress, less is known about sudden shifts to WFH
arrangements. Although remote schooling, shifting household
demands, and elevated concerns over children’sadjustmentmay
have increased parenting responsibilities and put pressure on
WFH parents [1,4,5], WFH may have also offered the opportunity
for increased parental warmth through expanded family time and
reduced work-family conflict [13,14]. In fact, the increased flexi-
bility of WFH arrangements has been associated with improved
work-family balance [13 ] and increases in available time to spend
with family [14]. Few studies, however, have examined the direct
effects of job loss and WFH on parental warmth. In sum, parental
employment status during the pandemic is of critical import
owing to consequences for family interpersonal dynamics and
adolescents’psychological well-being [2,4,7].
The impact of parental employment status on parent-
adolescent relationships and adolescents’psychological well-
being may vary by the family’s socioeconomic status [7]. Family
economic stability has been shown to buffer the negative impact
of job loss on family relationships [3], and parents who WFH may
have more resources to mitigate family stress caused by COVID-
19 [13]. Low-income families, unfortunately, may suffer the most
pervasive financial consequences of shelter-in-place policies, as
wage-earners in these families tend to work jobs that are not
amenable to WFH (e.g., essential workers) [3]. Owing to demands
on time and limited resources, low-income parents are more
likely to experience parent-child conflict and may also struggle
to maintain warm, supportive relationships with their children
[10]. Combined, economic vulnerabilities may place low-income
families at a greater risk for household contention and
maladaptive adjustment. However, studies have shown that the
presence of conflict does not necessarily imply the absence of
warmth. Even in circumstances of poverty, parental warmth has
been shown to mitigate the negative impact of economic strain
on youth well-being [15]; ergo, it is critical to understand how
job loss and WFH impact family relationships across different
socioeconomic backgrounds and how those family relationships
in turn impact adolescent well-being.
The present study
To understand whether and how parental employment status
has affected parent-child relationships and adolescents’emotional
well-being during the pandemic, we used dyadic parent-
adolescent daily-diary data with 9,372 assessments to examine
the mediating role of family relationships (i.e., parent-adolescent
conflict and parental warmth) in the link between employment
status (i.e., job loss, WFH) and adolescents’emotional well-being
(i.e., positive and negative affect). We also investigated whether
the link between employment status, family relationships, and
adolescents’affect varied by families’economic circumstances.
Daily-diary approaches provide a rich perspective into partici-
pants’real-time daily experiences and behaviors, thereby mini-
mizing systematic recall bias and permitting in-depth analysis of
within-person processes over time [16].
Because extant literature has suggested that the stress of
COVID-19 employment status is expected to increase conflict
between family members, we hypothesized that job loss would
be associated with increased family conflict, which in turn would
predict decreased positive affect and increased negative affect in
adolescents. Conversely, there may be decreased warmth within
the family as it convalesces and adapts to new ecological cir-
cumstances. Owing to a dearth in literature surrounding this
topic, we propose no specific hypotheses regarding the relations
between WFH and parental warmth aside from expecting
warmth to be associated with increased positive and decreased
negative affect of adolescents.
Methods
Participants
Participants included 447 parent-child dyads from an ongoing
nationwide longitudinal study examining school experiences,
family dynamics, and adolescent well-being in the United States.
The original study worked with a research company to recruit a
national sample of parents and adolescents (i.e., middle- and
high-school-aged) by using a representative, random sampling
method. When COVID-19 was declared a national emergency in
the United States in March 2020, the original longitudinal study
M.-T. Wang et al. / Journal of Adolescent Health xxx (2021) 1e82
was leveraged by inviting a subsample of parent (i.e., primary
caregiver) and adolescent participants to engage in a 15-day
daily-diary study focusing on family functioning and adjust-
ment during the pandemic. Because of the prevalence and
saliency of unemployment rates and adoption of WFH arrange-
ments in states with stay-at-home orders, participation required
residence in a state where stay-at-home orders mandated
schools and nonessential businesses to close.
Approximately 79% of the qualified participants from the
original study agreed to partake in the daily-diary study. The final
sample included 447 parents and their adolescent children from
38 states (parent/primary caregiver sample: M
age
¼43.7, range ¼
27e52 years; child sample: M
age
¼15.0; range ¼12e18 years;
see Table 1 for demographic information). This subsample did
not differ from the original sample regarding sociodemographic
characteristics or psychological adjustment, but there were some
geographic differences. The subsample had slightly more par-
ticipants from the Northeast and South regions (vs. West and
Midwest) as compared with the original study sample (see
Table 1). The higher number of participants from Northeast and
South regions was attributed to the fact that these states
implemented stay-at-home orders at the time of the study.
