Every city has a network you can't point to on a map. It runs under the hours of the day, not the streets. It is built from calendars and keys, not concrete. Its substations are living rooms and church basements, corner daycares and adult day-health centers, the spare bedroom of a licensed home provider, the Tuesdays when a cousin can cover, the midnight shift of a home-care aide whose bus never comes on time. Economists call this "care." Cities act as if it were weather. In truth it is a grid: a system that converts attention into time so other work can occur.
This essay treats childcare and eldercare as infrastructure—with uptime targets, redundancy, dispatch rules, and capacity planning—rather than as a private puzzle dumped into the laps of households. We will avoid sentimental arguments. We will draw the network, define its physics, calculate its failure modes, and propose instruments a city can actually build: isochrones for reachable affordable care, "care latencies" analogous to grid restoration times, a Care Grid Index that a mayor can understand at a glance. No sermons. Blueprints.
I. The Missing Utility
Imagine a care network as a power grid.
- Nodes: childcare centers, licensed family homes, in-home sitters, pre-K classrooms, after-school programs, adult day-health centers, respite programs, home-health agencies, memory-care wings, and—often forgotten—informal kinship nodes that carry a large fraction of the load.
- Edges: transit lines and sidewalks, but also the time edges—opening and closing hours, the frequency and predictability of schedules, the notice window in which capacity can be dispatched.
- Voltage: the hours of operation; higher voltage means evening/weekend coverage, not just 8–5.
- Amperage: the workforce—trained, paid, retained. Without amperage, voltage is a sign on a door.
- Load: children under care, elders needing supervision or assistance, and the schedules of the adults around them (parents, adult children, employers).
- Control Room: enrollment systems, eligibility rules, subsidies, vouchers, case managers; today they resemble bulletin boards. A grid requires dispatch.
A city can go dark in two ways: the lights fail, or the grid works for the wrong hours. Our care grid does both—no capacity where demand peaks, and the wrong voltage for the people who keep the city running nights and weekends.
II. Uptime, Not Vibes
Every real utility lives by a dashboard. The care grid needs one too. Borrow the discipline from power systems and transit.
- Care Uptime (CU): share of scheduled care hours delivered as promised. A center open 10 hours/day, 5 days/week, 50 weeks/year offers 2,500 hours. If closures, staff shortages, or illness cut 150 hours, CU = 94%. Track by neighborhood and provider type.
- Household Care Latency (HCL): median hours between a failure (provider closure, canceled sitter, discharged elder needing supervision) and a safe substitute being in place. In neighborhoods with long HCL, entrepreneurship and second jobs die quietly.
- Unserved Shift Hours (USH): evening and weekend gaps between stated need and actual capacity, measured in person-hours. USH tells you where transit-adjacent evening care is not an amenity but a production function.
- Affordability Pressure (AP): copay as a share of net income at the 30th, 50th, and 70th percentiles, after subsidies. A grid that is "available" at 40% of income is a museum exhibit.
- Reliability Events (RE): closures per 100 child-months or elder-months, categorized (staffing, illness, inspection, landlord issues, transit failure). Power grids log SAIDI/SAIFI; care grids must log why a Tuesday broke.
Targets? Treat care like power: CU ≥ 99% for subsidized slots; HCL ≤ 24 hours; USH trending downward each quarter. If that sounds ambitious, remember: when electricity flirts with 94% uptime, it's a scandal.
III. Isochrones, But For Care
Urbanists use isochrones to map where you can travel in N minutes. Build the same for care: reachable, affordable care within 20 or 30 minutes by walking and transit, at hours that match your shift, priced below an income-indexed threshold.
Care is a coverage cube: space × time × price.
- Space: 30-minute isochrone around home and around work.
- Time: exact opening/closing windows; swing hours (6–9 a.m.; 5–9 p.m.); weekends.
- Price: after-subsidy copay relative to income bands.
A city's care map may appear generous at noon on Wednesday and desolate at 6 p.m. on Saturday. The cube makes the emptiness visible. With it, you can place care hubs where the cube thins: near logistics parks with night shifts; at bus transfer nodes; inside public buildings that die at 4 p.m.
IV. The Subsidy Cliff and the Marriage Penalty of Time
Households don't just face high prices; they face cliffs—income thresholds where a small raise deletes a large subsidy. The cliff is not merely unfair; it is grid-unstable. It induces demand to disappear abruptly, then reappear as informal care at unsafe latencies.
