Introduction

Expulsion from early care and education (ECE) settings throughout the United States has been a topic of interest in recent years. Gilliam initially presented the issue in 2005 when he reported that early childhood expulsion occurred at a rate three times higher than with school-age children (Gilliam, 2010). Exclusion of a preschool child from an educational setting is associated with future negative educational and social-emotional outcomes (Noltemeyer et al., 2015). Giordano and colleagues’ (2022) study explored expulsion from community childcare centers during the first year of the COVID-19 pandemic and found a reduction in expulsion rates when compared to pre-pandemic rates, despite the disruption of routines and levels of stress that occurred at during this time period. These counterintuitive results led to the current study, where we examined expulsion from community childcare centers during the second year of the COVID-19 pandemic, as vaccine rates increased, restrictions were removed, and centers were able to operate more fully.

Early Childhood Expulsion

Expulsion involves the permanent removal of a child from an early learning setting (Gilliam & Shahar, 2006). As ECE enrollment is not mandated, young children without diagnosed disabilities are not entitled to access to education (Loomis et al., 2021) and can be dismissed from many early learning settings. Prior research has demonstrated notable demographic discrepancies within preschool expulsion rates. Giordano and colleagues (2020) found that Black preschoolers who were enrolled in community childcare programs were at a greater risk of being expelled than their counterparts who attend public schools. Findings from this same study indicated that children who are Hispanic/Latino/Spanish (Hispanic) were twice as likely to be expelled than their peers within community childcare settings. Gender and age discrepancies have also been observed with regard to preschool expulsion rates (Gilliam, 2005). Older preschoolers were at greater risk of being expelled than younger preschoolers, indicating that age was also a factor that played into expulsion decisions (Giordano, McKeating Giordano et al., 2022). More specifically, four-year olds were 50% more likely than children two or three years of age to be expelled, while five-year olds and six-year olds were twice as likely to be expelled (Gilliam, 2005).

Suspension has been described as a set amount of time when a child is temporarily removed from an educational setting (Gilliam & Shahar, 2006). A 2016 report indicated that Black preschoolers received out-of-school suspension at 3.6 times the rate of their White classmates (U.S. Department of Health & Human Services, 2016). Giordano and colleagues (2020) calculated the risk ratio of suspension for Black preschoolers and determined that they are almost three and a half times more likely than students of all other racial backgrounds to be suspended.

Soft expulsion occurs when a program suggests that the child should leave, such as implying that the child is not a match, that they are not ready for school, or otherwise communicates to the child or caregivers that this child is unwanted in the program, without requiring them to leave (SRI International, 2016). Numbers surrounding soft expulsion are difficult to evaluate because children who are treated in this way and leave by their caregiver’s choice are not included in the reported expulsion data. It seems that when including instances of soft expulsion, rates of expulsion may be significantly higher than typically disclosed. Interestingly, there is no evidence in the literature that suggests that suspension or expulsion decreases challenging behavior. After a child is expelled, their families often re-enroll them in other ECE settings which, without appropriate supports, often resulting in repeated expulsions for the child (Loomis et al., 2021).

Program Quality and Expulsion

Quality may be measured in a variety of ways depending on who is assessing the quality (Layzer & Goodson, 2006). Teachers may focus on aspects such as employment conditions or class size, while researchers might be more likely to focus on things such as staff-child ratios and teacher training as indicators of quality. A teacher’s educational level, credentials, or years of experience teaching young children are reliable predictors of expulsion and preschool teachers with more years of experience tend to foster more positive relationships with students (Gilliam, 2008). Other factors associated with quality found to predict expulsion from a program are quality rating and improvement system (QRIS) participation, beliefs about expulsion, suggesting that the child is not a match for the program, and teacher:child ratios (Giordano, Goldberg et al., 2022). However, multiple factors that make up quality are not guaranteed to be a protective factor against expulsion. Giordano, Goldberg and colleagues (2022) also found that other factors related to program quality (accreditation, administrator education and experience, support available for children with behavior challenges) did not predict expulsion.