Procedures
All consented parents and children first completed baseline
and demographic measures one week before the daily survey
period. Both parents and children then completed daily survey
assessments between 5:00
P
.
M
. and 12:00
A
.
M
. using Internet-
capable devices across 15 consecutive days from May 18 to
June 1, 2020. Participants received two to four email or SMS re-
minders each day to complete the daily survey. Research staff
contacted any participants who missed a diary entry to trou-
bleshoot any issues with accessing the survey. Participants
received $40 for completing the baseline survey and daily-diary
entries. All materials and procedures were approved by the au-
thors’university institutional review board.
Measures
Parental employment status. We measured job loss and WFH as
two indicators of employment status during COVID-19. All parent
respondents were employed before COVID-19. Parents were
asked whether they had lost their job since the COVID-19
outbreak (0 ¼no,1¼yes). Parents who remained employed
were asked whether they worked from home during the COVID-
19 outbreak (0 ¼no,1¼yes). None of the parent respondents
worked from home before the pandemic. Those with another
adult in the household were asked to report on that adult’s
employment status as well.
Daily child affect. Adolescents’positive and negative affect were
measured daily using the Positive and Negative Affect Scale for
Children, a well-validated psychological scale [17]. We assessed
positive affect with four items (i.e., grateful, energetic, happy,
hopeful) and negative affect with seven items (i.e., sad, anxious,
depressed, hopeless, nervous, lonely, scared). Participants
reported their mood during the past 24 hours on a 5-point scale
ranging from 1 (not at all)to5(extremely). To generate appro-
priate reliability to detect change [18], items were averaged to
form daily composite scores of positive (R
C
¼.70) and negative
affect (R
C
¼.75).
Daily family relationship. For both parents and children, family
dynamics and functioning were measured on a 5-point Likert
scale using the Network of Relationship Inventory, a well-
validated and widely-used scale (1 ¼not at all,5¼a lot)[19].
We focused on parent and child reports of parent-child conflict (4
items; e.g., Today, I experienced conflict or tension with my child/
parent; I was yelled at by my parent) as well as parent and child
reports of parental warmth (4 items; e.g., Today, I did something
fun or relaxing with my child/parent; I felt supported by my parent).
Given that the correlations between child and parent reports
ranged from moderate to high, a mean composite score for each
dimension was calculated by summing the item scores of parent
and child reports and dividing the total by the number of items
(Parent-child conflict: R
C
¼.87; parental warmth: R
C
¼.73).
Family Socioeconomic Status. An income-to-needs ratio was
calculated by dividing the family’s total income by the federal
poverty threshold adjusted for family size. This ratio estimates
family income in a widely-used, valid, and reliable metric that
uses poverty as its critical referent [20,21]. To designate low- and
middle-high income families, an indicator was set to one if the
income-to-needs ratio was less than 2.0 (i.e., designating low-
income families) and was set to 0 if the ratio was equal to or
greater than 2.0 (i.e., designating middle-higheincome families).
Covariates. Because of their associations with family relation-
ships and well-being, we accounted for several covariates. We
included the numerical day of reporting (1e15), weekend (0 ¼
weekday,1¼weekend), and parent daily psychological distress
(item from parent report using the Psychological Distress Scale;
e.g., I feel stressed out today) as time-level covariates. Six person-
level covariates were also collected from child or parent reports:
(1) child age, (2) child sex (0 ¼girls,1¼boys), (3) child race, (4)
parent race, (5) parent role (0 ¼mothers,1¼fathers), and (6)
prepandemic parent-child relationship (2 items from parent
report using the Network of Relationship Inventory; e.g., Overall I
get along with my kid).
