Ask a family to draw its financial plan and you will almost always get a two-generation picture. Parents earn, children receive. The grandparents, if they appear at all, appear as a duty: a monthly remittance, a medical...
Ask a family to draw its financial plan and you will almost always get a two-generation picture. Parents earn, children receive. The grandparents, if they appear at all, appear as a duty: a monthly remittance, a medical bill, eventually a funeral. They are treated as the past of the family's money, not as part of its machinery.
The data says this picture is wrong by a full generation.
The measurement comes from Fabian Pfeffer and Alexandra Killewald's study "Generations of Advantage," published in Social Forces in 2017. Working with the Panel Study of Income Dynamics, which has followed American families since 1968, they could observe three generations of the same families as adults, something almost no dataset on earth allows. Their headline finding for our purposes: the correlation between a grandparent's wealth and a grandchild's wealth is 0.23.
To feel the size of that number, put it next to the parent figures from the same study. Parent and child wealth correlate at about 0.33 when the child is a young adult, rising to around 0.44 by late middle age. The grandparent signal, at 0.23, is not far behind the early parent signal, and it belongs to people the grandchild may barely have known. More striking still, roughly half of the grandparent association does not pass through the parents at all. If wealth flowed the way the two-generation picture assumes, the grandparent effect would be fully absorbed into parent wealth: grandfather helps father, father helps you, chain complete. The data refuses that story. A meaningful part of the grandparent's influence reaches the grandchild directly, around the middle generation rather than through it.
This is not one study standing alone. The sociologist Robert Mare, in his 2011 presidential address to the Population Association of America, published in Demography as "A Multigenerational View of Inequality," argued that the entire field had been measuring mobility on a two-generation ruler and missing the longer machinery of families. Two years later, Tak Wing Chan and Vikki Boliver tested the idea against three British birth cohort studies covering roughly 17,000 people. Their paper, "The Grandparents Effect in Social Mobility" in the American Sociological Review, found that even after fully accounting for the parents' class position, grandchildren of professional and managerial grandparents had at least two and a half times better odds of reaching the professional class themselves than grandchildren of unskilled manual workers. Same conclusion, different country, different measure: the third generation is connected to the first by channels the second generation does not fully control.
So the practical question is not whether grandparents transmit. They do, measurably. The question is what exactly travels down that wire, because a family that can name the cargo can load it deliberately.
Money. The most visible channel is direct transfer that skips a generation. Grandparents pay school fees, and fees paid at the right term matter more than the same sum bequeathed thirty years later. Grandparents gift land or contribute the deposit on a first plot, which puts the grandchild onto the property ladder early; in Pfeffer and Killewald's decomposition, homeownership is the single largest transmission channel between generations, accounting for 28.4 percent of the wealth correlation, with education next at 25.5 percent. A grandparent who funds a degree or a title deed is not giving a gift. They are operating the two heaviest channels in the published data, one generation deeper than anyone expects them to.
Time. The least counted channel, and possibly the largest in African households. When a grandmother minds children, the family books it as help. The correct entry is a transfer of earning capacity to the middle generation. Janice Compton and Robert Pollak, in "Family Proximity, Childcare, and Women's Labor Force Attachment," found that married American women with young children who live close to their mothers or mothers-in-law have employment and labor force participation rates 4 to 10 percentage points higher than those who do not, an effect they attribute largely to available childcare, both routine and emergency. Read that as a wealth statement: the grandmother's presence shows up in the daughter's payslip. Every school run an elder covers is income the parents did not have to forfeit, and that income compounds for decades.
Norms. The third cargo is the money script itself, observed at the source. Children learn money less from what they are told than from what they watch, and a grandparent is often the only person a child ever watches handle money with full authority: negotiating at the market, refusing a bad loan, setting aside the tithe, recounting what the land cost and what it took to keep it. The template transmits in both directions, which is the uncomfortable half; a grandparent who hides money, hoards it, or gambles it teaches that with equal efficiency. But the mechanism itself is free, requires no estate, and runs on nothing but proximity and speech.
Here the usual direction of comparison should be reversed. Development literature tends to frame the African extended household as a burden: dependency ratios, black tax, too many claims on one salary. But look at it through the transmission data and the structure holds an advantage the Western nuclear household has engineered away.
In a typical American or European family, the generations live apart, often in different cities. The grandparent effect measured at 0.23 in the United States operates mostly by cheque and telephone. In much of Africa, the grandparent is not a voice on a call; she is at the table. Grandchildren in three-generation households watch money decisions being made by the person with the longest memory of the family's finances, daily, for years. The norms channel, which American data can only observe in weakened long-distance form, runs at full strength. If the direct grandparent effect is measurable across American distances, a family that eats dinner with its elders is holding a stronger version of the same asset. The question is whether it is being used deliberately or just ambiently.
Ambient is the default. The elder is loved, fed, consulted about weddings, and given no role whatsoever in the family's money formation. Her influence on the grandchildren happens in whatever hours accident allows. That is a 0.23 channel left to run on idle.
Deliberate looks like this: the family names the role, out loud, in the family council.
The storyteller seat at Numbers Night. If your family holds a regular money meeting, reserve a seat and a segment for the eldest generation. Not to approve budgets; to narrate. The first job, the first plot, the debt that nearly sank the family, the drought year and what survived it. This is the norms channel converted from accident to appointment. It costs nothing and it is the one item on the agenda children reliably remember.
The grandchildren's pot they co-fund. Where the elders have means, route their giving through structure instead of ad hoc envelopes. A named pot per grandchild, defaulted to education, with the grandparent as visible co-funder, does three things at once: it operates the education channel directly, it lets the elder watch the money work while alive, which is what elders consistently say they want, and it converts a private gift into a public norm the whole family can see.
The time transfer, acknowledged. If a grandmother's childcare is what lets both parents earn, say so in the family's accounting. A family that names her contribution can also reciprocate it deliberately, in housing, in health cover, in the seat of honor, rather than treating her labor as free and her upkeep as a burden. The channel runs better maintained than neglected.
None of these roles requires the grandparent to be wealthy. The storyteller seat and the time transfer run entirely on presence and memory. The mechanism Pfeffer and Killewald measured in American panel data is available to any family with a living elder and the discipline to assign the role.
Every channel in this essay has one property the correlation tables do not print: an expiry date. The fees, the childcare, the stories at the table all require the grandparent to be alive, and the deadline is invisible until it has passed.
So the decision is immediate. Your family either gives its eldest generation a named role in the money formation of its youngest, this year, or it accepts that its strongest documented transmission channel will keep running unaimed until the funeral closes it. Name the storyteller. Open the pot they co-fund. Or explain to your grandchildren, later, why the 0.23 in your family carried nothing at all.