Fix the physics:
- Replace cliffs with co-pay glides: smooth marginal effective tax rates (METRs) so that every kroner or dollar earned reduces benefits by a gentle, published slope. The control room can then predict load.
- Index subsidies to hours delivered, not just enrollment. Evening/weekend hours cost more to staff; pay more per hour and require higher uptime. A subsidy not yoked to hours is a dimmer switch that does nothing.
- Build eligibility grace windows for volatile incomes (gig work, shift work): 90-day averaging, not weekly write-offs. Care collapses on Thursday when forms are weekly.
V. Workforce as Capacity, Not Charity
Grids fail from neglect. Our care grid neglects the people who constitute its amperage.
- Create a career ladder with wage floors tied to credential steps (assistant → lead → master educator; aide → advanced home-care professional). Publish the ladder; tie public funds to it. If you subsidize seats while letting wages erode, you are subsidizing outages.
- Fund paid training time and ratio relief (a float pool) so providers can upskill without violating ratios. The grid's reserve margin is people who can step in when the classroom goes from nine toddlers to twelve.
- Offer portable benefits—health insurance, retirement micro-credits, paid sick time—across employers. The grid needs mobility; otherwise providers are prisons.
- Build a city float corps: a centrally managed reserve of vetted caregivers who can be dispatched when a center goes dark or an aide calls in sick. The dispatch window is measured in hours. Think of it as N-1 contingency: the grid survives the loss of one node.
VI. Eldercare Is the Other Half of the Circuit
Childcare monopolizes headlines; eldercare drains calendars. The physics are similar, with extra hazards.
- Adult day-health centers provide supervision, medication management, and socialization. Their uptime reduces emergency-room recidivism and caregiver burnout. Treat them as peaking plants: fund extended hours around discharges and holidays.
- Respite care is a maintenance outage for humans. Subsidize regular respite like oil changes; neglect leads to breakdowns measured in ambulance rides.
- Home-health scheduling is dispatch. Unreliable arrival times have the same effect as rolling blackouts: families cannot plan work. Require time windows that stick and publish on-time arrival rates.
- Discharge planning is interconnection. Hospitals should be grid operators: no discharge to an address without a verified care plan and transportation. Memory-loss households need anchored handoffs, not pamphlets.
VII. Employers as Large Industrial Loads
When factories connect to a power grid, they pay for the right to draw steady load and are penalized for violent spikes. Today, employers behave as if care were a private afterthought. Make them grid citizens.
- Offer preferred permitting and tax credits to employers that contract care seats with evening/weekend coverage and co-site care near shift changes. Seats are not perks; they are capacity purchases.
- Publish an employer care factor: the share of employees covered by contracted seats or vouchers, plus schedule predictability scores (two-week advance posting; cap on last-minute changes). Tie public procurement preferences to this factor.
- Penalize schedule volatility that breaks the grid (last-minute shift flips without premium pay). Volatility must carry a price like peak demand charges.
- Build multi-employer care hubs at logistics clusters, hospital districts, and university zones. Let employers subscribe like bandwidth.
VIII. Zoning, or How Cities Forbid the Obvious
Many cities outlaw the grid accidentally.
- Legalize small-footprint childcare in residential zones with light-touch safety rules and neighbor management plans. Parking minimums designed for stadiums will kill family childcare.
- Retrofit dead retail into care hubs. A failing strip mall with a bus stop is a perfect node; the landlord's fear of "children" is an administrative myth. Offer code-compliance blueprints and tenant-improvement loans.
- Create a single-permit concierge for care providers—one counter, one checklist, one timeline. The enemy of uptime is a six-office scavenger hunt.
- Lock long leases with escalation caps for subsidized providers in public buildings. Landlord churn is a rolling outage.
IX. Standards and Data: The Care-NAP
Transit has a GTFS (a standard feed for routes and times). Create Care-NAP (Node and Availability Protocol): a simple, open data standard for publishing hours, capacity, prices, eligibility, and real-time status. Require it for public money.
- Parents and adult children see inventory in real time. Case managers route like dispatchers, not like people trapped in PDFs.
- Platforms can compete on experience while sharing the plumbing. Forbid self-preferencing that hides inventory behind paywalls.