Analytic plan
This study examined whether family relational dynamics
mediate the association between parental employment status
Table 1
Demographic information for children and parent participants
Child participant Parent participant
Age
Mean (SD) 15.09 (1.66) 43.66 (6.86)
Range 13e18 27e52
Sex
% Males 39.1 13.0
Race
% Black/African American 44.8 46.1
% White/European American 35.6 38.7
% Latinx 9.8 8.5
% Asian/Asian American 7.4 6.3
% Native American 2.4 0.4
Family socioeconomic status
% Low income 62.0 62.0
Region in the United States
% Northeast 58.0 58.0
% South 24.6 24.6
% Midwest 10.3 10.3
% West 7.1 7.1
M.-T. Wang et al. / Journal of Adolescent Health xxx (2021) 1e83
and child daily affect using longitudinal multilevel modeling in
Mplus with daily observations (level 1) nested within partici-
pants (level 2). The outcomes of interest were children’s positive
and negative affect at level 1. The mediators were parent-child
conflict and parental warmth at level 1. The key predictors
were job loss and WFH at level 2. We first tested main and
mediation effects and then investigated whether these effects
varied by family socioeconomic status (SES). For the mediation
analysis, we used the “Model Constraint”command to estimate
indirect effects (i.e., a*b) using main effects for “a”(i.e., inde-
pendent variable /mediator) and “b”(i.e., mediator /
dependent variable) paths. For our moderation analysis, we
performed multigroup chi-square test comparisons to examine
whether we could constrain the paths for each SES group’s model
to be equivalent without causing a significant decrement in
model fit. All level 1 predictors were person-mean centered. All
analyses were controlled for level 1 and 2 covariates. The intra-
class correlation of the outcome variables justified the use of a
multilevel modeling approach, as 60%e70% of each outcome’s
variance was at the person level relative to the 30%e40% at the
daily level.
Missing data
The amount of missing data at both the daily and person
levels was relatively low. Of the possible 6,705 daily-diary as-
sessments (15 days, 447 children and parents), there was only
6.6% (n¼441) and 5.2% (n¼350) missing data at the daily level
for children and parents, respectively. There were varying levels
of missing data at the family level: 100% of families (i.e., both
parent and child) completed the baseline and demographic
surveys; 62% of children and 72% of parents did not miss any
daily-diary entries; 28% of children and 21% of parents missed
one or two daily entries; and 6% of children and 4% of parents
missed 3e4 daily entries. On average, both children and parents
each completed 14 of 15 daily-diary entries.
Little’s missing completely at random test suggested that the
data were missing completely at random,
c
2
(20) ¼31.10, p¼ns.
An examination of missing data patterns indicated that adoles-
cents or parents with complete data did not differ from those
with missing data on key constructs or demographic character-
istics. To retain all participants in analyses, we accounted for
missing data through full-information maximum likelihood
estimation.
Results
Table 2 presents means, standard deviations, and correlations
among all study constructs. Table 3 presents the mean differ-
ences by family socioeconomic status. Notably, parents of low-
income families were more likely to experience job loss (24%
vs. 13%) and less likely to WFH (44% vs. 73%) than middle-high
income parents during COVID-19. In addition, low-income fam-
ilies tended to experience more parent-child conflict and less
parental warmth than middle-high income families.
Direct effect of employment status on family relationship
As shown in Table 4, parental job loss predicted increases in
parent-child conflict (B¼.15, SE ¼.07, p<.05, 95% confidence
interval [CI] [.01, .29], ES ¼.12). In addition, WFH predicted
increases in parental warmth (B¼.23, SE ¼.08, p<.01, 95% CI
[.07, .38], ES ¼.15).
Direct effect of family relationship on child affect
Parent-child conflict predicted increases in children’s nega-
tive affect (B¼.14, SE ¼.02, p<.001, 95% CI [.09, .18], ES ¼.14)
and decreases in children’s positive affect (B¼.16, SE ¼.02,
p<.001, 95% CI [.20, .11], ES ¼.11; see Table 4). Moreover,
parental warmth predicted increases in positive affect (B¼.15,
SE ¼.02, p<.001, 95% CI [.11, .18], ES ¼.17) and decreases in
negative affect for children (B¼.04, SE ¼.01, p<.001, 95% CI
[.06, .02], ES ¼.07).
Mediation effect of family relationship on the link between
employment status and child affect
As shown in Table 5, parents who experienced job loss
demonstrated increases in parent-child conflict, which in turn
predicted decreases in child positive affect (indirect effect:
B¼.02, SE ¼.01, p<.05, 95% CI [.04, .01]) and increases in
child negative affect (indirect effect: B¼.02, SE ¼.01, p<.05, 95%
CI [.01, .5]). Furthermore, parents’WFH status predicted in-
creases in parental warmth, which in turn predicted increases in
child positive affect (indirect effect: B¼.04, SE ¼.01, p<.01, 95%
CI [.01, .07]) and decreases in child negative affect (indirect effect:
B¼.01, SE ¼.00, p<.05, 95% CI [.02, .01]). The model fit the
data well,
c
2
(6) ¼80.93, p<.001, RMSEA ¼.05, CFI ¼.93,
SRMR
within
¼.01, SRMR
between
¼.04.
