USVI Public Records

Document text

2022 UNITED STATES VIRGIN ISLANDS DEPARTMENT OF HUMAN SERVICES OFFICE OF CHILD CARE & REGULATORY SERVICES MARKET RATE SURVEY & NARROW COST ANALYSIS REPORT By: Dr. Elizabeth Jaeger on behalf of the University of the Virgin Islands Center for Excellence in Leadership & Learning Submitted on August 31, 2022 Response to OCCRS comments submitted October 21, 2022 Response to OCC comments submitted December 9, 2022 VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 2 INTRODUCTION The University of the Virgin Islands Center for Excellence in Leadership and Learning (UVI CELL) was contracted by the Virgin Islands Department of Human Services (DHS) to conduct a market rate survey of private child care providers throughout the Territory. Recipients of Child Care Development Funds (CCDF) from the U.S. Department of Health and Human Services (HHS), Administration of Children and Families are mandated to conduct market rate surveys (or an alternate methodology) every three years to assist with setting subsidy payment rates. In addition to considering prevailing market rates, CCDF lead agencies are also required to consider the actual costs of providing child care when setting their subsidy payment rates. This includes the costs associated with meeting basic licensing requirements as well as costs related to providing higher quality care. Therefore, data to inform this “narrow cost analysis” were also collected as part of the market rate survey.1 The survey was administered in June 2022. This report details the findings of the survey. In addition, to the data obtained from the survey, information about costs associated with providing Head Start services to families obtained from DHS are included in this report as a means of assessing the costs associated with higher quality care. The USVI Child Care Market Before discussing the survey methodology and results, however, it is important to provide the reader with some context about the child care market in the U.S. Virgin Islands. This information is important to understanding both the absolute rates charged for child care in the VI as well as the variability within rates. First, like its stateside counterparts, the USVI has a mixed delivery child care system that includes both public and private settings. However, unlike many markets in the continental U.S., funding streams are not mixed in the latter. In the VI, all fully publicly funded early learning settings, including Head Start, Early Head Start, and the Virgin Islands Department of Education’s PreK program (i.e., Granny Preschools) are standalone programs. In other words, private settings in the VI do have any slots funded through these programs. The government of the Virgin Islands also provides funding for some afterschool programs in the Territory. Collectively, these free, or very low cost, programs serve over half of children in the VI who are in early care and education settings. Both public and private early childhood programs are regulated by DHS. Second, many programs operate under multiple child care licenses that are defined somewhat differently than those in the continental U.S. In the USVI, there are four types of child care 1 Given time constraints, it was decided to include questions pertaining to the narrow cost analysis in the market rate survey rather than address them through other methodology, such as focus groups. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 3 licenses granted. These include licenses for a family day care home, a group family day care home, a day care center, and a school-age program. The family day care home licenses do not necessarily mean that care is provided to children in a residential home. Rather, these are settings that can serve children from birth to 14 years of age and are licensed to serve up to six children (family day care home) or between 7 - 12 children (group family day care home). Often these licenses pertain to classrooms within a day care center setting. Indeed, child care settings in residential spaces are very uncommon in the USVI and at the time of this survey, none were currently operating in the regulated market. Day care centers are licensed to serve children from two through twelve years of age. School age programs serve children from kindergarten through age 14 during non-school hours (almost entirely after-school hours). Third, a substantial minority of child care programs in the USVI (29.7%) are part of faith-based and secular private K+ schools. Settings in faith-based schools are the most common. Early childhood programs and/or classrooms in these settings are also required to be licensed by DHS. The early childhood programs at these schools operate during the school year only. Some may offer summer camp options, which are also licensed by DHS. When considering a typology for different sorts of private child care programs in the VI, it is most useful to distinguish a) community-based (for profit) child care programs, 2) faith-based school programs, and c) secular school programs. Fourth, the notion of a unitary child care “market” is a bit more difficult to apply to the Virgin Islands. Child care programs are distributed across three islands (St. Croix, St. Thomas, and St. John). The cost of living varies across all three islands which could impact variability in market rates. Although this is true when one considers differences in market rates between urban and rural settings within a state, the increased geographic isolation of islands may exacerbate the variability in rates and impact the true costs of providing services in each. Finally, Hurricanes Irma and Maria of 2017 and the COVID – 19 pandemic have taken their toll on the overall early care and education (ECE) system in the VI. The total number of programs in the ECE system has dropped substantially due to these events. In 2018, the market rate survey was distributed to 111 programs. In 2022, it was distributed to 64 programs. The VI Quality Rating and Improvement System (QRIS) is also not currently operational. Although a QRIS was developed and piloted, its implementation stopped in 2017. The hurricanes, followed by the pandemic, substantially delayed its restart. Thus, unlike many other jurisdictions in the U.S., the VI has not yet been able to implement the use of a tiered reimbursement system based on the QRIS. In sum, although the mixed delivery early care and education system in the VI is similar to systems stateside, there are some notable differences. These should be kept in mind when considering current market rates, the cost associated with providing child care services, and options for revising subsidy payment rates. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 4 METHODOLOGY The survey instrument was initially drafted jointly by DHS -OCCRS and UVI CELL. It was informed by previous market rate surveys and feedback on these surveys from those who provide technical assistance related to CCDF funding to the VI. In addition to questions concerning rates charged by child care programs, information about program demographics, enrollments, other fees collected, subsidy participation, and some of the annual costs associated with providing care was collected. As there are no formal child care provider organizations in the USVI, the initial survey draft was reviewed by a group of six providers who have been acting as an informal advisory group to DHS’ Office of Child Care and Regulatory Services (OCCRS). Edits were made in response to the feedback received. Consistent with previous market rate surveys, DHS and UVI CELL opted to conduct a census survey of licensed, private child care providers in the Territory (N = 64). However, unlike previous surveys, it was administered online using the SurveyMonkey platform as the pandemic required that OCCRS staff and providers interact virtually. This provided OCCRS the first opportunity to use an online platform to collect data. In addition, all providers were given the option to complete the survey and/or ask any questions about the survey by phone. Surveys were initially sent to licensed providers via emails from the Administrator of OCCRS on June 7th, 2022. The email explained the purpose of the survey and also provided phone numbers of those who could assist programs with its completion, including staff from OCCRS and UVI CELL. The original email requested that the survey be returned by June 10th. Programs were emailed again on June 9th asking them to participate in the survey and to again encourage them reach out for any assistance needed to complete. As responses were slower to come in than expected, the survey remained open until the end of June. Unfortunately, response rates by island varied considerably at the close of the survey, with St. Croix having a particularly low response rate. In mid-July, OCCRS staff directly called programs on St. Croix that had not yet participated to encourage their participation. Additional responses were received between July 18th through July 22, 2022. Data Clean-up and Analysis Individual responses in SurveyMonkey were examined to determine if providers had any common difficulties with providing complete responses to any questions. In a few cases, providers did not provide numbers in the unit the response called for (e.g., monthly versus annually). When this occurred, rates were converted to the appropriate unit. For the market rates, all figures for rates and teacher salaries were annualized as almost half the sample only operated during the school year. Finally, full-time care was defined as a minimum of 7 hours per day. One program reported full-time care for a 5.5 hour day and this rate was converted to a 7 hour day. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 5 Response Rates The survey was sent to a total of 64 private child care programs that were currently in operation as of May 1st, 2022 in the USVI2. A total of 46 programs submitted completed the online survey. Two of the 46 programs had a substantial amount of missing data, including their rates, and were excluded from the final sample (N = 44). The response rate was 69.7%. Although the minimum response rate (65%) was exceeded, the margin of error for the overall sample is approximately +/- 8% overall because of the small overall sample size. However, given OCCRS’s intimate knowledge of the population of programs it serves, there is every reason to believe that the sample is good representation of the population of programs. Private child care providers were located on all three islands – St. Croix, St. Thomas and St. John. Table 1 describes the response rate by island. As can be seen in the table below, St. Croix and St. John programs were somewhat under-represented in the sample. Table 1. Response Rate by Island St. Croix St. Thomas St. John 15 (58%) 26 (79%) 3 (60%) Table 2 describes the number of private programs both in the population and the sample by program type. As described above, program type was defined as 1) community-based ECE programs, 2) ECE programs in a K+ faith-based school, or 3) ECE programs in a K+ secular school. As can be seen from the table, the sample as a whole was fairly representative of the different program types in the population, though ECE programs in secular schools were somewhat underrepresented in the sample. This has implications for some implications for the market rates, which will be addressed below. Table 2. Private Programs in Population and Sample by Type Type # (%) Population # (%) Samplea Community-based 45 (70%) 31 (74%) Faith-based School 14 (22%) 10 (24%) Secular School 5 (8%) 2 (2%) a = One program responded anonymously so type could not be determined. Table 3 further breaks down the sample by island and program type. Percentages in the table reflect the percentages of the population of programs represented in the sample by island and program type. As can be seen community-based programs on St. Croix, ECE programs in faith- based schools in STT, and ECE programs in secular schools are somewhat under-represented in the overall sample. 2 There are other private programs licensed to provide child care in the VI but these were either still closed due to the pandemic and/or were standalone afterschool programs, most of which had not yet re-opened. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 6 Table 3. Private Programs in Sample by Island and Type Type St. Croixa St. Thomas St. John Community-based 8 (44%)b 21 (88%) 2 (67%) Faith-based School 5 (83%) 5 (63%) 0 (0%)c Secular School 1 (50%) 0 (0%)d 1 (100%) a One program on STX responded anonymously so type could not be determined. b Percentage of population of programs by type and island c n = 1 d n = 2 Because of these differences, analyses of market rates are presented both for the overall sample, and by island and type of program. Sample characteristics Of the programs participating in the survey, 43% served infants, 52% served toddlers, 93% served preschoolers (ages 2 -5), and 50% served school-aged children in ECE settings. Only two programs served school-aged children exclusively.3 Just over 82% of programs in the sample accepted child care subsidies. While 95% of programs provided preschool services for 7 or more hours per day, what was considered full-time care for a preschooler varied from 5.5 hours to 12 hours per day, with an average of 9.41 (SD = 1.30) hours. This variability reflects the substantial number of programs in the sample that offer preschool services in school-based settings (29.5%). On the whole, programs in school-based settings provided offered significantly less hours of service (M = 8.13, SD = 1.39) than those in community-based settings (M = 9.57 SD = 1.67).4 Relatedly, only 57% or programs provided services year-round, of which, none were school-based settings. Approximately, one-third of programs (31.8%) accepted children on a part-time basis. However, only one-third of these had any children enrolled part-time. The vast majority of children in ECE settings attend full-time. The presentation of the market rate survey results will therefore focus on full-time rates only. MARKET RATES Current market rates are described in this section. Please note that this year’s survey did not include rates for summer camp options as it has in previous years because camps were not open when the survey was first fielded. Because of concerns of the validity of a unitary early care and education “market” in the USVI, in addition to age, rates are presented by type of private ECE setting and by island. 3 Very few stand-alone private after school programs had yet resumed delivering services after the pandemic. 4 t (41) = 2.70, p = .01; d = .897, 95% CI [.36 to 2.50]. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 7 Market Rates by Age Served Market rates were first examined for infants, toddlers, preschoolers, and school-aged children. Table 4 describes the monthly rate for full-time services by age as reported for the entire sample. Table 4. Monthly Rates in the VI in 2022 by Age Served Age served N M(SD) Minimum Maximum Median Infants 24 $502.29(73.07) $350 $600 $500 Toddlers 25 $509.20(79.76) $350 $675 $525 Preschoolers 41 $489.48(164.23) $310 $1295 $475 School-ageda 22 $235.41(99.52) $90 $429 $200 a The school age rates were calculated after excluding a private program that only charged an annual registration fee of $35.00. The large standard deviation in the rates for preschool care reflects a statistical outlier ($1295), a rate charged by a private secular school. When the monthly rate for preschool ECE is re- calculated excluding the outlier, the average market rate drops from $489.48 to $469.35 (SD = $102.99), with fees ranging from $310 to $725 per month. The median drops slightly to $472, and the 75th percentile of the market rate remains at $550. To give the reader a visual representation of the VI market rates, the four figures below depict market rates by age as reported by the entire sample. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 8 Figure 1: Monthly Rates for Full-time Infant Care Figure 2: Monthly Rates for Full-time Toddler Care VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 9 Figure 3: Monthly Rates for Full-Time Preschool Care Figure 4: Monthly Rates for School Age Care VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 10 Market Rates by Island and Age Served Table 5 depicts the monthly rates by island and age served. Please note that STJ has only has five licensed programs in total, and only three are included the sample. Thus, results for St. John should be viewed with caution. Table 5. Monthly Rates by Island and Age Served Age Served St. Croix St. Thomas St. John N M(SD) Median N M (SD) Median N M(SD) Median Infants 9 $500(68) $550 14 $497(76) $500 1 $600a NA Toddlers 9 $519(89) $550 15 $497(74) $500 1 $600 NA Preschoolers 15 $496(114) $550 23 $453(90) $468 2b $463(194) $463 School Age 9 $244 (114) $350 12 $215(81) $258 1 $400 NA aOnly one program serves infants on St. John. bExcludes the outlier. As can be seen from the Table 5, average preschool rates on Thomas (M = $453) appear somewhat less expensive than those on STX (M = $496); however, this difference was not significant.5 Market Rates by Type of Setting and Age Served As described above, there are at least three distinct types of private early care and education settings in the VI: community-based ECE; faith-based school ECE, and secular K+ school ECE settings. Table 6 describes the average monthly rates for preschoolers and school-aged children in these programs as the school-based settings do not serve infants or toddlers (< 24 months of age). As can be seen from Table 6, preschool early care and education programs in faith-based school settings are substantially less expensive on average than preschool programs based in the community or in secular K+ schools.6 Although this appears to be the same for programs serving school-aged children, the result is not statistically significant. Table 6. Monthly Rates by Type of Setting and Age Served Setting Preschoolers School Age N M(SD) Min. Max. Median N M(SD) Min. Max. Median Community 27 $488(93) $ 310 $675 $500 14 $246(81) $160 $400 $238 Faith-based School 11 $407(83) $325 $544 $352 5 $187(142) $90 $429 $120 Secular School 2 $1010(403)a $725 $1295 $1010 2 $300(141) $200 $400 $300 aThe outlier for preschool program rates is included in this average. 5 t(36) = 1.31, p = .197, d = .436, 95% CI [-.225, 1.10]. The difference between the medians was also nonsignificant. 6 Because of the small number of secular school programs, t-tests were conducted comparing rates in community- based and faith-based school settings only. The difference for rates in preschool settings was significant, t(36) = 2.50, p = .017, d = .895, 95% CI [.159, 1.620]. The difference in rates for school-aged programs was not significant, t(17), p = .113, d = .587, 95% CI [ -.461, 1.618]. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 11 Programs in secular schools also appear to be more expensive, on average, than settings in community-based or faith-based school settings. Although caution should be used in interpreting the results because of the small number of ECE settings in secular schools, examination of the costs of other ECE programs in secular schools not included in the sample show that the rates reported by the programs in the sample are representative of the costs of such settings in the VI.7 Additionally, previous market rate surveys that included more of such settings in the sample, clearly demonstrated a very positively skewed distribution in the preschool market rates, with all programs in secular schools grouped at the highest end of the market rates. Such a figure would also vividly depict the segregation of ECE programs in the VI market as there would be a large gap in rates between programs in these settings, and nearly all of the other preschool programs in the VI. A number of factors might contribute to the lower monthly fees for preschool programs in faith-based schools. On average, they are open significantly fewer hours per day (M = 8.00, SD = 1.48), on average, than community-based settings M = 9.57, SD = 1.67).8 Programs in faith- based schools also serve more preschoolers on average (M = 25.27, SD = 13.01) than community- based programs (M = 13.20, SD = 15.54).9 Further, preschool classrooms in faith- based schools are feeders into their K+ programs so rates may be kept low to promote higher enrollments in later grades. Finally, as mission-driven institutions, they may be more willing to absorb losses in revenue to serve families. Comparison of Market Rates and Maximum Subsidy Reimbursement Rates Table 7 shows the monthly market rates by age associated with the 75th percentile, as well as the 50th to 90th percentiles in 10 percent increments. Based on the programs sampled, the 75th percentile of market rates is $550 across infant, toddler, and preschool care. This figure also represents the 80th percentile in market rates for infant and preschool care. Table 7. Monthly Market Rates and Associated Percentiles by Age Served Age Served PERCENTILES 50th 60th 70th 75th 80th 90th Infants $500 $550 $550 $550 $550 $600 Toddlers $525 $550 $550 $550 $590 $600 Preschoolers $475 $500 $550 $550 $550 $600 School Age $200 $245 $264 $300 $340 $400 Unfortunately, maximum subsidy reimbursement rates are well below the 75th percentile. For infants and toddlers (18 months old or younger), the maximum reimbursement rate is $450 per 7 Rates for the other secular schools can be found on the school webpages. These ranged from $995 - $1650 per month. 8 t(39) = 2.73, p = .01, d = .897, 95% CI [.37, 2.50]. 9 t(39) = -2.49, p = .009; d = .878, 95% CI [- 1.88, - 2.2]. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 12 month, which is at the 29th and 28th percentiles of the market rates for infants and toddlers respectively. For preschoolers, the maximum reimbursement rate is $400 per month, which is at the 25th percentile.10 Finally, for school-aged child care, the reimbursement rate is $200 per month, which is at the 54th percentile of market rates. In the 2018 Market Rate Survey Report, it was noted that there was not much difference in the absolute prices of programs at the 75th percentile and those below it. This no longer appears to be the case as the variability in overall rates has increased substantially. As can be seen in Table 8, there has been marked shift upward in maximum prices and median prices, especially for infant and preschool ECE programs. Even though subsidy rates have increased since 2018, the difference between the subsidy reimbursement rates and the median rates charged has also substantially increased for infant and preschool-aged care. Table 8. Market Rates and Subsidy Reimbursement Rates by Year and Age Served Age Served 2018a 2022 Min. Max. Med.b 75th Per.c Subsidy Rated Min. Max. Med. 75th Per. Subsidy Rate Infants $320 $460 $375 $400 $310 $350 $600 $500 $550 $450 Preschoolers $280 $980 $350 $375 $300 $310 $1295 $475 $550 $400 School Age $90 $350 $150 $195 $150 $90 $429 $200 $200 $200 a Rates for toddler care were not included in the 2018 Market Rate Survey Report. b Median. b Market rate at the 75th percentile. c Maximum subsidy reimbursement rate In 2018, because of the more limited variability in market rates, the families who received subsidies had relatively similar access to child care programs to those that did not. However, when considering access to programs that cost > $50 more per month, an increase in cost that could be burdensome for families receiving subsidies, the current subsidy reimbursement rates would not be considered promoting equivalent access to care settings. For infant care, 68% of the programs charge > $50 per month more than the maximum reimbursement rate. For preschool care, 50% of programs charge > $50 per month more than the maximum reimbursement rates. For school-age child care, although more accessible, 40% of programs charge more than $50 per month for care than the maximum subsidy reimbursement rate. Do market rates cover the true cost of care? If market rates are to be used as a basis for setting subsidy reimbursement rates, particularly in ways that assure equity in access to high quality child care, they should be closely aligned with the true costs of providing child care. This is far from true for ECE programs in the VI market. Although an improvement over 2018,11 only 25.5% of programs reported having revenues that met or exceeded their annual expenses in the previous fiscal year. 10 The outlier in the preschool rates is excluded. 11 In 2018, 13.7% of providers reporting charging enough in fees to cover expenses. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 13 There are some notable differences between programs which reported at least breaking even and those that did not. Programs were asked about assistance they received with their expenses and, not surprisingly, programs that broke even were more likely to report receiving financial assistance than those that did break even. As can be seen in Table 9, programs which were able to at least break even reported being significantly more likely or tended to be more likely to have assistance with expenses including the operation of their building, utilities, maintenance, furniture, and supplies than programs which reporting losing money. However, they were not more likely to have revenues from fundraising, a sponsoring organization, grants, or the USDA Food program. Table 9. Percentage of Programs Receiving Assistance with Expenses by Fiscal Status Expenses Breaks even/revenues exceed expenses (N = 11) Expenses exceed revenue (N = 32) Operation of the building1 66.7% 23.3% Rent 36.4% 34.4% Utilities2 72.7% 40.6% Furniture3 63.6% 28.1% Supplies4 72.7% 40.6% Maintenance5 72.7% 35.5% Large Repairs 30.0% 36.7% 1 !2 (1, N = 39) = 5.58, p = .016. 2 !2 (1, N = 43) = 3.38, p =.066. 3 !2(1, N = 43 ) = 4.42, p = .036. 4 !2 (1, N = 43) = 3.76, p = .066. 