Determinants of Observed Price and Quality Distributions
To structure our empirical work, we adopt a model of supply and demand for child care developed by Chipty and Witte (1994). For the purposes of this paper, we estimate the reduced form implied by this model. The estimation of a carefully specified reduced form increases our confidence in results we obtain for the effect of R&Rs on the distribution of prices.
The PCS provides detailed information on prices and staff/child ratios 18 for each group at a center. Prices vary with age of the child and with the period for which care is purchased (e.g., daily, weekly, monthly, annually). Centers generally supply care and quote prices on a weekly basis. To reflect this, we employ weekly prices in our analysis. Centers generally have at least three distinct groups, and groups generally contain only children of specific ages. For each age group, we construct a center level measure of price and stafi/child ratio using information for the groups that care for children falling in that age group. Units for prices are chosen to reflect the most frequent hours per day and days per week combination for each age category.
PINFANT is the weekly price for the care of children that are less than a year old for five days a week and ten hours a day. PTODDLER is the weekly price for care for children between the ages of one and three, for five days a week, eleven hours a day. PPRESCH is the weekly price of care for children between ages three and six for five days a week and three hours a day. PSCHOOL is the weekly price of care for school-age children for five days a week three hours a day. If a center offers multiple packages for care of children of a particular age, the price for the age group is computed as the average price for that age group at the center. All prices are deflated by a regional cost of living index at the county level (American Chamber of Commerce, 1990).