The following analysis of access will be necessarily brief. Only the key findings are presented in the narrative while supporting data as well as the raw data included for this report will be available online with a capacity to view and interact with the data. Public users will be able to create maps and tables for their own analyses. Final website development is forthcoming; check back on Advancement Project’s website (http://www.advancementprojectca.org/) for more information.
The analysis of access also carries important limitations based on the data presented and methods employed.
This report contains some of the most pertinent data from local, state, and federal data sources that report on ECE, but Advancement Project staff could not collect every piece of data surrounding access to ECE within the timeframe for this project. Also, in some cases, newer data have been published since the start of this study for some data analyzed in this section. Finally, it is worth noting limitations specific to the population data and ECE services data.
Population data from the American Community Survey are estimates and subject to error. It is important to examine margins of error and to review the non-sampling errors for estimates of interest. For this study, margins of error associated with each estimate are presented online, and source information for these estimates is presented in the report.
ECE Service data here rely on ECE providers reporting their capacity numbers by year of child’s age for the purpose of licensing standards. ECE providers do not operate at full capacity 100% of the time, so this data limits our ability to understand accessibility in comparison to unavailable real-time seat usage data. Additionally, seats in a part-day classroom are weighted the same as seats in a full-day classroom, when they can be used twice in one day.
Capacity rates calculated from the ECE Service data are limited in precision, because they subdivide children based on their year of age and not month age, but available data does not enable that level of specificity. The resulting calculations contained in this report thus reflect necessary assumptions that our research team made in order to best approximate current conditions given the limitations of existing data.
There are limitations on our ability to interpret information from maps and understand the availability of different types of ECE programs using the methods employed in this analysis.
The interpretation of maps by ZIP code tends to favor ZIP codes with larger areas because they tend to stand out more. A note of caution to the reader: ZIP codes that appear larger on maps generally have smaller populations per square mile than smaller ZIP codes. View maps here with an eye towards how the population in Los Angeles County is distributed to get a better understanding of densities. Additionally, conclusions reached in this analysis are based on Los Angeles County, Best Start Communities, and county ZIP codes. They may be sensitive to the geographies selected and may change somewhat in an analysis based on, for example, Southern California, all place-based initiatives, and census tracts. Termed the modifiable areal unit problem, understanding how sensitive these conclusions are to geographic selection would be advisable for future research.
Interpretation of ECE programs and for different ages of children is limited by the fact that parents may choose child care in a ZIP code other than the ZIP code in which they live. Studies of child care choices demonstrate that parent’s choices of child care are “interwoven” with the nature of their employment, such as work demands and schedule, type of work, and location (Burstein and Layzer, 2007). And though ECE providers in larger networks may accept only children from within a given geographic area, this is not a general practice, and ECE providers generally accept families with children who come from ZIP codes beyond the ZIP code of the facility location.
Also, there is little information regarding wait lists for different types of child care and no information on wait times associated with these lists. The 2011 Los Angeles County Health Survey reported that, among primary caretakers of children five years of age and under, 26.9% reported it is very or somewhat difficult to find childcare (of any type) on a regular basis. This figure represents the population of parents/primary caregivers of children five years of age or under, regardless of family income and need for subsidies.
Finally, there is a limited amount of data on license-exempt child care data and for ethnic groups within different types of child care. For these reasons, only ZIP codes where ECE gaps are greatest are identified. Beyond these highest-need ZIP codes, only ZIP code quintiles are directly reported. Exact figures are reported when on firmer methodological ground, such as when analyzing data for all Best Start Communities, and at the county level.
There are several limitations of the data that should serve as cautions for interpreting the results of this section.