5 !2 (1, N = 42) = 4.55, p = .033. Whether programs could make ends meet was also related to program characteristics including location and whether or not they currently accepted child care subsidies. As can be seen in Table 10, programs in STX tended to be less likely to break even or make money (7.1%) than programs on STT (30.8%) or STJ (66.7%).12 Programs that broke even or better tended to be less likely to accept subsidies (63.6%) than programs which did not break even (87.5%).12 This may reflect the financial status of the families the programs served. Fiscal status, however, was unrelated to program type or size. Table 10. Fiscal Status by Program Location Fiscal Status St. Croix (%) St. Thomas (%) St. John (%) Breaks Even+ 7.1 30.8 66.6 Loses Money 92.9 69.2 33.3 12 2 X 3 !2 = (1, N = 43) = 5.528, p = .063; Follow-up 2 x 2 test for St. Thomas versus St. Croix also showed a trend, (!2 (1, N = 40) = 2.913, p = .086. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 14 Relationship of market rate setting practices to fiscal status In an ideal world, program rates would be set to cover almost all expenses. However, this is far from true for the majority of programs in the VI. When asked how they set their rates, only 22.7% reported that they set them to cover their expenses. The most common responses were that they set rates to be near the subsidy reimbursement rates (36.3%) or the rates most other programs charge (36.3%). Unfortunately, even programs that set their rates to cover costs were no more likely to break even than programs that set their rates by other means. Results from this survey confirm previous market rate survey results that indicate market rates in the VI are driven by subsidy reimbursement rates, rather than the other way around. A longitudinal view of market rates in the VI confirm this pattern – as reimbursement rates have risen, so too have market rates. Why do programs not raise their rates? The answer is that providers are acutely aware that parents simply cannot afford to pay more. Several providers explained this in their open-ended comments: “Rates can’t go any higher for the finances of parents is not there”. “Unfortunately, I set my rates to accommodate the families in my program due to the economy in which we live. Therefore, I take a major loss in revenue for a passion in which I love.” “It’s a small island I would like to raise but we living in hardship times so I think of others.” “The rate can be charged more but because I serve a lot of low income families it is difficult to do so.” “Due to the limited class sizes sanctioned by Human Services our existing rate does not cover all of our expenses. Even if we increase our rates to cover our expenses there will still be a challenge since we don’t want to outprice ourselves and lose our current enrollment. To avoid this ripple effect, we will need other means to subsidize the program.” According to some providers, the pandemic further exacerbated the financial challenges experienced by programs. “We are school based on assisting low income families, but parental involvement is so important to conduct fund raising activities to ease the cost of the tuition. Presently, the COVID-19 pandemic has provided substantial losses and grants are one of the areas we are relying on to help us continue our mission.” VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 15 Finally, the inability to raise rates to cover the costs of care has negative consequences for providers and programs. According to providers, “We spend more on licensing, taxes, payroll, insurance and the upkeep of the school so much we barely earn anything”. “I service with loans and credit cards to run my school and compare to the mainland were child care is over 13000 per child a year we in the islands are making less than half of that per child a year we are very underpaid.” “As Owner/Director, I am not able to earn my salary because my employees come first.” “The rate that we have helps the families, but not the program.” Participation in the Subsidy Program Most programs in the sample (and population of programs) accept child care subsidies as payment. Approximately, 82% of programs in the sample accepted child care subsidies and the number of children that received subsidies at these programs ranged from 0 – 30, with a median of 4 children per program. Another 11.5% of programs said they do not currently accept them but would be open to doing so in the future, while another 6.8% said they would not accept them. Of the nine reasons given for not accepting subsidies, one-third was because of delays in receiving reimbursement, and another third was due to difficulties in collecting co- payments from families. Other reasons given included that the administrative burden was too high, the reimbursement rates too low, and that programs are only reimbursed for the days when child attended the program. Programs that accept subsidies appear to charge less (M = $471.32, SD = $102.91) than programs which do not accept subsidies (M= $595.42, SD = $357.71). However, this difference was not statistically significant, likely due to the small N and much larger variability in the rates for preschool services observed in the group of programs not currently accepting subsidies. It is possible that with a larger sample that included more of the secular private schools (which currently do not have students whose families receive subsidies), the difference in fees charged for preschoolers in programs that accept subsidies versus those that do could be significant. NARROW COST ANALYSIS The narrow cost analysis focused on estimated costs associated with maintaining licensure, and costs associated with higher quality services as defined by teacher qualifications. The costs associated with maintaining licensing are discussed first. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 16 Annual Licensing Costs As can be seen in Table 11, there was great variability in the amount of expenses programs reported associated with maintaining licensing in the past year. Table 11. Expenses Associated with Maintaining Licensing Expense M(SD) Min. Max. Median Licensing feesa ($35 -135 per site) $170.81(172.83) 0 $900 $100 Fire safetya ($150 per site) $133.93(84.16) 0 $310 $150 Health inspectiona ($75 per site) $110.12(64.77) 0 $300 $100 Criminal background checks $68.50(122.88) 0 $500 $0 Health Cards ($60 per person) $190.12(202.38) 0 $720 $125 Supplies $727.82(957.79) 0 $5000 $400 Indoor equipment $1541.43(2109.69) 0 $7500 $500 Outdoor equipment $3167.22(4997.89) 0 $20000 $710 Repairs licensing $2467.22(2422.45) 0 $8000 $1750 Professional Development $2095.83(3654.56) 0 $15000 $400 Total Expenses $10,248.88(10,260) $830 $42160 $7030 a = These expenses are part of the fees required as part of the annual license renewal process. The fact that some programs reported no expenses in these categories suggests some providers may have interpreted the question differently. The other expense categories vary depending upon the providers' specific needs. Some costs associated with licensing were quite expensive. Although these are not likely to be expenses all programs have every year, these figures speak to the possible expenses programs might face in any given year. Fortunately, OCCRS often makes grants available to providers that can help programs defray some of the larger, less frequent expenses associated with licensing. However, given that the large majority of programs struggle simply to make ends meet, the impact of these expenses on program budgets should not be underestimated. In addition, proportionally, some of these expenses are much higher in the VI than the continental US as obtaining materials, supplies, and furnishings also involve exorbitant shipping costs. Simply put, meeting the demands of licensing is proportionally more expensive in the VI than in the US. The Cost of Quality Because the VI QRIS is not currently operating, we used teacher qualifications as a proxy for program quality. Although an imperfect metric, research suggests that teacher education is correlated with program quality, as defined by metrics such as the ECERS, including research conducted in the VI (Jaeger, Mills, & Braithwaite-Hall, 2017).13 Teacher salaries are also often 13Jaeger, E., Mills, F. & Braithwaite-Hall, S. (2017). Virgin Islands Partners for Success. Final Report submitted for grant # 90YE0152 to the Office of Planning, Research and Evaluation of the Administration for Children and Families, Washington, DC. Caution should be exercised in interpreting these results, however, since the assumption that program quality in the VI can be adequately captured by ERS or CLASS instruments has not yet been adequately tested. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 17 the largest line item in program budgets so increases in teacher qualifications should be reflected in increased market rates. The association of teacher education at the program level and market rates first considered how the percentage of teachers reported as having an AA degree or higher was related to market rates. The percentage of college-degreed teachers in a program ranged from 0% to 80%. Market rates for programs who percentage of degreed teachers fell above and below the 75th percentile (33%) did not differ. Next, the relationship of the total number of degreed teachers to market rates was examined. The total number of degreed teachers ranged from 0 to 8, with 75% of programs with less than two degreed teachers. As can be seen in Table 12, programs with 2 or more degreed teachers, on average, charged more, or tended to charge more, for infant and preschool care than programs with fewer than two degreed teachers.14 Table 12: Average and Median Market Rates by Age Served and Teacher Education Age Served < 2 degreed teachers > 2 degreed teachers N M(SD) Median N M (SD) Median Infants 17 $485.59 (79) $480 7 $542.86 (35) $550 Toddlers 18 $497.50(90) $478 7 $539.29(100) $550 Preschoolers1 29 $455.82(100) $440 11 $505.00 (106) $525 School Age 11 $239.45 (106) $225 10 $231.36(98) $200 1Does not include outlier. Interestingly, although reflected in increased market rates, having < 2 or > 2 degreed teachers was not strongly related to the annual salaries paid to full-time lead teachers.15 This may reflect the limited range in salaries and the small number of degreed teachers at most programs. However, teacher salaries overall were related to market rates for preschool care.16 Perhaps, as the workforce becomes more qualified, a stronger impact will be seen on teacher salaries. Programs were also asked to estimate how much the annual salaries of FT Lead teachers would have to be increased if all teachers were required to have at minimum a Child Development Associate (CDA) credential, an Associate’s degree, or a Bachelor’s degree, as these benchmarks represent different levels of quality that were included within the original VI QRIS. Moreover, although not yet fully enforced, amended licensing regulations in the VI stipulate all teachers must have a CDA as the minimum education requirement. As can been from Table 13, there is considerable variability in the additional expense programs projected to be associated with increasing the standards for teacher qualifications. In part this reflects that some programs already have more qualified staff, but it may also reflect providers’ 14 For infants: t(21.78) = -2.48; p = .022, d = -.823, 95% CI [-1.73, .10]; for toddlers: t(22.95) = -1.68; p = 054; d = - 528, 95% CI [-1.41, .363]; for preschoolers: t(39) = -2.13; p = .02; d = -.731; 95% CI [-1.42, -.034]. 15 t(13.29) = -1.94, p = .127, d = -.535, 95% CI [-1.22, .16]. 16 r(38) = .588, p = .001, 95% CI [.284, .742]. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 18 abilities to estimate what the actual costs could be in the future. Regardless, the median increases suggest that many already struggling programs will have to come up with substantially more revenues to meet increased expenses associated with higher minimum teacher qualifications. Table 13. Projected Salary Increases Associated with Increases in Minimum Teacher Qualifications Degree M (SD) Min. Max. Median CDAa $4734.53(4322.05) $0 $20,720 $4000 AA $8368.24 (6014,38) $0 $25,720 $7000 BA/BS $13,429(9062.27) $1000 $32,000 $10000 a Child Development Associate Credential. Finally, the costs per child per month associated with Head Start and Early Head Start services were examined to assess the potential increased costs associated with higher quality care in the VI.17 Both programs have minimum staff requirements that far exceed that of private ECE settings in the VI. Early Head Start services in the VI cost $1856 per child per month, while Head Start services cost $1436.50 per child per month. These monthly costs per child are nearly 3x the median rates charged for infant and preschool care in private settings in the VI, respectively. Even assuming high quality community-based settings would not typically