The Capital Problem, the Markets, and Who's Building
Americans are angry, startups need to act, the markets are clear, the problem is capital, who's with me?
Americans are angry. That’s not news - since the Vietnam War and Cold War eras, Americans’ satisfaction ratings with the government, their jobs, society, just about everything have ticked downwards.
But what is new is the level of violence and despair wracking our country. And much of this despair stems from a core economic problem: Americans feel they’re paying more for essential services and getting less in return.
When I say essential services, what I mean are the fundamental things Americans must have to live: healthcare, housing, dependent care, and education (or job-dependent skills training). These services serve as the bulwark against financial ruin. And they rely upon each market functioning to support the others. The maxim ‘without health, you have nothing’ is the basis for this observation. But healthcare drives the American economy. With 17 million Americans depending on it for their income and $5.3 trillion exchanged for healthcare goods and services in 2024 alone (18% of GDP), it's not a mission-driven enterprise. It’s big business.
You can’t move out of your body. Frustrating as housing is, it is expensive, and Americans are right to be angry about it – the market has release valves. Startups are working on financing. More homes are being built. States and cities are changing zoning laws and taking on the NIMBYs who have blocked supply for decades. Austin is the textbook case for what happens when you actually build - inventory grows, prices ease. Someone may want a cheaper house in San Francisco, but San Francisco has regulatory and geographic constraints that other markets don't. Same with Boston, NYC. They are desirable and there are good jobs. But people can move. The market is working, however imperfectly, and the feedback loops exist.
Education, on the other hand, is littered with broken markets. We have an oversupply problem in America - too many people with an (expensive) education that they thought would lead them to a stable upper-middle-class life, and not enough upper-middle-class jobs for them to fill. For decades, we sold high school graduates on the four-year degree. A graduate degree is even better. But there was never a nationwide recognition that the cost of the four-year degree, the job it led to, and the income it produced often didn't add up. The crisis of people getting degrees that outstrip their annual take-home pay has resulted. Technology has also eroded, and will continue to erode, many of the white-collar jobs that these very degrees churn out graduates for (see Anthropic). We’ve seen other trends emerge, too - more young people and mid-career people retooling themselves to take on skilled work. A growth in apprenticeships and skills training programs to fill the very jobs this push for a 4-year degree decimated - plumbers, welders, electricians. Even PE has taken note - the focus on these ‘non-software’ ‘real’ jobs has become a greenfield opportunity to apply operational best practices to meet social need, while producing significant profit. Nursing is among the highest in-demand jobs, and is viewed as one of the most surefire pathways to a comfortable economic future. While Academia, particularly four-year institutions, is rightfully course-correcting amid changes in economic demand and a declining American birth rate, this market, too, is working in many respects.
But as for the dependent care essential services? Well, those markets are in such disarray that Healthcare Delivery seems rational. What makes them challenging is the dynamic - the services themselves require human care, which is great because these jobs will remain in demand because AI can’t touch them (see, again, Anthropic), bad because they’re low wage, low skill, largely filled by immigrants, and generally ‘hard jobs’ due to the physical and emotional toil.
While childcare isn’t universal in the same way that healthcare and housing are, and isn’t inevitable in the way elder care is, it is a precondition of workforce participation. And the 40% of the population with children, who participate in the labor force, need it to work. Despite this need and the increase in workforce participation, wages have not kept up with the meteoric rise in childcare costs. What was once $85/week in 1997 now averages $400-$500/week. Inflation alone would put that cost closer to $170.
What's behind this rise? Childcare workers earn between $12 to $15/hour, which means labor isn't the cost driver. Instead, insurance, real estate, and overhead are. And while in other markets substitute goods have emerged (i.e., you can opt for outpatient vs. inpatient care), the fundamental need for a human to watch a young child, care for their needs, and aid in their development is immutable. There isn't a strong technological fix or real 'opportunity' for AI to change the market. But financial mechanisms are shifting. Section 45F expanded under OBBBA, increasing the credit rate from 25% to 40%, and the cap from $150K to $500K, with small businesses getting up to 50% on $600K. Toyota and others are building on-site childcare. TOOTRiS is matching families to slots and bringing in employer dollars. When I worked on the Hill, there was bipartisan appetite for this. Imperfect and uneven, but moving.
Eldercare, or aging care, on the other hand, is stalled if not spiraling downward quickly. Sixty-three million Americans are family caregivers right now, and by 2034, for the first time in history, there will be more Americans over 65 than under 18. Seven in ten will need long-term care, against a median older-adult income of $36,000 and care costs that range from $24,700 to $288,288 a year. Only 14% of adults over 65 carry private long-term care coverage, and roughly 22% of older Americans have no spouse, children, or close relatives to rely on for care. Similar to childcare, the workers who provide elder care - nursing home attendants and home health aides - are low-wage, low-skilled, and predominantly immigrants. Also, as with childcare, unless you are very poor and qualify for Medicaid, these costs are entirely borne out of pocket.