Moderation effect of socioeconomic status differences
Although the means of our key constructs varied by family
SES, family SES did not moderate the links between these con-
structs,
c
2
(12) ¼8.82, p¼ns.
Discussion
COVID-19 precipitated unparalleled social and economic
disruptions in family life, exacting an especially heavy toll on
low-income families [3]. Because of soaring unemployment rates
and rapid, widespread transitions to WFH arrangements, it is
imperative to understand how parents’employment status
during the pandemic has influenced family relationships and
children’s emotional well-being. Using 15 consecutive days of
dyadic parent-adolescent data, we found that parental
employment changes (i.e., job loss and WFH) influenced family
relationships (i.e., parent-adolescent conflict and parental
warmth), and family relationships were connected to adoles-
cents’daily positive and negative affect. Family relationships also
mediated longitudinal links between parents’employment sta-
tus and daily affect among adolescents. These findings not only
highlight family factors shaping adolescents’affective states but
they also reveal psychosocial consequences of parental job loss
and WFH arrangements during COVID-19.
Our results suggest that family stress processes may underlie
associations between parental job loss and curtailed family
functioning. Indeed, numerous studies have indicated that
heightened economic pressure incites parental distress that fo-
ments harsh parenting and increased antagonism between par-
ents and children [9,22,23]. For example, recent work has
indicated that parents who experience COVID-related job losses
M.-T. Wang et al. / Journal of Adolescent Health xxx (2021) 1e84
were nearly five times more likely to psychologically abuse their
children [24]. Family economic stress related to job loss also
imperils the family system by stimulating coercive family
relations [25,26].
Unlike job loss, we found that WFH was connected to
increased parental warmth. It is likely that WFH helped parents
to minimize COVID-19 infection risks and supervise children
themselves during school closures, thereby alleviating the
stress of grappling with alternative childcare arrangements or
being unable to monitor children during remote schooling.
Additional protective factors related to WFH (e.g., financial
stability, job security, professional autonomy, schedule flexi-
bility) may have buffered parents against psychological distress
and allowed them to avoid decompensating parenting behav-
iors [27,28]. Some parents have even reported finding
emotional respite in spending more time with their children
during the pandemic [29]. Despite the stress of integrating
family and professional life resultant from unexpected transi-
tions to WFH, such working arrangements may confer
protection and contribute to family resiliency via the opportu-
nity for increased warmth and togetherness.
Consistent links between family relationships and adoles-
cents’emotional well-being also emerged in our analyses. Family
systems research emphasizes the importance of nurturing and
supportive parenting for youth’s affective states [30,31]. Parent-
adolescent conflict has been shown to decrease adolescents’use
of parental social supports in times of crisis, ultimately deterio-
rating relationship quality and posing negative consequences for
adolescent’s emotional well-being [25,32]. High, stable parental
warmth may therefore act as an efficacious protective factor
against negative affect and distress arising from family stressors
(e.g., parental job loss) [1,33].
Although family economic status did not moderate the direct
effects of parental employment status or the mediation effects of
family relationships on adolescent affect, our findings nonethe-
less have major implications for economically disadvantaged
families. In our sample, low-income families were 67% less likely
to enter WFH arrangements and twice as likely to experience job
Table 2
Descriptive statistics and zero-order bivariate correlations among key constructs
Variable 1234567Mean (S.D.)
Within-Person
1 Day 1 8.00 (4.32)
2 Weekend .28** 1 .26 (.44)
3 Child positive affect .03*.01 1 3.24 (1.05)
4 Child negative affect .04** .01 .18** 1 1.56 (.75)
5 Parent-child conflict .08** .03** .06*.37** 1 1.36 (.61)
6 Parental warmth .05*.05** .53** .15** .04** 1 2.97 (1.04)
7 Parent psychological distress .03*.03*.13** .32** .23** .09** 1 1.49 (.65)
Between-Person 1234567891011
1 Low-income family 1 .62
2 Parent lost job .10** 1 .20
3 Parent work from home .34** .32** 1 .58
4 Child’s age .12** .10** .10** 1 15.09 (1.66)
5 Male child versus female child .01 .09** .02 .18** 1 .39
6 Black child versus white child .34** .02 .16** .14** .03*1 .45
7 Other race child versus white child .06** .09** .11** .04** .09** .47** 1 .22
8 Father versus mother .04** .08** .04** .10** .02 .08** .12** 1 .13
9 Black parent versus white parent .36** .01 .18** .13** .03** .89** .32** .02 1 .46
10 Other race parent versus White parent .09** .09** .11** .05** .07** .33** .70** .04** .42** 1 .17
11 Prior parent-child relationship .06** .03*.09** .10** .07** .14** .07** .01 .12** .09** 1 4.22 (.91)
*p<.05, **p<.01, ***p<.001.
