WomensBureau Supporting Statement Part B

B. COLLECTIONS OF INFORMATION EMPLOYING STATISTICAL METHODS

1. Describe (including a numerical estimate) the potential respondent universe and any sampling or other respondent selection methods to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corresponding sample are to be provided in tabular form for the universe as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.

Market Rate Surveys. The federal government provides states with funding to offer financial assistance to low-income families to pay for child care. This funding is provided through the Child Care and Development Fund (CCDF) childcare subsidy program.¹ Subsidy payment rates are determined by states, but payment rates are required to be informed by childcare market prices. States are encouraged, but not required, to set payment rates at the 75^th percentile of the market price.² To obtain market prices, states are required to conduct a Market Rate Survey (MRS) no more than two years prior to the submission of states’ Child Care and Development Fund (CCDF) Plan. The CCDF Plan serves as the mandatory state application to receive Child Care and Development Block Grant (CCDBG) funds.³

MRS data used in the CCDF Plan must be current and complete, with data collection practices statistically valid and reliable.⁴ MRS must be conducted at least every three years. States must include variation in prices based on provider type, child age, and geographic location. Market prices must reflect local prices so parents who receive a subsidy have access to the range of providers in their local community. The goal of MRS is to demonstrate that subsidy payment rates provide equal access to quality childcare to families receiving subsidies when compared with unsubsidized families, while factoring in the age of the enrolled children and their geographic location.

MRS are based on a universe or sample of providers in each state. Most states rely on existing administrative databases of providers to define the universe: regulatory agencies (licensing), child care resource and referral (R&R) agencies, and subsidy agencies.⁵ States sample childcare centers and home-based providers that charge a price for childcare. MRS do not capture care provided by family members or providers operating illegally. The price of childcare is the fee that providers charge to parents for childcare services. Prices are differentiated by the age of the child (e.g., infants, toddlers, preschoolers, school-age children) and are expressed in terms of the hours of care per a particular period of time (e.g., hourly, weekly, monthly). HHS reports that price data are provided accurately by providers and price information is tightly linked to what parents pay in the market.⁶

Respondent Universe. MRS reports and data will be requested from all 50 states, the District of Columbia, and Puerto Rico. Data requests will be sent to each state or territory’s childcare administrator responsible for filing CCDF Plans (i.e., Lead Agency).

Activity	Number in population of respondents	Number of respondents in the sample
Provide existing copies of market rate survey reports and data files	50 states, District of Columbia, and Puerto Rico (52 government Lead Agency respondents)	50 states, District of Columbia, and Puerto Rico (52 government Lead Agency respondents)

2. Describe the procedures for the collection of information.

Market Rate Surveys are conducted in 3-year cycles. The Women’s Bureau will request the most recent cycle of MRS reports and tabulations covering years 2019 through 2021 from all states, the District of Columbia, and Puerto Rico. Respondents will be provided with at least 30 days to respond. As data will be requested from all states, no sampling selection is necessary. Some states may have conducted multiple MRS during this period, including some additional data collection and tabulation to provide supplemental data on disruptions to the childcare market and pricing during the COVID-19 pandemic. Additionally, some states were granted a waiver to continue data collection and reporting into 2022 due to pandemic-related disruptions. The Women’s Bureau will request published and unpublished MRS reports, tabulations, and data that would allow for the creation of county-level estimates of childcare prices.⁷ We will also request any available supplemental data reflecting pandemic disruptions to more accurately develop pricing estimates and imputation models over a time period that spans pre-pandemic to pandemic years (2019-2022). Some states or jurisdictions (e.g., New Mexico, District of Columbia) conduct their studies using cost-based, alternative methods in lieu of a MRS. These reports and files will be collected and may be included in the NDCP if the data can be made comparable to data derived from a MRS. An expert panel will be convened to evaluate options for inclusion of data from cost-based, alternative methods.

We seek to publish county-level estimates because county-level data will more precisely reflect the prices parents pay in the market than state-level childcare price estimates available from other sources. We are not requesting data below county level because smaller geographic areas would pose more significant disclosure risks, as smaller areas may have one or few providers. American Community Survey (ACS) demographic and economic indicators are also more extensively and consistently available for counties on a yearly basis, and the NDCP is linked with publicly available ACS data.