Program location has also been shown to be related to program quality. Maher and colleagues (2008) found that infants who attended centers in rural areas had higher child:teacher ratios than their urban counterparts, where preschool students were found to have lower ratios comparatively. These researchers also found that children who receive childcare subsidies tend to have higher child:teacher ratios, regardless of setting. A higher number of children per teacher predicts an elevated likelihood of expulsion (Gilliam, 2008), therefore ratios and center capacity also seem to play a role in suspension and expulsion rates.

COVID-19 and Childcare

During the COVID-19 pandemic, childcare centers did not operate as they previously had. A series of state-wide regulations were put in place to maintain safety for both the children and their teachers. In the state where this study was conducted, in March 2020, there were executive orders put in place in response to the influx of COVID-19 cases that closed childcare centers throughout the state (Exec. Order No. 110, 2020b). As part of Executive Order 110, the governor determined that programs providing care for the children of essential workers would be paid a set stipend weekly per enrolled child (Exec. Order No. 110, 2020b). In June 2020, all childcare centers were allowed to reopen, however, there were a series of safety and health requirements that were put into place during this time. The regulations for how childcare centers could run as of June 2020 included: daily temperature checks, distancing requirements, limits of 10 or fewer children in a group with no mixing of groups, both adults and children had to wear masks, and children were prohibited from sharing toys and materials (Exec. Order No. 149, 2020a). When these regulations were at their strictest, the childcare programs adapted in order to meet these requirements. For example, in order to minimize the sharing of toys and materials, childcare centers sometimes prepared individual sets of materials and stored them in the children’s personal spaces or cubbies. As of October 2022, public health recommendations for those in childcare settings were again updated. Masks were no longer required for children, staff, or visitors of childcare facilities, since individuals six months and older were eligible for the COVID-19 vaccine, staff and students were recommended to stay “up to date” with their vaccinations, and both students and staff were encouraged to stay home when sick (New Jersey Department of Health, 2022).

According to the literature, during the COVID-19 pandemic, expulsion rates were lower than they had been before the pandemic. Prior to the pandemic, research found that 35.7% of the childcare centers surveyed had expelled a child in the past year (Giordano et al., 2020). In a follow up study that occurred in November 2020, eight months after initial closures and five months after the centers reopened, 17.4% of centers reported expelling at least one child. Out of the 161 programs that were surveyed, only 34 children were expelled during the height of the pandemic (Giordano, McKeating Giordano et al., 2022). In this same study, many directors reported they did not see a change in either the intensity or frequency of children’s challenging behavior compared to a year prior to the pandemic (Giordano, McKeating Giordano et al., 2022). A smaller percentage than expected indicated that the behavior was in fact worse than prior to the pandemic.

COVID-19 and Stress

There were many new demands placed on teachers during the pandemic concerning new safety policies and mandates, in addition to the stress of maneuvering the unknown world, supporting students in their care, and maintaining their own health and well-being. Job stress and burnout during COVID-19 affected most professions, but it affected teachers and school administrators particularly hard (Kush et al., 2022). Depending on where a school was located, more or less support may have been necessary during the COVID-19 pandemic to enable it to function smoothly. Support may have been technology-based, instructional, or emotional (Pressley & Ha, 2022). Many teachers were not provided with the necessary support, as the majority of schools were forced to shut down suddenly and transition to remote or hybrid learning, leaving educators stressed and scrambling to transition curricula to an online structure. Stress and burnout in jobs occurs when there is an appraised mismatch between job demands and resources (Bakker & Demerouti, 2007), meaning a job is demanding more responsibilities of someone than is possible based on the resources available. Pressley and Ha (2022) found that the level of administrative support, as well as anxiety about COVID-19, current teaching, and communication with parents, were factors that predicted teacher burnout. Professionals who care directly for children were more prone to compounded stressors, causing them challenges when attempting to identify and solve the problems that lead to stress (Baumgartner et al., 2009). High rates of burnout during the COVID-19 pandemic led to many teachers leaving their positions and not returning (Pressley & Ha, 2022).