Table 3
Mean group differences by family income
Key variable Low-income household
62% of analytic sample
Middle-high income household
38% of analytic sample
Independent samples t-tests/
c
2
test comparisons
Child positive affect 3.24 (1.08) 3.22 (1.01) t (5,468) ¼.73, p¼ns
Child negative affect 1.55 (.78) 1.56 (.70) t (5,692) ¼.60, p¼ns
Parent-child conflict 1.39 (.65) 1.32 (.53) t (6,073) ¼4.66, p<.001
Parental warmth 2.95 (1.08) 3.00 (.97) t (5,767) ¼2.12, p<.05
Parent psychological distress 1.52 (.70) 1.45 (.57) t (6,140) ¼4.73, p<.001
% Parent lost job 23.90 12.70
c
2
(1) ¼95.53, p<.001
% Parent work from home 44.00 73.40
c
2
(1) ¼26.19, p<.001
Child’s age 14.93 (1.64) 15.34 (1.66) t(447) ¼2.56, p<.05
% Male child versus female child 38.80 39.80
c
2
(1) ¼.04, p¼ns
% Black child versus white child 57.60 22.80
c
2
(1) ¼51.83, p<.001
% Other race child versus white child 19.90 25.10
c
2
(1) ¼1.68, p¼ns
% Father versus mother 11.80 14.40
c
2
(1) ¼.64, p¼ns
% Black parent versus white parent 60.00 23.60
c
2
(1) ¼57.35 p<.001
% Other race parent versus white parent 14.30 21.30
c
2
(1) ¼3.71, p¼ns
Prepandemic parent-child relationship 4.16 (.96) 4.33 (.82) t(402) ¼2.04, p<.05
M.-T. Wang et al. / Journal of Adolescent Health xxx (2021) 1e85
loss than families with more economic resources. Our results also
show that levels of parent-adolescent conflict were higher
among low-income families. Low-income families are more
likely to work in hourly and essential service sector jobs, and
during the pandemic, these families were more vulnerable to
layoffs and furloughs and less likely to WFH as compared with
office-based, white-collar roles [3,34]. Furthermore, many
economically disadvantaged parents who escaped job loss dur-
ing the pandemic worked in essential roles (i.e., “essential
workers”) that required them to place themselves at risk for
COVID-19 infection. These contextual hardships can quickly
accumulate in ways that undermine family functioning and
emotional well-being [2,4,9,35]: As job loss foments negative
family relationships and WFH fosters parental warmth,
economically disadvantaged families confront more contextual
adversities and less access to protective factors.
Limitations
This work is a first step into understanding family dynamics
and adolescent emotional well-being within the context of a
pandemic-related economic crisis. Although this longitudinal
study highlights the key role family relationships play in linking
COVID-19’s economic impacts to adolescents’emotions, these
data were collected during the nascent stages of the pandemic.
As family well-being may change over the course of the
pandemic, future studies should investigate whether and how
our findings persist, attenuate, or strengthen over time. In
addition, researchers should focus on increasing generalizability
by using more racially proportionate, geographically diverse
samples, as our nationwide sample was not representative of the
national population, nor could these results be applied to youth
from states where stay-at-home orders were not enacted.
Because the daily-diary assessment occurred once per day for
15 days, future studies should consider collecting information
about the within-day sequencing of events over a longer period
of time. In addition, we did not use multidimensional or clinical
scales to measure family dynamics or psychological adjustment
as a means of minimizing daily participant burden.