provide as comprehensive of services as those offered in Head Start, it is clear that the costs to have degreed teachers and resources typical of Head Start classrooms accessible to subsidy- accepting facilities would require a substantial increase in market rates for all programs, except perhaps those located in secular schools. CONCLUSION In considering the current market rates, the maximum subsidy reimbursement rates for infant, toddler and preschool ECE in the VI are well below the 75th percentile. Although absolute differences in programs that were somewhat above and below the 75th percentile in 2018 were not reflected in large differences in absolute costs, this no longer holds true in 2022, especially for infant care. Assuring equity in access to care for subsidy eligible children will require an increase in reimbursement rates. Using the current market rates to set subsidy reimbursement rates, however, is a flawed proposition. This study confirmed findings of previous market rate surveys in the VI that suggest the overall ECE market in the VI is depressed. As has been true historically, prevailing market rates do not cover the true cost of care, with almost 3 in 4 programs not able to break even at the end of a fiscal year. In fact, most programs do not attempt to set their rates to cover costs, but rather attempt to align them with what other programs charge or the maximum subsidy reimbursement rates (which are generally one in the same thing). In the VI, 17 The cost was provided upon inquiry by the Director/Administrator of the VIDHS Head Start Program Masikia Lewis and from Lutheran Social Services of the Virgin Islands (LSSVI) Executive Director Junia John-Straker. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 19 the subsidy reimbursement rates drive the market rates, which, in turn, determines what providers charge. According to a report by the Bipartisan Policy Center (2020), “Basing subsidy reimbursement rates on already depressed market rates instead of the cost of providing child care puts child care programs in a precarious business model and reinforces a cycle of poorly paid staff and low-quality care, even when the program’s leadership and staff are committed to quality improvement.”18 Setting reimbursement rates based on the true cost of care in the VI, however, could also have unintended negative consequences. Increasing subsidy reimbursement rates by too much could inadvertently lead to market rate increases that will make child care inaccessible to the many working families who are not eligible subsidies but are still low-to-middle income. The latter is a real possibility in the VI as most private programs serve both subsidized and unsubsidized children. However, not increasing reimbursement rates to cover the true cost of care means that is unlikely that the subsidy reimbursement rates will ever be able to promote equitable access to high quality child care. Therefore, it is recommended that true cost of care analysis also be conducted to help set future subsidy reimbursement rates. The dilemma facing the VI in setting its subsidy reimbursement rates demonstrates that resolving issues of equitable access to high quality child care at the family level cannot be done without also considering how systemic racial inequality has impacted the larger contexts in which child care systems operate. In the VI, the child care system is very precarious, subsidized through lost wages, free labor, and reliance on external resources that are not guaranteed. An astonishing one-third of programs reported having staff, including directors/owners, who regularly go without pay - and even this is not enough to help programs break even. These issues – which communities of color stateside also face - are further compounded by the VI’s geographical isolation, the challenges of being an island economy, as well as its political status as a territory, which is also deeply rooted in racism and colonialism. The impact of the VI’s territorial status can be seen in the segregated nature of ECE settings serving preschoolers described above. The divide in market rates for preschoolers almost perfectly aligns with a divide in access to ECE settings in the VI based on race. The secular schools predominantly serve the minority of white families who moved to the VI (as well as economically advantaged families of color). The vast majority of families in the VI - who are of Black and/or of Caribbean descent - do not have access to these programs. To be sure, by the number of slots, families who receive child care subsidies have some access to the majority of care available in the VI and raising reimbursement rates to the 75th percentile will promote even greater access. But, if we consider the possibility that resources are related to program quality, then we cannot pretend that families receiving subsidies in the VI have, or will ever have, equal access to high quality care. Subsidy reimbursement policies, no matter how set, will never alone be able to close this very large gap in access to quality. Until equity is also addressed at these larger systems levels, the vast majority of working families in the VI will 18 https://bipartisanpolicy.org/report/the-limitations-of-using-market-rates-for-setting-child-care-subsidy-rates/ VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 20 continue to have access to a depressed child care market that affords them relatively similar access to the majority of very under-resourced ECE programs. VI MARKET RATE SURVEY AND NARROW COST ANALYSIS 21 APPENDIX A: Market Rate and Narrow Cost Analysis Survey Instrument
             !    " !# $          %&'(  $      )  ))         "#  '  &          )   '*     +,    +'& -      ' *)     $              ' ,           ' .$   )      + '"&  /0+/   +$        '#          "      '#'*  )  )1232   ' .          '  & $     456  1789:8:77; - <  '   '           '
                  !   "    #        $$  "    %&"'        ()    ( *$   * (!  *$  $    $ +,-  (  
     . &   - /

   01/1 2  $      
     !  - /

   01/1 3  $      
       4-  
     !   $ $    50 6 7. !  
 (    !  $   
89" $ "    $ 
    !  ( =>1211.

9" $ "    $ 
    !   (=>1211? +
-  !  
      . &   - /

   01/1
  $ "     
    !!  - /

   01/1
#  $ "     
    ! 
%9" $ "    $ !
    !   (=>1211.
,9" $ "    $  
    !   (=>1211. +
2-  !        . &   - /

   01/1
3  $ "         !!  - /

   01/1
4  $ "         !  89" $ "    $      !  ( =>1211. 
9" $ "    $      !  ( =>1211. -       !!   $ $   50 6 7. !  
