You think your wealth plan is a two-generation story: you and your children. The data says you are wrong by one generation in both directions.
You think your wealth plan is a two-generation story: you and your children. The data says you are wrong by one generation in both directions.
Your bank balance today carries a measurable imprint of a man or woman who may have died before you opened the account. And the decisions you make this year will leave the same imprint on grandchildren who do not exist yet. This is not mysticism. It is one of the most striking quantitative findings in the study of family wealth, and almost nobody plans around it.
The evidence comes from Fabian Pfeffer and Alexandra Killewald's "Generations of Advantage", published in Social Forces in 2017. Their raw material was the Panel Study of Income Dynamics, which has followed American families continuously since 1968, long enough to observe three generations of the same families as adults: 4,608 parent-child pairs, with grandparent wealth measured decades earlier.
Two numbers matter here.
First: the correlation between grandparent wealth and grandchild wealth is 0.23. Grandparents you may barely have known predict your net worth today at a strength most people would find uncomfortable.
Second, and this is the part that should reorganize your planning: only about half of that grandparent effect flows through the parents. The rest reaches the grandchild directly, bypassing the middle generation entirely. If wealth moved the way most people assume, grandparent influence would be fully absorbed into parent wealth: grandpa helps dad, dad helps you, story over. The data refuses that story. Grandparents reach around their own children and touch the third generation by separate routes.
For calibration, the same study measured the parent-child correlation at 0.33 for adults aged 25 to 34, rising to 0.44 by ages 55 to 64. Advantage compounds across a lifetime, and the grandparent signal at 0.23 is not far below the early parent-child signal. The three-generation link is not an echo. It is a channel.
The researchers point to a set of concrete mechanisms, none of them mysterious once named.
Direct help with education. Grandparents pay school fees, fund university, cover the gap year the parents could not afford. This is money that lands on the grandchild's human capital without ever appearing in the parents' net worth, which is exactly why it shows up as a direct effect.
Housing deposits. The grandparent contribution to a first plot or a first home deposit is one of the most common direct transfers in wealthy families worldwide. It puts the grandchild on the property ladder, the single largest transmission channel in the same study at 28.4 percent, years earlier than their income alone would allow.
Norms and expectations. Children who grow up around a grandparent who owned land, ran a shop, or held title deeds absorb a template: people like us own things. Expectations about education, saving, and ownership transmit in conversation and example, at zero monetary cost, and then quietly steer decades of decisions. The reverse template also transmits, which is the uncomfortable half of the finding.
Safety nets. A wealthy grandparent is an insurance policy the grandchild never pays premiums on. Knowing a fallback exists changes behavior: the grandchild takes the better-but-riskier job, starts the business, stays in school through a hard year. Families without the net make defensive choices at every branch point, and defensive choices compound too.
Evidence-first cuts both ways, so here is the honest critique.
The main academic objection to multigenerational findings like this one is measurement error. Wealth is hard to measure at one point in time; parents' full economic position is captured imperfectly in any dataset. Skeptics argue that some of the "direct" grandparent effect is really unmeasured parent advantage: if we could see the parents' complete position perfectly, the grandparent's independent contribution might shrink. This debate, sometimes framed around whether social status has "excess persistence" beyond parent-child links, is live and unresolved.
Second, this is American data. Extended-family structures, co-residence patterns, and remittance norms differ sharply across societies, and in much of Africa the grandparent is not a distant figure sending an occasional cheque but a central node of the household economy. That likely makes the direct channel stronger in African families, not weaker, but the honest statement is that the 0.23 was not measured here.
Third, correlation includes the ugly inheritances too: debts, disputes, land conflicts, and the norms of families that never held assets. The channel is neutral. It transmits whatever the grandparents actually had.
None of these critiques erase the core result. Even the skeptics' corrected estimates leave a real three-generation link, and the mechanisms are ones you can watch operating in your own extended family this year. The finding survives its critics; it just asks to be held with appropriate confidence.
Run the finding forward and it lands on you.
If grandparent wealth predicts grandchild wealth at 0.23, with half the effect bypassing the middle generation, then the title deed you register this year, the education pot you fund, the business you formalize, these are not gifts to your children. They are gifts to your children and your grandchildren, in measurable proportion, whether or not you ever meet the grandchildren.