Present Study

The current study revisited early childhood expulsion rates in community childcare programs during the COVID-19 pandemic (May 2022), approximately two years after childcare centers in this state reopened. This research follows a previous study that examined expulsion practices in one state’s community childcare centers during the first year of the pandemic (Giordano, McKeating et al., 2022). In the present study, childcare program administrators were asked about: the frequency and intensity of children’s challenging behaviors; whether children were asked to leave the program for a number of reasons; the number of staff members; perceived stress levels of staff; COVID-19 program alterations (and which of these programs intended to keep); training provided to teaching staff; the presence of a waiting list; and overall descriptions of the program-type and population it served. The present study aimed to address two main research questions:

  1. 1.

    What patterns are seen in expulsion two years into the COVID-19 pandemic? and

  2. 2.

    What factors predicted expulsion?

Method

Procedures

After IRB approval was obtained, researchers accessed a publicly available list of licensed childcare programs in the state. Those clearly identified as public schools or as only providing care to children six years old or older (after-school programs, for example) were removed, leaving 2,800 email addresses. It was unknown at the time of distribution how many of these programs were closed, either temporarily or permanently, due to the pandemic. For this reason, it was not possible to determine a response rate. Program directors were sent a recruitment email explaining that researchers were interested in examining changes in practices responding to children’s challenging behavior in the state during the pandemic and inviting them to complete an online, anonymous survey on the topic. No incentives were offered for participation. The survey remained open for two weeks and email reminders were sent after one week.

Participants

Initially, 176 surveys were completed; data from those who did not meet inclusion criteria (working in a community childcare program and responding to the question asking whether or not a child was expelled) were removed, leaving data from 131 participants for analysis. Participants represented 20 of the state’s 21 counties and primarily ran programs in suburban (n = 74, 56.5%) or urban (n = 40, 30.5%) areas. Most programs reported serving families in the lower (n = 43, 32.8%) or middle (n = 47, 35.9%) socioeconomic range. Director’s reported providing care to infants through pre-kindergarten (pre-k) (n = 49, 37.4%) or toddlers through pre-k (n = 24, 18.3%). Some provided care only to preschool/pre-k (n = 37, 28.2%) and only 3 (2.3%) provided care exclusively to infants and toddlers. Participants reported a mean enrollment capacity of 103.14 children (range 12–394) and a current enrollment of 74.83 children (range 8-275). Only eight reported capacity that matched enrollment; the remaining programs had lower enrollments than allowed capacity. A slight majority of participants reported having a waitlist (n = 68, 51.9%), suggesting that some programs intentionally enrolled to a number lower than their licensed capacity.

Measures

The authors developed a multiple-choice, online, anonymous survey examining expulsion and suspension practices two years into the COVID-19 pandemic. Participants were permitted to skip questions, and the median survey completion time was 11.32 min. One set of questions examined current program enrollment, the existence of wait lists, and training being provided to staff members. Participants were asked which COVID-19 protocols they planned on keeping even when no longer required. Participants were prompted to report the number of children suspended and expelled in the past 12 months and to provide the reasons for these decisions. Demographic data (i.e., age, race, gender) were requested for each child who was expelled. The survey asked questions regarding how the frequency and intensity of behavior had changed since before the pandemic and if there was support available if a child displayed challenging behaviors. Finally, participants were asked to provide information on their own and on their teaching staffs’ level of stress and levels of staff turnover within their programs.

