Social Capital

Social capital is the component of human capital you build out of other people. It is the web of relationships — neighbors, classmates, professional societies, coworkers, mentors — that turns skill and knowledge into actual opportunity. Unlike the genetic and family endowment discussed at the end of this section, social capital is almost entirely developable: you choose where to stand, whom to stand near, and how to be useful. For a young adult deciding where to study or work, and for a parent deciding where to raise a child, this is one of the highest-leverage decisions on the board.

A strong network provides access to jobs, promotions, deals, and advice that never reach a public posting. Social capital is a multiplier on your human capital — the same skill earns far more inside a good network than outside one. Inherited social capital — the connections your family hands you — is simply a head start on that multiplier: a parent’s introduction, a relative’s mentorship, a family friend who vouches for you. If you were handed one, your job is to extend it and pass it on. If you were not, your job is to build it from scratch — and it can be built.

Two properties of the position are worth getting right before you spend a decade building the wrong one. The first is how you measure it. Your social capital is not your contact count; it is the number of people you could actually mobilize, multiplied by what each of them holds in their own right. Two hundred acquaintances who can do nothing for anyone is a smaller position than four who can move money, headcount, or a decision. That product is also why the same hours of sociability spent inside a wealthy institution and outside one build two entirely different assets, and why Chetty’s economic connectedness predicts mobility below while generalized neighborliness does not. It compounds unevenly, too: the more you already hold, the more each additional hour returns, because the well-connected are known to more people than they know and never have to make the acquaintance of their own acquaintances. The curve is steepest at the top, which is an argument for building early rather than an argument for giving up.

The second is that this is the one asset on your balance sheet with a carrying cost you cannot pay in dollars. A portfolio sits there whether or not you look at it. A network depreciates — Bourdieu called its upkeep an “unceasing effort of sociability” (section “Four Balance Sheets and the Rate Between Them”), and the accounting translation is that relationships you stop servicing quietly stop being assets. Nobody sends you a statement when it happens; you find out the day you need the call returned and it is not. Budget the maintenance the way you budget an insurance premium, and understand that you cannot delegate it, because the thing being maintained is that a specific person thinks well of you specifically.

The research is consistent: Putnam’s Bowling Alone: The Collapse and Revival of American Community¹² found that communities rich in social capital show better educational outcomes, lower crime, and faster economic growth. Granovetter’s The Strength of Weak Ties found that the highest-yielding job leads come not from close friends but from acquaintances — the people one step removed, who travel in circles you do not. Breadth of network, not depth, surfaces opportunity.

The most consequential finding is Chetty’s. Across two large studies,^13,14 economic connectedness — the degree to which low- and high-income people are actually friends with each other — proved one of the strongest predictors of upward mobility ever measured. Other forms of social capital — civic participation, tight-knit communities, generalized trust — did not predict mobility. What moves a child up is not how neighborly the neighborhood is; it is whether the rich and the non-rich in it know each other.

That finding is narrower than it first sounds, and the distinction is worth money. Chetty measured friendships — reciprocal ties, where the higher-income person also knows you and can act on your behalf. Watching wealthier people from a distance is a different exposure with a different sign: it delivers the comparison cost documented in section “Lifestyle Inflation: The Hedonic Treadmill”¹⁵ while delivering none of the introductions, and it raises your spending floor toward a peer group that will never take your call. The outcome is simple — the proximity to wealth pays when it is two-way and costs when it is one-way. Build the network; do not settle for being the audience.

Geography decides much of this for you. Low-income individuals make roughly four times as many of their friends inside their own ZIP code as high-income individuals do; the affluent form more of their friendships at college, because they are far more likely to attend. Place therefore constrains the poor far more tightly than the rich — the lower your starting point, the more your address is your destiny. You can inspect this directly: the Social Capital Atlas publishes exposure and friending-bias data for every high school, college, and ZIP code in the country. Treat it as a screening tool, not a mere curiosity.

The old boys’ network is real — so build your own The informal system by which the already-connected route jobs, deals, and capital to each other is not a paranoid fantasy. It is proximity, trust, and reciprocity operating among people who share institutions. You cannot abolish it, and resenting it changes nothing. You can join one — and build your children into one. The mechanics are unglamorous and entirely learnable:

Choose the institution instead of the individuals

You do not befriend your way into a wealthy peer group one person at a time — that is transactional, and transactional networking fails. You choose the room, and the room assigns your peers. Selective universities, certain employers, professional associations, nonprofit boards, alumni clubs, and a short list of neighborhoods come pre-stocked with the network. Pay the price of admission and the friendships form on their own.

Use the credential as a passkey

A degree from a selective school is worth more as a membership card than as a transcript. Chetty found that universities are nearly the only venue where the wealthy reliably form friendships across class lines;¹⁴ high schools, workplaces, religious groups, and neighborhoods mostly do not. The credential buys a seat in the one room that mixes.