As is the case in all nonexperimental research, we cannot
firmly establish causal relationships among the patterns of
associations observed in this study, nor can we definitively
attribute changes in children’s emotional states to parent-child
relational dynamics. Children’s characteristics and parenting
are often reciprocally related. Analyzing extensive longitudinal
data collected over an extended period of time in a cross-lagged
Table 5
Direct and indirect effects of family employment status and family relationship on child affect
Predictors Child positive affect outcome Child negative affect outcome
Parent lost job Parent-child conflict as the
mediator
Parental warmth as the
mediator
Parent-child conflict as
the mediator
Parental warmth as the
mediator
B (SE) 95% CI B (SE) 95% CI B (SE) 95% CI B (SE) 95% CI
Parent lost job /Mediator .15 (.07)*.01, .29 .05 (.11) .16, .26 - - - -
Mediator /child affect .16 (.02)*** .20, .11 .15 (.02)*** .11, .18 .14 (.02)*** .09, .18 .04 (.01)*** .06, .02
Parent lost job /child affect .09 (.10) .29, .10 - - .28 (.09)** .11, .46 - -
Parent lost job /mediator /Outcome .02 (.01)*.04, .01 .01 (.02) .03, .05 .02 (.01)*.01, .05 .00 (.01) .01, .01
Parent lost job /child affect .06 (.11) .28, .16 - - .39 (.10)*** .20, .59 - -
Parent works from home (WFH)
WFH /mediator .09 (.05) .01, .19 .23 (.08)** .07, .38 - - - -
Mediator /outcome .16 (.02)*** .20, .11 .15 (.02)*** .11, .18 .14 (.02)*** .09, .18 .04 (.01)*** .06, .02
WFH /outcome .06 (.07) .08, .21 - - .01 (.06) .13, .10 - -
WFH /mediator /outcome .01 (.01) .03, .01 .04 (.01)** .01, .07 .01 (.01) .01, .03 .01 (.00)*.02, .01
WFH /outcome .23 (.09)*.04, .41 - - .02 (.07) .11, .15 - -
*p<.05, **p<.01, ***p<.001.
Table 4
Direct effect of family employment status and family relationship on child affect
Predictor Child positive affect Child negative affect Parent-child conflict Parental warmth
Within person
Day .01 (.00)** .01 (.01)
Weekend .02 (.02) .01 (.01)
Parent-child conflict .16 (.02)*** .15 (.02)***
Parental warmth .15 (.02)*** .04 (.01)***
Between person
Low-income family .04 (.08) .06 (.06) .10 (.05)*.03 (.08)
Lost job .11 (.10) .37 (.10)*** .15 (.07)*.04 (.11)
Work from home .06 (.07) .04 (.06) .09 (.05) .23 (.08)**
Child age .02 (.02) .05 (.02)** .01 (.02) .07 (.02)**
Child boy versus girl .01 (.07) .19 (.06)** .02 (.05) .09 (.07)
Black child versus white child .12 (.21) .20 (.16) .02 (.10) .27 (.18)
Other race child versus white child .04 (.16) .13 (.11) .07 (.06) .06 (.13)
Father versus mother .03 (.10) .03 (.10) .05 (.05) .04 (.10)
Black parent versus white parent .34 (.21) .03 (.15) .02 (.09) .23 (.17)
Other race parent versus white parent .05 (.16) .03 (.11) .01 (.08) .27 (.15)
Prior parent-child relationship .02 (.05) .02 (.05) .14 (.04)*** .37 (.05)***
*p<.05, **p<.01, ***p<.001.
M.-T. Wang et al. / Journal of Adolescent Health xxx (2021) 1e86
panel framework could yield additional insights about the di-
rection of effects. Relatedly, we cannot rule out the possibility
that other parental characteristics (e.g., parenting style)
contributed to family processes. Future research may benefit
from assessing other dimensions of family relationships or
parental characteristics as potential moderators or mediators
and using prepandemic comparison data to determine differ-
ences in prepandemic and peripandemic parenting. Researchers
interested in continuing this line of inquiry should also incor-
porate clinical evaluations to explore adolescents’psychological
well-being during COVID-19 more thoroughly.
Using multiinformant daily-diary data with a racially and
socioeconomically diverse U.S. sample, this study showed that
COVID-19’s effect on parental employment status shaped family
processes which in turn influenced adolescent emotional
adjustment. With emerging evidence showing increased mental
health issues among U.S. adolescents during COVID-19 [36], the
need for research that identifies key risk and protective factors
for youth well-being persists. The American Academy of Pedi-
atrics policy guidance emphasizes the central role pediatricians
play in promoting children’s socioemotional health, and re-
searchers and policy makers continue to advocate for mental
health screening and referrals in schools [37,38]. Clinicians and
schools can screen youth for signs of psychological distress and
support parents by providing information about best practices
for managing stress, dealing with family conflict, and meeting
their children’s emotional needs.
More broadly, a multipronged policy response is needed to
address the negative societal impacts fueled by COVID-19 [39]:
At-risk families need substantive financial and health supports
that stabilize income for parents who lost their jobs and increase
access to school and community-based socioemotional supports.