Results

Expulsion

Data from 131 participants were included in the analysis. Missing or unknown responses were eliminated from individual analyses. When it came to expulsion, 44 (33.6%) participants stated that they had asked at least one child to permanently leave their program during the year the study was conducted. Of those who did expel, the majority (n = 28, 63.6%) expelled one child, followed by two children (n = 13, 29.5%). Only three programs (6.8%) expelled three children and one (2.3%) expelled four. No program reported expelling 5 or more children. Overall, at least 67 individual children were reported expelled across the 131 participating programs in the academic year. Demographic information (i.e., age, race, gender) was collected on children who were expelled (see Table 1). This information was not required and six to nine participants, depending on the question, did not provide this information. The majority of children were three (n = 20, 29.8%) and four (n = 24, 35.8%) years old, White (n = 24, 35.8%), and male (n = 43, 64.2%). It is noted that although White students represented the highest number of expulsions, this does not take proportionality into account. As there were not enough students expelled in non-White groups to calculate a meaningful risk ratio (Pyramid Equity Project, n.d.), it is not known at this point if these results represent disproportionality.

Table 1 Characteristics of children expelled from programs
Full size table

When asked to provide the main reason for expulsion, the most frequently reported reason was The child was displaying challenging behavior that did not respond to typical discipline techniques (n = 16, 23.9%), followed by The child had uncontrollable temper tantrums (n = 11, 16.4%). The least frequently cited reasons had to do with parental behaviors (The parent failed to pay tuition (n = 3, 4.5%) and The parent demonstrated “behavior problems” (ex: does not adhere to policies, verbally or physically threatens staff, etc.) (n = 4, 6.0%). It is noted that no children were reported expelled due to lack of adherence to COVID related rules and protocols (see Table 2).

Table 2 Reason cited for child expulsion & suspension from programs
Full size table

Finally, we looked at what factors predicted expulsion with a Bonferroni adjusted alpha level of 0.01 (0.05/8) per test. A series of binary logistic regressions were run to examine if the availability of support (B = 0.80, p = 0.06), prior suspensions (B=-0.42, p = 0.29), whether it was suggested that the program was not a match (B=-0.78, p = 0.04), higher reported turnover (B=-0.05, p = 0.91), the presence of waiting lists (B=-0.11, p = 0.79), administrator reported stress (B = 0.25, p = 0.84), enrollment capacity (B = 0.01, p = 0.24), and teacher perceived stress (B = 20.51, p = 0.88) predicted expulsion (yes or no). None of these factors significantly predicted expulsion.

Suspension

Participants were also asked to report the number of times a child was required to be picked up early for reasons other than illness during this academic year (suspended). Most programs reported that they had suspended at least one child (n = 72, 55.0%). Of these, the majority suspended one child (n = 29, 40.3%), followed by two (n = 22, 30.6%) and three (n = 14, 19.4%) children. Three programs suspended four children (4.2%) and only one suspended five or more (1.4%). Overall, at least 136 individual children were suspended from early learning programs during this time. While we did not collect individual demographic data on these children, participants were asked to select the reasons for suspension; note, participants were permitted to select more than one reason. The most frequently reported reasons were The child was displaying challenging behaviors which did not respond to typical discipline techniques (n = 59, 43.4%), and The child was hurting others (n = 55, 40.4%). Similar to expulsion, the least reported reasons had to do with parent behaviors: Parent failed to pay tuition (n = 1, 0.7%); The parent failed to complete required forms (ex: medical documentation, updated application paperwork, etc.) (n = 2, 1.5%); The parent demonstrated “behavior problems” (ex: does not adhere to policies, verbally or physically threatens staff, etc.) (n = 5, 3.7%). (See Table 2). Almost half of participants (n = 57, 43.5%) reported suggesting to a family that the program was not a good match, without forcing them to leave.