- Privacy is sacred: Care-NAP shares availability, not personally identifiable information. Audit the API like payments, not run like social media.
- Publish reliability reports: on-time opening rates, cancellation rates, complaint resolution time. If a provider fears publication, they fear their own numbers.
X. Financing: Capex Is a Building, Opex Is a Civilization
Cities love ribbon cuttings; they neglect maintenance. Care punishes that habit.
- Establish a Care Infrastructure Bond for capex: build or retrofit centers, adult-day sites, respite suites, transit-adjacent hubs. Capex follows the coverage cube, not politics.
- Pair with a Care Operations Fund that pays for the unfashionable: extended hours, float pools, supervisor training, workforce wage floors. Tie payouts to uptime and HCL improvements, not attendance alone.
- Use wage pass-throughs: a percentage of public dollars must show up in paychecks, audited quarterly. When wages rise, HCL falls; we measure both.
- Create a Countercyclical Care Reserve: during recession, demand spikes (job search, stress) and supply falls (closures). A reserve prevents sudden blackouts.
XI. A Dashboard the Mayor Can Read
Build a Care Grid Index (CGI) from sub-scores:
- Coverage (30%) — Care-NAP isochrone score at 20/30 minutes by income band and hour band (day/evening/weekend).
- Reliability (25%) — CU, RE, HCL medians by tract.
- Affordability (20%) — AP at income bands; slope of co-pay glide.
- Workforce (15%) — wage floors vs. median local wages; training hours; turnover; fill rates for float corps.
- Governance (10%) — permitting speed; data compliance; employer care factor average.
Publish CGI quarterly. Red means outages the way transit maps do. If elected officials don't like the color, they can change the grid, not the legend.
XII. Three Model Cities (Working Theories)
The Freight Crescent. Logistics parks on the edge, a river of night shifts, a desert of evening care. Solution: a triad of care hubs at transfer nodes with 6 a.m.–midnight coverage, employer seat contracts, and a bus line re-timed to shift changes. CGI rises as USH collapses.
The University-Health District. Daytime care is decent; eldercare is fragile; residents bounce between ER and home. Solution: co-located adult day-health with respite bays; hospital discharge interconnection; on-time home-health metrics published; vouchers for scheduled respite. HCL for eldercare shrinks; ER recidivism falls quietly.
The Tourist Coast. Weekend spikes; weekday lulls. Solution: pricing that pays for weekends; micro-centers near service-worker housing; school gyms converted to Saturday care; transit runs later. Reliability events drop; AP stabilizes even as seasonal wages swing.
These are working theories, not scripture. The test is the dashboard. If CGI rises and households' hours-freed per month go up, the model is earning its keep.
XIII. Scheduling Is Governance
You can't fix the grid without fixing time. Many outages are born in scheduling software: 48-hour notice of shift changes, random weekend rotations, algorithms that spread misery evenly.
- Mandate predictable scheduling for large employers: two-week advance posting; penalty pay for last-minute changes; automatic care-seat holds that expand when shifts expand.
- Require platforms that dispatch home aides to respect human time: stable routes, minimum guaranteed hours, paid travel time. Route optimization that ignores the aide's life is theft by spreadsheet.
- Treat school calendars as grid backbones: align breaks across districts; publish years ahead; coordinate with care providers for surge coverage. The school calendar is a load profile; plan for it.
XIV. Quality Without Pretension
A grid that hits 99% uptime by warehousing children or elders is a moral failure. Quality is not an afterthought; it is the reason the grid exists.
- Define quality floors: age-appropriate ratios; evidence-based curriculum; fall-prevention and cognitive engagement for elders. Publish them and audit lightly but often.
- Fund coaching, not just compliance: a master educator or nurse who rotates among sites to improve practice. The gain is visible in retention and complaint rates, not just in checklists.
- Align inspection with uptime: closures for trivial infractions are blackouts disguised as virtue. Triage violations; cure fast; publish timelines.
XV. The Informal Circuit
Grandmothers, neighbors, cousins—informal care is the shadow network that saves the formal one. Do not romanticize; support.
- Offer micro-stipends and training for informal caregivers who pass basic checks. A thousand small stipends stabilize ten thousand Tuesdays.