First, the most recent study of the workforce done by the Center for the Study of Child Care Employment, UC Berkeley (MacGillvary and Lucia, 2011) involves a secondary analysis of data from 2006, which is over 8 years old. The economic impact study conducted by the Los Angeles Area Chamber of Commerce (n.d.) cites this work and also includes secondary data from the California Education Development Department that represents the year 2008, but this appears to be the most current data available for Los Angeles County. Original data collection studies are extremely valuable and more timely data are required to represent the current ECE workforce. However, such efforts are beyond the scope of this study. Clearly, more timely data based on a representative survey of all members of the ECE workforce in Los Angeles County, including teachers, assistants, directors, and family child care providers would give a more precise picture of the current situation in the county.
Second, there is no centralized repository of key information such as qualifications and backgrounds for all staff, caregivers, teachers, and others working in ECE environments in Los Angeles County. The California Department of Social Services Community Care Licensing Division database provides some of this information, such as a list of all licensed child care centers and family child care home providers. In order to piece together all elements of the ECE workforce, it was necessary to analyze a variety of datasets. These sources included The Los Angeles County Office of Child Care Investing in Early Educators stipend program, the LAUP ASPIRE CARES Plus stipend program, LAUP, and Head Start program database.
For example, data are available describing educational attainment and ethnicity from the California Head Start State Collaboration Office reports, but there is little information on compensation, and no data are available from a database (as opposed to reports) at the level of the individual workers. The Investing in Early Educators Stipend Program (AB212) and ASPIRE CARES provided the most comprehensive data sets but these focused primarily on participants of these programs. Further, analyses of some data, such as that regarding compensation and fields of study, relied on separate datasets that varied by date, data collection methods and populations, making it impossible to combine the data or conduct comparisons. For example, compensation levels were provided for the majority of the participants in these two programs but were not categorized according to full or part time work, making it difficult to interpret the results. Moreover, there was some overlap in the populations included in several data sets, for example; the Head Start data included only staff working in Head Start programs, whereas the ASPIRE CARES Plus data set included some Head Start staff as well as staff in LAUP ECE funded programs and from other non-CDE funded programs. Finally, the AB212 data set included staff and family child care providers working in CDE contracted programs such as State Preschool and Family Child Care Home Education Networks, which could also include some of the same providers in the other datasets. As a result, there was a strong possibility of duplication that further limited the analytic options in using these data.
The result is that each dataset described a subset of the overall workforce population, such as those who participated in each of the programs, but does not describe a representative sample or the entire census of the ECE workforce. Thus, the results displayed in the charts and graphs used only the limited data that were available and represented a very small proportion of the total Los Angeles County ECE workforce. For this reason, analyses of data from all Best Start Communities were not performed and, if they were done, would have resulted in inconsistent results. Caution should be exercised when interpreting the findings from these analyses, because the samples are highly selective and only pertain to a subset of the ECE workforce universe.
Additionally, the working definition of the workforce excludes those caregivers employed in home visiting or the newly created transitional kindergarten program operating as part of the K-12 system. The rationale for excluding these populations from the ECE workforce is that the programs, and the populations of families and children enrolled in them, may be very different from ECE settings. The workforce requirements for staff working in these programs are also likely to be different. To the extent that the staff working in these programs is similar to the ECE workforce, the findings presented in this report cannot be generalized to the workforce staffing these programs. But it is more likely that, by excluding the workforce from home visiting and transitional kindergarten programs, the results are more clearly focused on the key members of the workforce, that is; staff working in a variety of center- and home-based ECE settings.
Finally, license-exempt caregivers were not included in these analyses. As noted earlier in this section, it is difficult to estimate the numbers of license-exempt caregivers because there is no database available. Since the individuals providing license-exempt care usually include professional nannies as well as relatives of the child under care, it is likely that these individuals are very different from the licensed care workforce, making it difficult to interpret the results if they were included in the analyses. New light has been shed on license-exempt providers in the recently released report by Harder+Company Community Research titled License Exempt Early Care and Education Provider Needs Assessment (Harder+Company, 2014).