Although flexible and remote work options are not feasible for all
working parents, public policies and workplace benefits that
buoy parental well-being, such as paid time off, livable wages,
and adequate health care, may improve the qualitydif not the
quantitydof parents’time with their children, thereby
enhancing adolescent emotional well-being [40]. By focusing on
macrostructural factors and family-level processes, policy
makers and practitioners can promote adolescents’well-being,
even in the face of a pandemic.
Supplementary Data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.jadohealth.2021.07.016.
References
[1] Russell BS, Hutchison M, Tambling R, et al. Initial challenges of caregiving
during COVID-19: Caregiver burden, mental health, and the parentechild
relationship. Child Psychiatry Hum Dev 2020;51:671e82.
[2] Gassman-Pines A, Ananat EO, Fitz-Henley J. COVID-19 and parent-child
psychological well-being. Pediatrics 2020;146. e2020007294.
[3] Martin A, Markhvida M, Hallegatte S, Walsh B. Socio-economic impacts of
COVID-19 on household Consumption and poverty. Econ Disasters Clim
Chang 2020;4:453e79.
[4] Kalil A, Mayer S, Shah R. Impact of the COVID-19 crisis on family dynamics
in economically vulnerable households. SSRN Electron J 2020. https://doi.
org/10.2139/ssrn.3706339. Available at: https://papers.ssrn.com/sol3/
papers.cfm?abstract_id¼3706339. Accessed August 21, 2021.
[5] Vaziri H, Casper WJ, Wayne JH, Matthews RA. Changes to the work-family
Interface during the COVID-19 pandemic: Examining predictors and impli-
cations using Latent transition analysis. J Appl Psychol 2020;105:1073e87.
[6] Fegert JM, Vitiello B, Plener PL, Clemens V. Challenges and burden of the
coronavirus 2019 (COVID-19) pandemic for child and adolescent mental
health: A narrative review to highlight clinical and research needs in the
acute phase and the long return to normality. Child Adolesc Psychiatry
Ment Health 2020;14:20.
[7] Prime H, Wade M, Browne DT. Risk and resilience in family well-being
during the COVID-19 pandemic. Am Psychol 2020;75:631e43.
[8] Masten AS. Ordinary Magic: Resilience in development. New York, NY: The
Guilford Press; 2014.
[9] Kalil A, Wightman P. Parental job loss and family conflict. Bowling Green,
OH: National Center for Family & Marriage Research at Bowling Green State
University; 2010. Working Paper Series: WP-10-07.
[10] Acquah D, Sellers R, Stock L, Harold G. Inter-parental conflict and outcomes
for children in the contexts of poverty and economic pressure. London: Early
Intervention Foundation; 2017. Available at: www.eif.org.uk/files/pdf/
interparental-conflict-children-poverty-economic-pressure.pdf. Accessed
August 21, 2021.
[11] Weymouth BB, Buehler C, Zhou N, Henson RA. A meta-analysis of parent-
adolescent conflict: Disagreement, hostility, and youth maladjustment.
J Fam Theor Rev 2016;8:95e112.
[12] Silva K, Ford CA, Miller VA. Daily parent-teen conflict and parent and
adolescent well-being: The moderating role of daily and person-level
warmth. J Youth Adolesc 2020;49:1601e16.
[13] Allen TD, Johnson RC, Kiburz KM, Shockley KM. Work-family conflict and
flexible work arrangements: Deconstructing flexibility. Pers Psychol 2013;
66:345e76.
[14] Kelly EL, Moen P, Tranby E. Changing workplaces to reduce work-family
conflict: Schedule control in a white-collar organization. Am Sociol Rev
2011;76:265e90.
[15] Wadsworth ME, Rindlaub L, Hurwich-Reiss E, et al. A lo ngitudinal examination
of the Adaptation to Poverty-Related Stress Model: Predicting child and
adolescent adjustment over time. J Clin Child Adolesc Psychol 2013;42: 713e25.
[16] Bolger N, Davis A, Rafaeli E. Diary methods: Capturing life as it is lived.
Annu Rev Psychol 2003;54:579e616.
[17] Heuchert JP, McNair DM. Profile of mood states e2nd edition (POMS-2).
2nd ed. North Tonawanda, NY: Multi-Health Systems, Inc.; 2012.
[18] Cranford JA, Shrout PE, Iida M, et al. A procedure for evaluating sensitivity
to within-person change: Can mood measures in diary studies detect
change reliably? Personal Soc Psychol Bull 2006;32:917e29.