Practices & Beliefs

When reporting on the amount and intensity of challenging behaviors, compared to prior to the pandemic, participants were most likely to report that there was more challenging behavior and that this behavior was more intense. Comparing amount and intensity to one year prior, participants were most likely to say that these had stayed the same; although with amount and intensity, participants were not likely to indicate that behavior was less or less intense than it was before the pandemic or 1 year ago. See Table 3. When participants were asked if they had the support when faced with a child with severely challenging behaviors, the majority said no (n = 87, 66.4%). If a participant indicated that they did have support, they were asked in an open-ended question from whom the support was provided. The most frequently reported sources of support included: the director, local school district, training for staff, parents, therapists, and state or local agencies. Less reported sources of support included: a curriculum that addresses behavior, researching solutions (e.g., written materials, webinars), referral to early intervention, and the hiring of additional staff members.

Table 3 Changes in perceptions of behavior across the pandemic
Full size table

When asked about training provided to staff in the past academic year, the most frequently reported topics included health and safety (n = 111, 84.7%), social-emotional learning (n = 37, 61.2%), and teaching and learning (n = 37, 63.4%). Less frequently reported topics included inclusion (n = 33, 25.2%), diversity and equity (n = 31, 23.7%), and self-care (n = 2, 1.5%).

When provided with a list of nine procedural changes that were required due to COVID and asked which providers planned to keep, participants endorsed all nine items. The most frequently endorsed was enhanced sanitization procedures (n = 106, 80.9%) and drop-off and pick-up protocols (n = 103, 78.6%). The least supported was keeping teachers with a single pod of students (n = 22, 16.8%). See Table 4 for all nine practices. In addition to the listed changes, participants also indicated that they will continue to restrict access to their buildings, keep health related policy changes, and continue differing aspects of masking. When provided with the opportunity to explain their answers to this question, the most prevalent theme described changes in drop-off and pick-up policies. They discussed how having drop-off and pick-up outside of the classroom eases the transition, makes the classroom less chaotic, and helps to stop the spread of germs. Some of these participants expanded this response to indicate that families will still be allowed access for special events. Although less mentioned, two additional themes arose in the responses to this question. First, was an appreciation for the increased cleaning and sanitization requirements. While participants acknowledged the increased time and funding needed to adhere to this, they appreciated how this helped to reduce overall illness in their programs. Finally, participants discussed their appreciation of reduced class sizes and enrollment. They reflected on how this reduction allowed for more individual time with children and calmer classrooms. Several participants noted that while they have kept this for now, in the future they will need to increase these numbers again for their programs to remain financially solvent.

Table 4 COVID procedures & practices participants are planning to keep
Full size table

Participants were also asked to report on staff turnover and stress levels. Respondents were fairly evenly split on whether staff turnover was less than (n = 34, 26.0%), more than (n = 41, 31.3%), or the same (n = 44, 33.6%) as it was prior to the pandemic. When asked about their own stress levels, the majority of participants agreed or strongly agreed to currently feeling stressed, that they find their jobs stressful, and that they feel more stressed than they did before the pandemic (see Table 5). Participants were asked to expand upon their responses, and several themes arose. First was a concern about the inability to hire and retain qualified staff. Participants also described how parents have become more short-tempered and demanding. Participants reported needing to adhere to more regulations and do additional work and all of this occurred when policies were regularly changing. They also expressed concern about the consistency in communication of all of these changes. Participants were worried about finances, as many programs were under enrolled. Finally, participants expressed concern about their own health and the lack of work-life balance. They were also asked to report on how they perceived the levels of stress in their teaching staff. They similarly agreed or strongly agreed that teachers currently felt stressed, that they found their jobs stressful, and that they felt more stressed than they did before the pandemic (see Table 5). When asked to explain their responses, participants most frequently indicated that child behavior is the biggest cause of teacher stress. This was followed by concerns about their own health and lack of parental support combined with increasing parent demands. Administrators again described the lack of staffing and increase in mandates and requirements and how these resulted in teachers needing to do more work and work longer hours. Participants also discussed how children were entering programs with overall delays and teachers were stressed about trying to catch them up while still teaching grade and age-appropriate material.