The experience of seeking and examining the limited data on Los Angeles County ECE workforce has confirmed the need for a more organized, centralized, and standardized way of collecting demographic and other data on this critical workforce. To this end, a California Early Care and Education Workforce Registry (the Registry) has been developed to track information about the education, training, and work experience of early educators (California Early Care and Education Workforce Registry, n.d.). It includes sections for those working in classrooms and licensed homes, as well as for supervisors, managers, trainers, and directors. While the Los Angeles pilot phase of the Registry began in July 2012, at this writing, the data were not readily accessible to assist with addressing the research questions related to describing the current workforce characteristics.
There are several limitations that should serve as cautions for over-interpreting the results of the following analysis of quality. The analysis was constrained by the number of key informant interviews that could be completed within the report scope and timeframe. For instance, the majority of providers were interviewed using an online survey due to time constraints, which has the potential to limit the richness of the data garnered as compared to an in-person interview.
The key informant sample is a voluntary sample of convenience, selected for informant expertise and to provide a wide variety of perspectives. As a result, the key informant interviews cannot be considered representative of other administrators or providers who did not participate. The open-ended survey administering key informant interview questions, as well as the key informant interviews, were analyzed for the most part using qualitative, content analytic strategies. All key informant survey and interview responses were reviewed and common themes were identified, along with less frequent but equally meaningful individual perspectives that might create a fuller picture of the issue at hand as well as providing useful anecdotes or details. This is particularly important for questions 6 and 7, which deal with barriers and challenges to QRIS participation and quality improvement within QRIS. Responses from ECE and QRIS experts were extremely informative, but should not be considered representative of all experiences and perspectives in Los Angeles County or statewide.
The scope and timeline of the project also dictated reliance largely on existing and readily available data sets, except for the information gathered as a part of key informant interviews and key informant interview survey. This is a particular challenge when considering the landscape of quality in Los Angeles County because standardized assessments of program and provider quality have not been performed and reported for most ECE sites in Los Angeles County. Because this report is relying by necessity only on available data from the three local QRIS efforts, in which participation is limited, it would be expected that a full countywide picture of overall quality and variations in quality could not be obtained. The QRIS participants and rated sites comprise a small subset of the total ECE programs in the county and should not be considered representative of the county’s entire ECE system. It is also likely that there are a number of quality assessment and/or improvement efforts currently underway in the county where the data is not available. As a result, their stories and information on the characteristics, quality, or location of other quality improvement efforts are not included in this report.
The focus of this quality section was on licensed providers who were participating in a QRIS at the time of data receipt. License-exempt providers are not currently able to participate in QRIS so there are no data about license-exempt care available with regards to quality level for this report. License-exempt care is an important part of the mixed delivery ECE system in the county, but the level of quality of providers is difficult to measure or track and data are not currently available.
Data on provider characteristics, in the categories of ages served, program schedule, and funding source were not available for all providers from all three of the QRIS efforts in the county. Findings on provider characteristics detail the programs for which data was not available in the findings text and corresponding footnotes.
The lack of one unified QRIS in Los Angeles County limited our ability to more accurately compare ratings between these efforts because there are slight differences in the component tiers, standards, and how these standards are applied to derive a given quality rating. For discussing highly rated sites, all programs with the same numerical rating were considered broadly comparable across QRIS efforts.
Additionally, the data utilized for this analysis were current only as of the time of data receipt, and do not include new developments that occurred since then. For example, at the time the QRIS data were obtained, only a subset of all active QRIS participants were assigned an overall rating. For the purpose of the bulk of the analyses in this section, only providers who were active QRIS participants that had been involved in the QRIS effort long enough to have received an overall rating were included, intentionally limiting the analysis to those who had begun the process of assessing and improving their quality, and not focusing on those who were too new to have received an overall rating. There were additional, active participants who had not yet but were on the way to receiving an overall rating. Additionally, there may be even more prior QRIS participants who had received overall ratings but for whatever reason were no longer active in the process and so were not included in this analysis.