All states are required to make available their most recent MRS report online. However, public reports may not have the granularity of data the Women’s Bureau seeks. For the current collection spanning years 2008 through 2018, a total of 35 states responded to the data collection request and provided MRS reports during the data collection window.⁸ Of these reports, 25.5% included county-level childcare price data. Our contractor, ICF, was able to locate online 46 additional MRS reports covering multiple states, increasing coverage to 48 states and the District of Columbia. Many states only made publicly available childcare price data at the state or regional level. Our contractor developed an imputation model with input from many experts and stakeholders to account for missing county-level data in cases where reports or tabulations did not provide these estimates.⁹

Imputation. For states that refuse to provide county-level data or tabulations, or did not collect or do not have available county-level data, an imputation method using socio-economic variables and state-level estimates will be used to impute values for each county. Our contractor, ICF, will implement a county-level imputation model using data from states with county-level estimates, regional or cluster-based estimates provided by the state, and economic data from the American Community Survey. These imputations will be based on our current imputation methodology developed by ICF.¹⁰ These methodologies were informed by an expert panel the Women’s Bureau convened on imputation of county-level childcare prices in 2020. The panel had representation from experts from the Department of Health and Human Services, Child Care Aware of America, Lead Agency representatives from four states, ICF, and the Department of Labor. The imputation model was also informed by ICF’s work on MRS in California where a set of more than 500 variables are used to derive childcare price estimates, including housing costs, income, and educational attainment.¹¹

The current imputation model is a county-specific adjustment to the state average that varies based on the related socio-economic data from the ACS, information derived from state or regional childcare prices, information derived from different age groups for states that are missing data for only some age groups, and information derived from either the 50^th or 75^th percentile price if both estimates were not provided. Data in between study cycles are imputed based on known changes between study years and the consumer price index.

Imputing prices based on a different age group. To impute these missing values for a given county and year, the existing age groups (those provided by usable MRS reports) for that county and year are used. For instance, if a county is missing the childcare price for infants in center-based care but make available the childcare price for toddlers in center-based care, the latter was used in the imputation methodology. Specifically, the adjustment to impute the missing values are based on the ratio of the average prices for these age groups for other counties within the same state.

Imputing county-level childcare prices from statewide data. In the current collection, fifteen states provided MRS reports that only included state-level data. In these instances, ICF developed predictive models of county-level prices using county-level variables that are highly correlated with childcare prices. These include county-level data on educational attainment and median gross rent. ICF estimated regression models to predict the ratio of county childcare prices relative to the state level. The model had an r-squared value of 0.84.

Imputing 50^th and 75^th percentiles. The NDCP provides the median and 75^th percentile childcare price. In 18 states, one of these values was missing. To impute these values, ICF used a statistical model to estimate the missing percentile from the available percentile.

Imputing estimates for years between study cycles. Each state may conduct their MRS in a different year within the data collection cycle. To provide data for each year, our contractor will impute missing years based on known changes for that state between study years and the consumer price index.

Disclosure avoidance. The NDCP will not disclose any personally identifying Information (PII) or business identifying information (BII). The Women’s Bureau does not expect to receive any files that contain PII or BII from the states. In instances where PII or BII are identified, the Women’s Bureau or its contractor will remove and destroy the PII and BII data before the data are entered into the database. For any instances in which original source data are requested from states, our contractor will request that PII information be removed from the file before transferring the files. If there are any instances in which a state indicates that it will not be feasible to remove PII from the source file, our contractor will provide the state will access to a secure online portal into which the files can be transferred and the Women’s Bureau or its contractor will remove and destroy the BII or PII data before the data are entered into the database. The database will also be evaluated to identify counties with fewer than 10 providers for which these data were not already made public by the public state MRS report. Data for counties with fewer than 10 providers will be set to missing and imputed using the same imputation methodology described above.