- Create a kinship registry with optional inclusion in Care-NAP for last-minute coverage referrals through vetted channels.
- Protect tenancy for caregivers housing elders or children; evictions are rolling outages with paramedics attached.
XVI. Measurement So We Don't Lie to Ourselves
A serious grid keeps score.
- Hours-Freed per Household: net increase in usable adult work/education hours after care uptake, by decile. If this rises, labor force participation and small-firm formation follow.
- Startup Elasticity: change in new business registrations in tracts where USH falls sharply. Care unlocks risk-taking.
- Gender Wage Gap Compression: track in neighborhoods where HCL dropped below target. If the gap shrinks, the grid is doing economics, not merely ethics.
- Hospital Recidivism (for eldercare): 30-day return rates after discharge where adult day-health hours expanded.
- Queue Health: mean/median waitlist lengths and volatility. Long queues are voltage lies.
XVII. Objections, Answered in the Language of Systems
"This is too expensive." Outages are expensive. Missed shifts, lost taxes, ER trips, churn. Price the outages honestly; the grid will look cheap.
"Families can solve family problems." They already do. The grid's point is not to replace kinship but to remove the cliffs and latencies that turn love into panic.
"Quality will suffer." Define, fund, and audit floors; build coaching; pay wages that sustain skill. Uptime and dignity are not enemies.
"Markets can handle this." Markets handle commodities with predictable demand and patient buyers. Care is spiky, regulated, and geographically sticky. It behaves like a utility. Treat it like one.
"Data sharing risks privacy." Share availability, not identity. Audit systems like payment networks. Privacy is a constraint; grids live with constraints.
XVIII. The Minimal Program (One Year, Hard Stops)
- Publish the Cube: Care-NAP standard; 30-minute isochrones by income band and hour band; real-time availability feeds for subsidized providers.
- Buy Reliability: wage pass-throughs, float corps, evening/weekend rate multipliers. Tie funds to CU ≥ 99% and HCL ≤ 24 hours.
- Smooth the Cliff: co-pay glide that caps marginal clawback; 90-day averaging; backstop fund for arrears that prevents blackouts.
- Permit the Obvious: by-right small-footprint childcare; one-stop permits; retrofit dead retail; long-leases in public buildings.
- Make Employers Pay in Time: schedule predictability law; employer care factor; procurement preference for high scorers; multi-employer hubs near shifts.
- Interconnect Hospitals: no discharge without verified care plans; adult day-health surge funding; on-time home-health metrics.
- Publish the CGI: quarterly scores with subindices; red maps where outages cluster; a mayoral briefing that confronts time, not talking points.
If you can only do three, do Cube + Reliability + Cliff. The rest can follow.
XIX. The Geometry of Freedom
It is fashionable to speak of freedom as a landscape without fences. In cities, freedom is a schedule without panic. The care grid is not a utopia; it is a geometry of hours sturdy enough that a parent can say yes to a promotion, a home aide can plan a Tuesday, a daughter can sit through a meeting without watching the phone vibrate on the table like a small trapped animal. When the grid is strong, entrepreneurship is thinkable; second chances are not jokes; poverty loses one of its cruelest tricks—stealing time twice: once from today's work, again from tomorrow's possibility.
XX. Epilogue: The Room With Two Clocks
In an ordinary apartment, there are two clocks. One tells the city's time—shift changes, school bells, buses. The other tells the household's time—the pillbox, the nap, the pickup, the key exchange, the cousin's Tuesday. When the clocks disagree, people fall out of the economy through the crack between them. We have tried to fix this by telling households to be heroic. Heroes are unreliable infrastructure.
Build the grid instead. Give the clocks a common reference. Publish the uptime. Staff the float. Place the hubs where the cube is thin. Smooth the cliffs so raises don't break Tuesdays. Discharge to plans, not pamphlets. Pay the workers like capacity, not charity. Let employers subscribe to seats like bandwidth. Count hours-freed, not speeches given.
Do this, and a city becomes larger not by annexing land but by releasing time: a quiet enlargement measured in dinners made, assignments completed, fewer midnight returns to the ward, more startups that survive the first year, and the dull, shining fact that fewer Tuesdays explode. The grid will not sing about itself. It will hum, the way a good transformer hums, a sound you stop hearing because life is happening above it—steady, ordinary, and finally, free enough to be interesting.