 (    !   $    +-  !   (  !. &   - /

   01/1 #  $ "        !  - /

   01/1 %  $ "          ,9" $ "    $     (=>1211.  $  :  ; 0 <     $$       !  $ =( 5  6
 7 $$ 5
6# 7   5>6% 7 6 $5'$   
 $7 2  $    ($   $ !

   (" !=(  $   $!  !*!   ?8?      (   ')        ! (   ($ ( $@       !      ("  $(   =(  !$  !$     ! $*!  " ?0?   )   +39"$  ! (     ( " >1 # .=(  !!  *!    ?0? +49"$  ! (     (  ">3 1:# .=(  !!  *!    ?0? +89"$  ! (     (  "13  7 .=(  !!  *!    ?0? +
9"$  ! (     (  "3@  # .=(  !!  *!    ?0? +9"$  ! ( A    ?=(  !!  *!    ?0? +-  ! ! $ " ! 6  . &   +#9"$  ! (     (( 5  6
  7 .=(  !!  *!    ?0? +%9"$  ! (     (  $$ 5
6#  7 .=(  !!  *!    ?0? +,9"$  ! (     ($  $$ 56 $7 .=(  !!  *!    ?0? +29"$  ! (     (!  56%  $7 .=(  !!  *!    ?0? +39"$  ! (   ( /(    .=(  !!  *!    ?0?  &   1  ' !   4-    ! ( $   (". A   < ' $ :  : $ $ -!6 #8  $ " ! (  $$6 $   $   $  $  =( $ @  ! ( *!    ?8? ' (  ! (   )  '   !B! !   A@ ( $ $C     B $ *C  *$C ( !5'?;'D ?7 #
-  ! ! $        (  AC6 $ * C  *C ( 'E;'D F. &      *  !  $  #9" $  !   $  (  $  $ * C  $C ( !(1 
 *8. #  $  (      !$   ! $        (   $ *C  $C ( !.5. 7 C       " -      $ C     ! 5  !$($ $ $7   ( )( ($     $    $     -@   6!   ((  A  5  ! ( 7 ##<(  ("       $  " !      ( .    $ (  ' !         $           $ (  ' !  "   $           $ (  ' !        ! (     $ (      (   (    !$ $ =  ("( ")  *" "    !  (     !  $    $ "   $$!/(   G $  H=$5 A!   7  I  ()  $ "   !   (   ! ((  $ (  C   (  !! /1  1  '   =  !   0   /  (  A   #%A      $$  ! ( *!  $ " ! (  (  (" 5=( !$   ( *!    ?8? $0    $(!J 57   $$  G-0$6   A   #,9" $  !   (  (" 


 
?=( !$     ($( *!    ?8? A!  (;$ C  G      :)!  !!  1  /   5!(  K    !K $"'5$!7  C !5(*! * 7 #2=( !    (  ("   $ $! *!  $  !!   =(       $  $ $ * !    ?8? I! (   '  !$  $  !  (!       @   ' !    ! ( !  (!*"   ' $  $)       @     (" $ )  (   ! ( !   !   +#3<      !$(
 $    !. +#4<      (  $  .5=(  ! 6   $  *  ?0?7 +%8<      !$(
0      !. 5=(  0    *!    ?0?7 +%
<      (     .5=(   0    *  ?0?7 +%<      !$(
 *-  "   $  .5=(  !(6  6 -  *!   ?0?7 +%<      !$(  *-  5  $  "  !   7=(  !6 *! 6  * -  *!    ?0? +%#<     5 7(  A" . +%%<      (   
0$    @.5=(     (6  $    @*!    ?0?7 +%,<      (    $    @.5=(     (6  $    @*!    ?0?7   9 9-!/D:- $-  !  0  $   : $$:$   I    0 B-  : $$:$ /$-  !  0 B-    I $ ; B-  : $$:$ /$-  !  ; B-    I $ 1 B-  : $$:$ /$$  !  1 B-    I $ A   %2  $    (    !     ("  $   $-  !  0 B $   0 B$    $$ $  ; B$    $$ $  %39"1AC:$     !"$  !  $    (0 $    !"   ) $      %40    (6! 6   @"'  !"  "  ! $/  "'(( . &   ,8-    @    !$    !  $"  ! $. &   ,
=( @""'    "    (! "    !$*"  "$  !  !  ( @ .    )  '     !   $ @ $   $"     &"     )  " ! !  (     (!$   $ $     GH= ;$  /  G   5<00K! 7   :)!  !!  1    5 K   !K $"'7 C  C ! (  C !5(*! * 7 9  = ( @ '- /$9$  A  5!  ! (  ! $ 7 ,  $    ( !   (  (" !  =( $     I *       !   !   "  ;'$ '  (   ( !   9  !   9  $ !!     $$5 ' $  *I    *  *  *    * 7 =$! )!   $ $    $$ A $! )!   $ $  /   $$ 5! $ )!  *( * 7 C !/   (   /   $$ C ) $!( $  !   !  5 $  * !( *-07 A   ,<     $       !    "       ACL  $$$ 0  $"   * !!  "  ! $   $/ AC    .0*       !   " ( $    I    ,#0    $( I *"$   !M.        $ !  5'  7  !      $   5  7  '   ,%=*          !   $  $  " !      ( *$/""        !  (!   !.  A@ ($ NC     "$'   ' (!  " !    (!$ B $   (!$  C  (    !     ( $ $  !"6 ( $   $ $ !