Our Data Methodology & Calculation Models

Every published city dollar figure on Cost of Parenting is a curated planning estimate. Nothing in the current production codebase pulls from USDA Expenditures on Children, BLS CPI, local childcare rate surveys, MIT Living Wage, or any other live government / survey API. The numbers visitors see come from seed data in code.

Hand-modeled reference metros live in lib/mockData.ts as PRIMARY_CITY_SEEDS. Each seed stores stage monthly stacks (infant / toddler / school-age), housing quotes, healthcare factors, and food/supplies anchors that were entered by hand for:

• Austin, TX
• New York City, NY
• Los Angeles, CA
• Chicago, IL
• Denver, CO
• Seattle, WA
• Dallas, TX
• Miami, FL
• Boston, MA
• Atlanta, GA
• Phoenix, AZ
• San Francisco, CA

Additional metros are derived in lib/metroCatalog.ts (METRO_EXPANSIONS). Each expansion points at a nearest reference peer and applies manually set scale factors for childcare, housing, healthcare, and food. Those scaled values are what power the rest of the city catalog.

These curated city figures are not sourced from a live government dataset on published city pages outside Texas. For Texas metros and counties, see the section below — those pages use primary NDCP and HUD data for childcare and housing only.

For Texas only, public pages now use a partial Cost of Parenting Index built from primary datasets:

• Childcare — average of DOL NDCP median full-time weekly center-based prices across infant, toddler, preschool, and school-age bands, annualized × 52. The full age×care grid (including home-based rates) is retained for per-band views.
• Housing — HUD Fair Market Rent for a 2-bedroom unit, annualized × 12.
• Composite — childcare + housing only. This is not a total cost of raising a child.
• Not included yet for Texas — food, healthcare, transportation, clothing, education, and other categories. Those sections are labeled “not yet available” on Texas sourced pages; older mock dollar figures are not shown next to real ones.
• BEA Regional Price Parities — retained for context only. RPP is not multiplied into NDCP or HUD figures (already local dollars). Reserved for scaling unmodeled categories once those fetchers exist.
• Affordability (composite ÷ ACS median household income) on county pages when ACS publishes income. For Kenedy and Loving Counties, Census suppresses income due to small population — pages show an explicit unavailable note rather than blank or imputed values.
• Texas metros (Austin, Dallas, Houston, San Antonio) use population-weighted rollups of their MSA counties for childcare and housing. Non-Texas cities keep curated illustrative baselines unchanged.

County guides live at /counties/tx/[county]. Attribution on each page is built from fetch metadata (source, vintage, fetched date) — not hardcoded.

For a primary seed city, annual category totals are derived from the hand-entered monthly stage stacks and housing / healthcare / food-and-supplies fields (see buildBaselineFromSeed in lib/mockData.ts). Published pages show a data updatedAt stamp from that seed record.

For an expansion city, expandMetroFromPeermultiplies the peer seed's line items by the expansion's childcareScale, housingScale, healthcareScale, and foodScale. No external price feed is consulted at build or request time.

Interactive tools start from a location's school-age / elementary reference baseline, then scale by child age using a fixed matrix in code (globalAgeMatrix):

• Infant (0–2) — 1.4× the elementary reference year
• Toddler (2–4) — 1.25×
• Elementary (5–12) — reference band (1.0)
• Teen (13–17) — 1.15×

The free 18-year forecast (/api/calculate) walks each child forward year by year: age = starting age + year index, look up the band multiplier, and scale the location's annual category costs. Optional private childcareapplies a 1.65× multiplier to the childcare line only. The Family Cost Console's instant totals sum the relevant stage monthly stacks (with that same private-care multiplier when selected) and add the family housing premium once per household.

There is no separate household-size efficiency formula (for example “+27% for one child” or “−24% for 3+”) in the forecast path today — child counts add stage stacks; housing is not multiplied per child.

The site does not currently:

• Index historical survey vintages forward with BLS CPI (or any other inflation series) at runtime
• Import live childcare, rent, wage, or price-parity tables from government APIs
• Claim that published dollars are quotes from a specific daycare, landlord, or insurer

Dollar labels on city pages, calculators, and forecasts are marked as illustrative estimates and link here so the limitation stays visible next to the numbers.

See these planning baselines on live city guides, or open a side-by-side comparison:

Questions about a specific metro or the primary-data roadmap? Contact us or read more about Cost of Parenting.