Table 5 Administrator perceptions of stress for themselves and their teaching staff
Full size table

Discussion

As life with COVID-19 became the new normal, vaccines became available, and early learning programs began to reopen and operate more fully, we looked at expulsion rates two years into the pandemic. Prior research (Giordano, McKeating et al., 2022) has indicated a drop in expulsion rates during the first year of the pandemic, reporting a rate of 0.21 children per program (compared with a rate of 0.53 children per program prior to the pandemic). Our study found 67 children reported expelled from 131 programs, a rate of 0.51 children per program; a rate almost the same as the pre-pandemic rate of expulsion. While no suspension rates were reported during the pandemic, pre-pandemic expulsion rates from community childcare centers have been reported as 0.54 children per center (Giordano et al., 2021). Our study found suspension rates to be 1.04 children per center, almost double the previously reported rate. Similar to pre-pandemic results, approximately half of participants also reported that they have suggested that the program is not a good match for a child. It is notable that neither previous suspension or suggestion that the program is not a match for a child predicted expulsion. It may be that programs that use these suspension practices or suggestions that the placement is not a good match are doing so in lieu of expulsion (i.e., soft expulsion), not as steps preceding expulsion.

Behavioral Issues

Results of this study were disappointing as it appears as if exclusionary discipline practices, after experiencing a decline during COVID, have since returned to or exceeded pre-pandemic levels. When looking at the reasons for suspension and expulsion, we found that most children are being suspended and expelled for challenging behaviors, hurting others, and having uncontrollable temper tantrums. This is consistent with what was reported, both prior to and in early stages of the pandemic (Giordano et al., 2021; Giordano, McKeating Giordano et al., 2022). Participants reported that child behavior was worse than it was prior to the pandemic and worse than it was one year into the pandemic. Interestingly, research conducted one year into the pandemic (Giordano, McKeating Giordano et al., 2022) indicated that only 20% of participants felt that behavior was worse than it was prior to the pandemic. The First Five Years Fund (2022) revealed that, presumably the lack of exposure of young children to peers in group settings combined with pandemic-related stressors, have resulted in schools and parents reporting an increase in behavioral issues, which is consistent with results in this study. Research has indicated that outcomes for young children can improve through use of interventions. One intervention that has been to shown to be effective is the implementation of Early Childhood Mental Health Consultation (ECMHC; Hepburn et al., 2013), which demonstrated a lessening of the child’s challenging behaviors while increasing their prosocial behaviors (Loomis et al., 2021). ECMHC often focuses on relationships and the adult’s understanding and responses to behaviors they find challenging. Implementation of ECMHC increases the likelihood that the child can successfully remain in the educational setting, when provided with appropriate social-emotional and behavioral supports.

Adult Stress

The reported rise in challenging behaviors two years into the pandemic may also be associated with adult well-being, as the respondents reported feeling more stressed. The American Medical Association (AMA) and the Centers for Disease Control and Prevention are among the institutions warning of pandemic fatigue, explaining that people may feel exhausted after months of dealing with all of the challenges brought on by the COVID-19 pandemic (Centers for Disease Control and Prevention, 2022). According to the AMA, pandemic fatigue can lead to stress, which over time can result in exhaustion, grief, anger and anxiety (AMA, 2022). Therefore, while our results suggest that stress did not predict expulsion, stress may have increased perceptions of challenging behavior. If teachers and administrators are experiencing higher levels of stress, they may also be perceiving children’s behaviors to be more severe and more intense.

A newly released study of pre-K through 12th grade teachers in the U.S. during COVID found that teachers were 40% more likely to report anxiety symptoms than healthcare workers, 20% more likely than office workers, and 30% more likely than workers in other occupations (Kush et al., 2022). This is consistent with our study in which administrators reported high levels of current stress and also reported perceived high levels of stress for their teaching staff. Gilliam and Shahar (2006) found expulsion was relatively rare in classes where child to adult ratios and teacher stress were low. As stress levels did not predict if administrators expelled a child, perhaps the lower enrollments and ratios reported by administrators served as a protective factor against expulsion.