Standardization. MRS have some state variability in how metrics are presented in MRS reports. For example, states may use different childcare price modes (e.g., daily, monthly) and different age groups. To ensure that the NDCP is consistent across states and over time, the Women’s Bureau, through its contractor, will standardize variables across states. Childcare prices will be converted to weekly and annual estimates for each age group and care setting (e.g., center-based or home-based providers). The NDCP provides detailed age categories in 6-month increments for children ages 0 through 5 and one school-age price. The NDCP also aggregates age groups for infants, toddlers, and preschoolers to provide the 50^th percentile and 75^th percentile price for these groups. ICF used the standard age groups provided by Childcare.gov and examined the most common age groupings in the MRS reports to aggregate ages as follows: infant (0-23 months), toddlers (24-35 months), and preschoolers (36 up to, but excluding, school age¹²).

The NDCP also provides data standardized by geography. Counties are presented using standard Federal Information Processing Standards (FIPS) codes to be compatible with the American Community Survey and other major data sources.

3. METHODS TO MAXIMIZE RESPONSE

Response rate. The Women’s Bureau is seeking the most recent market rate survey (MRS) and supplemental COVID-19 assessments from each state. States are required to publish the most recent MRS online. This will minimize burden for states to respond and will increase state response rate. In the current collection, spanning years 2008 through 2018, 35 states responded to our data collection request and provided MRS reports and our contractor was able to obtain the remaining published reports and tabulations and data files for a total of 48 states and the District of Columbia. The response rate from the states was 70% but the contractor was able to secure MRS reports from over 90% of the states for at least one of the years in scope, yielding good coverage. Many MRS reports are publicly available and our contractor did not request these reports from the states which improved coverage significantly. Given the recency of data collection and the relative simplicity of providing the final studies and tabulations that were already produced, we expect to have a response rate of over 85% for 2019 through 2022. The NDCP is also no longer a new data collection and states have learned about the utility of the NDCP. Indeed, several states that did not provide data in the original request contacted the Women’s Bureau to seek to provide data at a later date once the NDCP was published and they noted missing data for their state.

The Women’s Bureau has taken a number of steps to increase response to the information collection. To secure a high response rate, the Women’s Bureau has partnered with the Administration of Children and Families, Department of Health and Human Services (HHS) whom have a direct relationship with states. State Lead Agencies work with HHS extensively to develop their Child Care and Development Fund Plans using data from their market rate surveys. The Women’s Bureau, through its contractor, will also implement tiered follow up outreach efforts. For states that are unable to provide the required county-level source data files, our contractor will follow up to request copies of the final published studies and would expect that at least 85% of states would be able to provide the final published studies. The Women’s Bureau, through its contractor, will evaluate which states have already published a sufficiently detailed MRS report online and will not request a response from those states, minimizing burden and increasing our response rate.

Partnerships. The Women’s Bureau has secured support from HHS to conduct the collection. The Women’s Bureau has consulted HHS on the best timing to contact the states to provide the data so it does not significantly conflict with other reporting requirements. HHS maintains extensive contact with the states and Lead Agencies. Prior to the initiation of data collection, HHS will advise states that the Women’s Bureau will be contacting them to request MRS data, including contact through Regional Program Managers and regional technical assistance specialists who work for the State Capacity Building Centers. This will increase state confidence in responding to the Women’s Bureau collection request. In addition to partnering with HHS, the Women’s Bureau will work with a contractor that has experience working with states through contracted MRS work or technical assistance.

Prenotification and notification letters. The Women’s Bureau, through its contractor, will provide state Lead Agencies with a prenotification letter and e-mail describing the information collection request, along with tiered follow-up notifications. The contractor will follow up with a solicitation letter and e-mail requesting that states provide unpublished MRS reports and tabulations, sufficient to derive county-level estimates of childcare prices. States will be provided with 30 days to respond to the initial request. The contractor will follow up with a reminder e-mail one week before the submission deadline to states who have not responded. The contractor will also send out an e-mail requesting additional information to states that provide an incomplete response. The contractor will follow-up with an additional round of e-mails to states that have not provided a response or provided an incomplete response one week after the submission deadline.