This changes the arithmetic of sacrifice. A parent deciding whether the titling fees are worth it is usually weighing cost against one generation of benefit. The data says the benefit line is drawn too short. Whatever you secure compounds through your children and reaches past them. And the same is true of what you fail to secure: the untitled plot, the unschooled child, the normalized debt. Those transmit at the same strength, down the same channel, with the same bypass.
Legacy is a three-generation game whether you plan it or not. The only choice you actually have is whether to play it deliberately.
The age pattern in the study sharpens the point further. If the parent-child correlation climbs from 0.33 in early adulthood to 0.44 by the late working years, then transmission is not an event but a slope, and the third generation inherits the slope, not the snapshot. A grandchild does not receive your net worth on one particular day. They receive the compounding trajectory your decisions set in motion: the title that appreciated for forty years, the education that raised two salaries which paid three sets of fees, or the dispute that consumed a decade of family attention and legal fees. Trajectories are set early and cheaply. They are corrected late and expensively. The month you formalize the plot or open the pot, you are choosing the slope for people two generations away.
Here is where most wealth content stops and LegacyPot does not. If grandparents are a proven, partly independent transmission channel, then a family plan that gives its living elders no role is leaving its strongest documented channel idle.
Most families do the opposite of deliberate. The grandparents are loved, visited, sometimes supported financially, and completely unstructured. Their influence on the grandchildren happens by accident, in whatever hours proximity allows. The data says this accidental channel carries a 0.23 signal. Imagine what it carries when someone actually aims it.
Three concrete roles, in ascending order of formality.
The elder as Habit Champion. Children adopt habits from grandparents with less friction than from parents, because the grandparent relationship carries authority without the daily power struggle. Give one grandparent explicit ownership of one family financial habit: the elder who asks every grandchild, every visit, what they saved this month; who hands over the small ceremonial deposit for each birthday; who tells the story of the first plot every time the family gathers. This costs nothing and formalizes the norms-and-expectations mechanism the researchers identified. The title is not decoration. Naming the role is what turns an accidental influence into an assigned one.
The grandparent-funded education pot. Where the elder has means, route their giving through structure instead of ad hoc gifts. A grandparent contribution that lands in a named education pot per grandchild does double duty: it is the direct education transfer the data shows bypassing the parents, and it arrives at the highest-yield point of the grandchild's timeline instead of as a bequest decades too late. Grandparents consistently say they prefer seeing their money work while alive. This structure gives them exactly that, with a statement balance they can watch grow.
The recorded story archive. The least financial role and possibly the most durable. Sit the grandparent down with a phone and record: how they earned their first money, what the land cost and what it took to keep it, the famine year, the debt that nearly sank the family, the rule they lived by. The Kurtz corpus on family legacy calls the grandparent relationship an "irreplaceable spiritual and psychological inheritance," and irreplaceable is the operative word: this asset has a hard expiry date, and no lawyer can probate it back after the funeral. A family that holds twenty recordings of its elders holds the norms channel in permanent, transferable form. Grandchildren not yet born will watch them.
Notice that none of the three roles requires the grandparent to be wealthy. The Habit Champion and the story archive run entirely on presence and memory, which means the three-generation channel is available to every family, not only the propertied ones. The mechanism the data measured in rich families is rentable by anyone with a living elder and a plan.
One more thing the correlation tables do not show: every mechanism above except the bequest requires the grandparent to be alive. The fee help, the deposit, the habit modeling, the recordings, all of it runs on a clock nobody can see. Families postpone structuring the elder's role the way they postpone writing wills, and for the same reason: the deadline is invisible until it has passed.
You likely have, right now, a living elder whose influence on your children is running at accidental strength. The data says that influence is one of the most powerful documented forces on your grandchildren's future balance sheet. It is currently unmanaged.
This month, give one grandparent one named role in the family plan. Pick the role to fit the elder: Habit Champion if they have presence, education-pot funder if they have means, story archive if they have memories, which is all of them. Say the role out loud to them and to the children, because an unnamed role is not a role.
If the grandparents are already gone, the assignment inverts: you are now the grandparent-in-training, and the 0.23 you build starts this year. Record your own first entry in the archive.
This piece did its job if one family stops treating its elders as beloved spectators, hands one of them a named position on the field this month, and starts running the three-generation game on purpose.