Continuation of Pandemic Practices

Results were also surprising when participants reported on practices they intended to keep after the pandemic. While some practices, such as increased sanitizing, do not likely have an impact on exclusionary discipline practices, others such as class size, ratios, additional materials, and changes to drop-off procedures might. Many programs reported that they would keep their COVID drop-off procedures, which, for the most part, restricts parent access to buildings. Gilliam has explained that when parents and teachers have a strong relationship, the risk for expulsion decreases (First Things First, 2019). Given changes to drop-off procedures and other restrictions on program access described by participants, one wonders about the impact this might have on parent-teacher relationships. Will this lack of interaction harm their ability to develop positive relationships? Or will less contact provide less opportunities to damage this relationship? Prior to the pandemic, parental behaviors (i.e., parental problem behaviors, failure to pay tuition) were a frequently cited reason for suspension. This seems to have decreased during the pandemic and remains low according to our participants; parent behaviors were the least frequently reported reasons for both suspension and expulsion in the current study. Perhaps the decreased interaction amongst school staff and parents during drop-off routines is related to this change.

On the other hand, restricting parent access to classrooms is counter to best practice, which stresses the importance of communication with families in early childhood settings. The National Association for the Education of Young Children (NAEYC) states in its program accreditation standards booklet,

Programs need to establish an atmosphere that continually invites and includes families in the life of the program in as many ways as possible. This starts with an open-door policy: families should be able to visit any area of the facility at any time during the program’s regular hours of operation. (NAEYC, 2019, pg. 93)

Policies restricting parental access seem to be in direct conflict with this best practice. Anecdotally, we have had teachers report pick-up and drop-off times as opportunities for informal communication with families. Could a reduction in this informal communication contribute to fewer opportunities for the sharing of contextual information relevant to the child’s state, disposition or well-being? An example of this might be an upset that the child experienced prior to coming to school that may impact their ability to regulate their behavior. If the teacher is aware that the child is having difficulty, they may experience the child’s responses differently and respond in kind, reducing the likelihood of an escalation of behavior. This decrease in parent-teacher communication is one potential explanation for the sharp increase in suspensions seen two years into the pandemic. It will be interesting to see what happens downstream if opportunities for parent/family contact have been eliminated and class sizes grow to pre-COVID numbers.

Regarding lower enrollment, reduced class sizes, and more materials, it seems as if these changes would decrease exclusionary discipline practices. It follows that fewer children and more materials provide for less opportunities for conflict, which often triggers the behaviors that result in expulsion. Since physical proximity and navigating sharing of materials are often antecedents for challenging behaviors, it makes sense that the reduction of these triggers, combined with fewer children in the room overall, may have resulted in an actual decrease in challenging behaviors, despite perceived increases in behavior.

Necessary Supports

It is notable that participants in this study reported that they did not have the supports necessary when faced with a child with severe challenging behavior. Prior to the pandemic, participants indicated that they did not have resources needed (72.5%; Giordano et al., 2021), yet one year into the pandemic (Giordano, McKeating Giordano et al., 2022), most programs reported (64.6%) that they did have these supports. Despite reporting that they did not have supports, it was found that availability of supports did not predict whether or not an administrator expelled a child. Given the shift in perception of supports, it may be that instead of reporting on actual availability of supports, participants are reporting on how supported they feel. It could be that a sense of solidarity and the feeling that we are all in this together was in effect through the first year of the pandemic. At the beginning of the pandemic, there seemed to be a sense of hope that the appreciation for the ECE workforce might result in systemic improvements in conditions and compensation and elevation of the profession in general. In October of 2020, the U.S. House of Representatives approved the updated HEROES Act, which included an investment of $57 billion to support child care (U.S. House of Representatives, 2020). Media outlets often referred to early care and education professionals as essential workers providing stability when routines were upended, enabling parents to continue going to work or work from home. The ECE workforce may have felt more appreciated at this time. Perhaps as the pandemic wore on in the following year, pandemic fatigue set in, also leaving the feeling of being supported behind. It is interesting to note that respondents one year into the pandemic were more expansive when listing who supported them, naming local public schools, program administrators, consultants/coaches and clergy as sources of assistance (Giordano, McKeating Giordano et al., 2022). In the current study, respondents did not report feeling supported by such a wide array of supports.