[19] Furman W, Buhrmester D. Methods and measures: The network of re-
lationships inventory: Behavioral systems version. Int J Behav Dev 2009;
33:470e8.
[20] Dearing E, McCartney K, Taylor BA. Change in family income-to-needs
matters more for children with less. Child Dev 2001;72:1779e93.
[21] Duncan GJ, Magnuson K, Votruba-Drzal E. Children and socioeconomic
status. In: Handbook of child Psychology and developmental Science.
Hoboken, NJ: John Wiley & Sons, Inc.; 2015:1e40.
[22] Conger RD, Wallace LE, Sun Y, et al. Economic pressure in African American
families: A replication and extension of the family stress model. Dev Psy-
chol 2002;38:179e93.
[23] McLoyd VC. The impact of economic hardship on Black families and chil-
dren: Psychological distress, parenting, and socioemotional development.
Child Dev 1990;61:311e46.
[24] Lawson M, Piel MH, Simon M. Child maltreatment during the COVID-19
pandemic: Consequences of parental job loss on psychological and Phys-
ical abuse towards children. Child Abuse Negl 2020;110:104709.
[25] Neppl TK, Senia JM, Donnellan MB, Bonnellan MB. The effects of economic
hardship: Testing the family stress model over time. J Fam Psychol 2016;
176:139e48.
[26] Wadsworth ME, Compas BE. Coping with family conflict and economic
strain: The adolescent perspective. J Res Adolesc 2002;12:243e74.
[27] Heinrich CJ. Parents’employment and children’s wellbeing. Futur Child
2014;24:121e46.
[28] Perry-Jenkins M, Laws HB, Sayer A, Newkirk K. Parents’work and children’s
development: A longitudinal investigation of working-class families. J Fam
Psychol 2019;34:257e68.
[29] McCrory CalarcoJ, Meanwell E, Anderson E, Knopf A. “Let’snot pretend it’sfun”:
How COVID-19-related school and childcare closures are camaging mothers’
well-being. OSF Prepr2020:1e28.Available at:https://doi.org/10.31235/osf.io/
jyvk4. Available at: https://osf.io/preprints/socarxiv/jyvk4/. Accessed August
21, 2021.
[30] Khaleque A. Perceived parental warmth, and children’s psychological
adjustment, and personality dispositions: A meta-analysis. J Child Fam
Stud 2013;22:297e306.
[31] Luebbe AM, Bell DJ. Positive and negative family emotional climate
differentially predict youth anxiety and depression via distinct affective
pathways. J Abnorm Child Psychol 2014;42:897e911.
[32] Brooks-Gunn J, Schneider W, Waldfogel J. The great recession and the risk
for child maltreatment. Child Abus Negl 2013;37:721e9.
[33] Brown SM, Doom JR, Lechuga-Peña S, et al. Stress and parenting during the
global COVID-19 pandemic. Child Abus Negl 2020;110:1e14.
M.-T. Wang et al. / Journal of Adolescent Health xxx (2021) 1e87
[34] Kochhar R, Barroso A. Young workers likely to Be hard Hit as COVID-19
Strikes a Blow to Restraurants and other service sector jobs. PEW Research;
2020. Available at: http://www.pewresearch.org/fact-tank/2018/10/19/5-
charts-on-global-views-of-china/. Accessed August 21, 2021.
[35] Gassman-Pines A, Oltmans Ananat E, Gibson-Davis CM. Effects of statewide
job losses on adolescent suicide-Related behaviors. Am J Public Health
2014;104:1964e70.
[36] Hussong AM, Midgette AJ, Thomas TE, et al. Coping and mental health in
early adolescence during COVID-19. Res Child Adolesc Psychopathol 2021;
49:1113e23.
[37] Foy JM, Green CM, Earls MF. Mental health competencies for pediatric
practice. Pediatrics 2019;144:1e18.
[38] Husky MM, Sheridan M, McGuire L, Olfson M. Mental health screening and
follow-up care in public high schools. J Am Acad Child Adolesc Psychiatry
2011;50:881e91.
[39] Dooley DG, Bandealy A, Tschudy MM. Low-income children and
coronavirus disease 2019 (COVID-19) in the US. JAMA Pediatr 2020;174:
922e3.
[40] MilkieMA, NomaguchiKM, Denny KE. Doesthe amount of timemothers spend
with children or adolescents matter? J Marriage Fam 2015;77:355e72.
M.-T. Wang et al. / Journal of Adolescent Health xxx (2021) 1e88