Optimize for weak ties

Stay loosely connected to many circles instead of deeply embedded in one. Keep in touch with former colleagues, classmates, and acquaintances — the contact you have not spoken to in two years is statistically the one who changes your career.

Give first, and be useful

A network is a reciprocity system, and reciprocity starts with a deposit. Make introductions, share what you know, solve small problems for people before you need anything in return. A reputation for being useful and reliable compounds exactly like money.

Place your children inside it

For parents, this is the highest-leverage move in the chapter. Choosing a neighborhood and a school does not merely buy a child an education — it assigns the peer group, the parents standing behind that peer group, and the friending environment the child inherits for decades. Use the Social Capital Atlas the way you would use a school-ranking site.

Keep the proportions in perspective: a network is a multiplier, not an engine. Nearly every wealthy person first earned a high income; the network is what compounds that income into more. A superb network wrapped around a weak earner produces a well-connected person of modest means. The play is to build the earning power and stand in the right room — never to treat one as a substitute for the other.

Marriage, Genetics, and the Endowment You Pass On

Networks are not the only thing institutions sort. Marriage is too. Federal Reserve economists¹⁶ found that people increasingly marry partners with similar education and skills ( assortative mating), and that this sorting accounts for roughly half the rise in household income inequality between 1980 and 2020 — education alone explaining about 35%. Two high earners marry, pool two incomes and two networks, and pull away from the household that pooled two modest ones. The university is not only where you meet your network; it is where you meet the person you compound it with.

There is a harder layer beneath this, and a book written for readers who intend to build something lasting should not flinch from it. Human capital is not assembled purely from schooling and effort. A substantial share of the traits that drive earnings — cognitive ability, conscientiousness, health, even height — is heritable, and nearly every dimension of individual difference that bears on earning, from risk tolerance to occupational choice, carries some heritable component.^17,18 This is not fringe science; twin and adoption studies have measured it for decades. But the useful question for a parent is not how much of a trait tracks DNA. It is how much of your advantage reaches your children, and through which channel — and on that question the evidence is both stronger and more encouraging than the heritability debate suggests.

Sorting concentrates the endowment, and it sorts on traits, not credentials. The largest study of the question — 1.5 million people across 212,070 extended families in Norwegian registry data, using the partners of twins and siblings to separate genetic from social sorting — found that partner similarity in education is explained better by assortment on underlying traits than by assortment on the diploma itself, with social homogamy playing a particularly large role. The implied genotypic correlation between partners was $r=0.34$, higher than sorting on observed education alone would produce.¹⁹ Earlier work reached the same conclusion from the other direction, finding spouses measurably correlated on the heritable correlates of education and height.²⁰ People match on the qualities that generate the credential, not the credential itself, and that concentrates the whole bundle — genetic and environmental — more tightly than the marriage statistics above imply.

Most of what transmits is not DNA. Three independent designs converge here, which is why the finding is worth acting on. In the same Norwegian data, roughly 38% of the parent–offspring correlation in education ran through environmental transmission; models that assumed simple assortment attributed less to environment but fit the data worse.¹⁹ A separate design compared children’s outcomes against the alleres their parents carried but did not pass on — alleles that can only act through the environment a parent builds. Those non-transmitted alleles predicted the child’s attainment at about 30% of the strength of the transmitted ones, and including this genetic nurture raised the variance explained by the parental endowment from 17% to roughly 30%. The authors are explicit about the mechanism: the score is tagging intelligence, conscientiousness, and future planning, and parents with it shape their children “through many behaviors,” not through their own schooling alone.²¹ In United States data, only one-sixth to one-third of the intergenerational correlation in educational attainment is attributable to shared genetics at all.²²

The practical reading: the advantage you transmit is large, and most of it travels through the environment your own traits construct — how problems are discussed at the dinner table, whose behavior is available to copy, which strategies for thinking are demonstrated instead of taught. That channel is not a consolation prize for families short on genetic luck. It is the larger channel and the only one you can deliberately operate.

Heritability is a population statistic — it describes how much of the variation in a trait tracks genes across a population. This is not an individual verdict, not a ceiling stamped on any one child. The malleable inputs — environment, schooling, social capital, the executive functions discussed next — remain large, and they are the entire subject of this book.²³ Also be careful which number you carry. The polygenic scores that predict educational attainment reasonably well predict income far worse — on the order of one to two percent of the variance in an independent sample.^24,25 Cognitive ability and temperament transmit; wealth transmits mostly through capital, networks, and behavior, which is why the rest of this book exists. And none of this licenses crude instrumentalism: you are choosing a life partner, not breeding stock.

What survives all the caveats is the decision itself. Choosing a partner is, among other things, choosing your children’s endowment — the heritable component, the household environment that endowment builds, and the network they inherit, all at once. Evaluate it with the seriousness you would give any other irreversible, multi-decade commitment. The economics of that commitment — prenuptial agreements, titling, the marriage penalty, and the cost of unwinding it — are taken up in chapter “The Economics of Marriage”.