A startup, CareYaya, explains this workforce problem exceptionally well: Most care companies capture more than half the value, while the caregiver captures less than half, even though the caregiver does 99% of the work. And to keep wages accessible to the people who pay them, they took grants and made the service a non-profit.
What capital has done with this market is extract from it: PE acquisition of nursing homes is associated with a 10% increase in 90-day mortality (NBER 2021), more emergency room transfers (JAMA Health Forum 2021), and lower frontline staffing (GAO 2022). Middle-class families finance the gap by spending down their assets to qualify for Medicaid, which is now the primary payer for long-term care in the country. After four years in a nursing home, 62% of private-pay residents have transitioned to Medicaid. Estate planning, asset protection, closing out a life — these are no longer specialty services. They are line items in the standard cost of aging in America. But 1 in 5 adults over 50 have no retirement savings at all, and 80% of older households cannot withstand a major financial shock, such as the need for long-term care. Americans as a society don't like to think about mortality and old age, but there is a 100% chance everyone will encounter it, and assuming Medicare and Social Security will cover isn’t just willful ignorance - it's life school malpractice.
I didn’t pick these markets just because they’re what make people angry (that's thanks to Annie Lowrey). I picked them to spend the past two years exploring because they share a few patterns.
First off, this is a problem of the commons. The government catches the very poor. Capital catches the billionaires — the Elon Musk tier, not the dentist with a vacation home. Everyone in between is balanced on financial quicksand. Quicksand may not seem like a big problem until you realize that financial health works like bodily health. One serious diagnosis, one ER visit, one $400 surprise bill that 37% of Americans can't cover, and the dominoes start falling. The credit score takes a hit. The loan terms get worse. The line of credit you needed to renovate the spare room for mom evaporates. The $400 you didn't have for the surprise gets pulled from this week's daycare bill, which means you can't go to work, which means you forgo income. People have a finite budget. There is no backstop. Everything held in precarious balance crashes at once.
Second, the companies trying to help people in quicksand are mostly built on extraction because the capital they take demands it. Capital is loaned on terms that anticipate a return, and different types of capital expect different returns on different timelines. Grants and philanthropy expect no financial return; the capital is tied to your execution of the mission. Debt expects principal back plus 6-12% interest on a fixed schedule. Angel investors expect 20-30% annual returns and 10-30x on individual winners over 5-10 years. Private equity expects 3-5x in 4-7 years, usually through operational restructuring. Venture capital expects 10-15x at Series A on a 10-year fund clock, with the math built around the assumption that most companies fail and a few have to return the entire fund. Most founders can't bootstrap and don't have a wealthy friend, so some mix of capital is necessary to bring a product to life. The decision of what kind of company you want in 10 years — one you can sell, one you're still running, or one set up to last beyond you — and how committed you are to the mission underneath it determines which capital you can take and which you can't. Most founders never run that calculation. But if you have a product, you have early indicators of traction and a track record to look to; there are options. They may be hard to find, and they may not yield a splashy Crunchbase alert, but they can help you make it to another day with your vision intact.
Third, profit isn't the problem. Extraction is. There's a difference, and Promise and Cost Plus prove it.
Promise is venture-backed. Phaedra Ellis-Lamkins took venture funding and still pivoted away from her first product, bail processing, about 18 months in, after participating in Y Combinator in 2018. She built the next product so that governments and utilities pay (the entities with money) and the people in distress don't. Zero interest. No customer collections. The company is profitable. The capital is venture. The architecture is non-extractive because she structured it that way.
Cost Plus is the other side of the same coin. Mark Cuban funded Alex Oshmyansky in 2018. The company didn't launch until January 2022 — four years between the check and any revenue. A VC fund would have forced a pivot, a sale, or a third fundraise long before launch. Cuban's capital bought Cost Plus the gift of time to set up the company carefully — partnerships, pricing model, regulatory posture — without the pressure to ship before the model was right. Manufacturing came later, after launch traction was established. That sequence isn't possible under standard VC clock pressure.
Most companies in these markets don't choose either path because the path of least resistance is generic VC, and VC math demands extraction in markets where users can't pay enough to fund the return. Promise and Cost Plus aren't anomalies. They're proof of what's possible when the founder deliberately chooses the capital and the buyer, rather than taking what's available and discovering that the optimization function has shifted beneath them.
The founders who can fix this are the ones willing to do the unglamorous work. Pull out the original business plan and see where you actually track. Read your term sheets again. Level with Product, Engineering, and Sales about what the capital you took is asking of you. Have the honest conversation about whether you're still on the path you started on, or whether the terms you signed pulled you somewhere you didn't mean to go.
That's the work. Not the next round. Not the next hire. The reckoning with what you signed up for and whether the mission survives it.