Burden reduction. The Women’s Bureau, through its contractor, will conduct a review

of all 52 state and territory childcare agency websites to locate and archive all MRS reports that are already publicly available. This will allow the contractor to avoid the duplication of asking states to provide studies that are already available online and reduce their reporting burden. The contractor will only request data that have not already been made publicly available. The contractor will provide states with enough prenotification and time to deliver the requested files. The contractor will collect responses electronically via a convenient and secure online portal.

4. TESTING OF PROCEDURES

Procedures will not be formally tested for this information collection. Communications with states and territories requesting MRS reports and data will follow statistical best practices to obtain a high response rate (Attachment D).

5. CONTACTS FOR STATISTICAL ASPECTS OF DATA COLLECTION

The contact for statistical aspects of the design and data collection and analysis is Liana Christin Landivar, Senior Researcher, Women’s Bureau ([email protected]; 202-693-6713).

1 Office of Planning, Research and Evaluation, Department of Health and Human Services. 2023. “Enrollment Size and Subsidy Density of Child Care Centers Receiving Child Care Subsidies in 2019.” Issued April 10.

2 Office of Child Care, Department of Health and Human Services. 2023. “CCDF Payment Rates – Understanding the 75^th Percentile.” Retrieved June 27, 2023. https://childcareta.acf.hhs.gov/sites/default/files/new-occ/resource/files/508ed-75th_percentile_exercise_1.pdf.

3 Office of Child Care, Department of Health and Human Services. 2022. “Child Care and Development Fund Program Plans.” Issued June 7. Available online: https://www.acf.hhs.gov/occ/plans.

4 Davis, Elizabeth, Lynn A. Karoly, Bobbie Weber, Pia Caronongan, Kathryn Tout, Patti Banghart, Sara Shaw, and Anne Partika. 2017. “Market Rate Surveys and Alternative Methods of Data Collection and Analysis to Inform Subsidy Payment Rates.” OPRE Report #2017-115. Office of Planning, Research, and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Services. Available online: https://www.acf.hhs.gov/opre/resource/market-rate-surveys-and-alternative-methods-of-data-collection-and-analysis-to-inform-subsidy-payment-rates.

5 National Center on Subsidy Innovation and Accountability. 2018. “Designing and Conducting Your Market Rate Survey.” Issued April; Grobe, Deana, Roberta B. Weber, Elizabeth E. Davis, J. Lee Kreader, and Clara C. Pratt. 2008. “Study of Market Prices: Validating Child Care Market Rate Surveys.” Technical Report. Oregon State University Family Policy Program, Oregon Child Care Research Partnership. Issued September.

6 Davis, Elizabeth, Lynn A. Karoly, Bobbie Weber, Pia Caronongan, Kathryn Tout, Patti Banghart, Sara Shaw, and Anne Partika. 2017. “Market Rate Surveys and Alternative Methods of Data Collection and Analysis to Inform Subsidy Payment Rates.” OPRE Report #2017-115. Office of Planning, Research, and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Services. Available online: https://www.acf.hhs.gov/opre/resource/market-rate-surveys-and-alternative-methods-of-data-collection-and-analysis-to-inform-subsidy-payment-rates.

7 Published reports and tabulations will only be requested if these items are no longer available online.

8 The current collection, spanning years 2008 through 2018, were collected between December 2019 and February 2020. This was the first time these data were collected by the Women’s Bureau so multiple MRS cycles were collected at once to obtain historical data.

9 ICF. 2020. “U.S. Department of Labor, Women’s Bureau National Database of Childcare Prices: Final Report.” Available online: https://www.dol.gov/sites/dolgov/files/WB/media/NationalDatabaseofChildcarePricesTechnicalGuideFinal.pdf.

10 ICF. 2020. “U.S. Department of Labor, Women’s Bureau National Database of Childcare Prices: Final Report.” Available online: https://www.dol.gov/sites/dolgov/files/WB/media/NationalDatabaseofChildcarePricesTechnicalGuideFinal.pdf.

11 Department of Education, State of California and ICF. 2017. “2016 Regional Market Rate Survey of California Child Care Providers.” Issued April 17. Available online: https://cappa.memberclicks.net/assets/CDE/2016-17/2016%20ca%20market%20rate%20survey%20final%20report%202.pdf.

12 School age is allowed to vary by state based on each state’s compulsory age for attending school full time.