Limitations and Future Research

Despite interesting results revealed in this study of childcare administrators as the COVID-19 pandemic wore on, we must address limitations of the study. First, it must be acknowledged that the study is based on a snapshot of time, the survey was disseminated approximately two years into the pandemic and was kept open for a period of two weeks. Administrators that volunteered to complete the survey may have experienced more favorable circumstances than those who did not respond. Administrators that did not respond to the survey may have experienced increased difficulties or more contentious conditions which left them less available to complete the survey. Additionally, these results only represent providers from one state. For these reasons responses may not be representative of the early care and education community at large. Relatedly, this survey was created by the researchers for the purpose of this study and information about the validity and reliability is not yet known. Therefore, further work is needed to determine if these results match actual behaviors of participants, as reported data may not map onto actual practices.

Results from this study show an interesting pattern of behavior and exclusionary discipline returning to pre-pandemic levels, after a brief drop during the pandemic, which brings into question areas for future research. The results of the study beg the question: What factors predict expulsion and suspension, if not supports, staff turnover, waiting lists, or stress in the workforce? Determining factors that predict expulsion are key in designing effective interventions. We also did not measure actual stress of teachers and administrators; instead our survey asked about administrator perception of their own and their staff’s stress levels. Additional work is needed to examine the relationship between teacher stress levels and suspension and expulsion. Also, while we looked at overall enrollment numbers compared with license capacity, we did not look at child:teacher ratios. As this has been shown in prior research (Gilliam & Shahar, 2006) to be related to suspension and expulsion, it may be that a center’s licensed capacity makes less of an impact on expulsion rates than whether a center has a small ratio of students to teachers. Further research is needed in order to determine how ratios, and the decreased ratios reported by participants, relate to suspension and expulsion.

We used anonymous surveys to encourage a higher frequency of submissions and honest responses, but anonymity may not allow us to see trends. For example, are the same centers expelling children at a higher rate than others pre-COVID and now? Are belief systems and interpersonal issues, complex stressors that require more in-depth queries, at play? More in-depth studies of practices in programs with higher rates of suspensions and expulsions may provide key insight into patterns of, and risk factors for, expulsion. Our results also did not allow for meaningful analysis regarding racial disproportionality in expulsion rates. Further research is needed to determine if these trends have been altered during the COVID-19 pandemic. This study also did not examine soft expulsion. Although we did find that suggesting that a program is not a good match does not predict future expulsion, we did not examine if these strategies are used in place of expulsion. If a caregiver chooses to remove a child, without being forced to do so, their removal would not be reflected in our data. Future work is needed examining the practices and prevalence surrounding soft expulsion. Relatedly, research is needed into interventions that address the factor that is consistently reported as the major reason for expulsion, namely, challenging behavior. Finally, similar research must be conducted across states. A consistent way of requiring and collecting data on expulsions and suspensions in our state and nationally would be useful. Practices within states with more success (i.e., fewer expulsion rates) should be analyzed and potentially replicated in other locations.

Conclusion

Two years into the pandemic, administrators reported suspensions and expulsion at or higher than pre-pandemic levels, after a decrease one year into the pandemic. The factors hypothesized to predict expulsion, including availability of support, prior suspension, staff turnover, waiting lists, enrollment numbers, and perception of stress in the workforce, did not. We recognize that there are many factors contributing to the increases in reporting of challenging behavior and the return to pre-pandemic rates of expulsion and that the topic requires additional investigation. This research has revealed the complexity of expulsion practices and we must continue to examine expulsion practices in the childcare community in an effort to effect change.