The entire estate planning industry is built on a premise the data quietly contradicts: that wealth transmission is an event. A will is read, assets move, the transfer is complete. If that were how advantage actually...
The entire estate planning industry is built on a premise the data quietly contradicts: that wealth transmission is an event. A will is read, assets move, the transfer is complete. If that were how advantage actually worked, the statistical resemblance between parents' wealth and children's wealth would spike at inheritance and be flat before it.
That is not what the data shows. In Pfeffer and Killewald's study of multigenerational wealth (Social Forces, 2017), the correlation between parental wealth and child wealth grows as the children age: 0.33 when the children are 25 to 34, rising to 0.44 by ages 55 to 64. Parental advantage does not arrive in a single envelope. It keeps working on a child's balance sheet for four decades, strengthening the whole way, long after the children are themselves middle-aged and, in many cases, after the parents are gone.
The same study found a grandparent effect of 0.23: grandparental wealth predicts grandchild wealth even after accounting for the parents in between. Advantage is not even a two-generation story. It echoes.
Attach the critiques before building on the numbers. This is US panel data from the PSID, so magnitudes will differ elsewhere. Rising correlation with age is consistent with several mechanisms at once, including simple compounding of early gifts, and the study cannot fully separate them. Correlation is not a lever you pull. Granted, all of it. The shape of the finding still carries a practical instruction most families ignore: if advantage operates for forty years, a transfer plan that concentrates everything at death is mistiming almost all of it.
Three mechanisms plausibly drive the age-strengthening pattern, and each one is something a family can either run deliberately or leave to chance.
The safety net. Wealthy parents change their children's risk mathematics without transferring a cent. A child who knows a failed venture will not end in ruin can take the concentrated career bet, hold equity instead of selling early, decline the safe job, survive a divorce or an illness without liquidating at the bottom. Economists call this insurance value, and it pays out over decades, which is exactly the signature the rising correlation shows. The net catches things in year 3 and in year 33. Most of its value is never drawn down at all. It works by existing.
Timely help at trigger moments. Life has a small number of doors that swing on capital. The first property title, where a down payment gift converts rent into the 28.4 percent channel, homeownership, years earlier than a salary alone would allow. The business start, where modest early capital buys survival through the fragile first years. The crisis, where a bridge loan from family prevents a fire sale that would have set the child back a decade. The same total sum, delivered at these hinges, buys multiples of what it buys as a lump at 60. The correlation keeps rising with age partly because these well-timed interventions compound: the house bought at 29 has appreciated for thirty years by the time the study measures the child at 60.
Norms. The slowest channel and possibly the deepest. Children absorb their parents' defaults about saving, debt, ownership, and patience, and then run those defaults for a lifetime. A norm set at 10 is still producing balance sheet outcomes at 55. This is also the most likely carrier of the grandparent effect: a grandmother's habits, transmitted as family culture, shaping a grandchild's choices decades after her death. You cannot bequeath a norm in a will. You can only install it while you are alive and present.
Take the mechanisms seriously and the design conclusion follows: a wealth transfer is not one decision, it is a 40-year drip, and the drip should be planned with the same care families currently reserve for the will.
Bill Perkins makes the aggressive version of this argument in Die With Zero (summary). His observation: on the standard plan, children inherit when the parents die, which in long-lived families means the heirs receive money around age 60, precisely when their capacity to convert money into a changed life has collapsed. Perkins argues for deliberate giving during what he treats as the high-conversion window, roughly ages 28 to 33, when recipients are old enough to handle money and young enough for it to redirect a life: a home, a family started without financial panic, a risk taken. He calls the deathbed default a failure of planning, not a display of prudence.
You do not need to accept Perkins' full program, and his critics fairly note that dying with zero is a poor fit for families intent on multigenerational continuity. But his timing point and Pfeffer and Killewald's correlation curve are two views of the same fact. Advantage delivered early compounds for decades. Advantage delivered at the end arrives after most of the compounding runway is gone.
So plan the drip. A serious 40-year transmission plan has at least four instruments, run in parallel:
Warm-hand windows. Identify the 28-to-33 window for each child and decide in advance what crosses during it: the down payment match, the seed capital, the education completion. Giving with a warm hand also returns information the grave never gets. You see what the child does with the first tranche, and the plan for the second tranche gets smarter.
Standing safety net. Make the net explicit instead of assumed. Children calibrate risk against what they believe the family will do. An articulated policy, what the family will backstop, on what terms, and what it will not, produces better risk-taking than a vague hope, and better discipline than an unlimited implied bailout, which is how safety nets curdle into dependency.
Trigger-moment readiness. Keep capacity liquid for the three or four hinge moments per child: first title, business start, crisis. These cannot all be scheduled, so the plan is really a reserve plus a decision rule, agreed while everyone is calm.
Elder mentoring. The 0.23 grandparent effect is the empirical case for keeping elders in the transmission loop, deliberately. Grandparents carry the norms, the stories, and the long view, and they typically have the time the parents in the middle do not. A family that structures elder-grandchild contact, through councils, apprenticed roles, or simple standing rituals, is running a documented channel. A family that lets it default to birthdays is leaving a measured effect on the table.
Notice what this reframes. The question "how much do we leave the children" becomes "what does each decade of each child's life need from us, and what does it need us to withhold." Withholding is part of the design. A drip is not a firehose; the correlation data describes advantage that supported children's own outcomes, not advantage that replaced them.
The standard pushback to early giving deserves a straight response: won't it soften them? Give a 29-year-old a down payment and you may be buying complacency instead of a head start.
Two answers. First, the risk is real but it attaches to the structure of the gift, not its timing. Unconditional cash with no expectations attached carries dependency risk at any age, including 60. A matched down payment, where the child brings half, or seed capital released against a written plan and reviewed like an investor would review it, transfers advantage while leaving the effort where it belongs. The families that get burned by early transfers almost always skipped the structure, not the timing.
Second, the deathbed default does not avoid the softening problem, it just hides it until you cannot correct it. A child who spends thirty years expecting a large inheritance is shaped by that expectation the entire time, and you learn what the money does to them only after you are gone. The drip gives you the one thing the lump sum never can: feedback while there is still time to adjust. A first tranche handled well earns the second. A first tranche handled badly is the cheapest tuition your family will ever pay, and it repriced the whole plan.
Here is the exercise, and it fits on one page per child. Sketch your child's four decades: 25 to 65, in ten-year blocks. Write what each block will likely contain, first home, first venture, growing family, peak career pressure, their own children's launches, your own decline. Then mark the three moments where your help, money, presence, or standing behind them, will matter most, and write what you intend to do at each one. Date the page. Revisit it annually, because the marks will move.
That page is not a substitute for the will. It is the part of the plan the will cannot do, covering the forty years the data says the transmission actually happens in.
This piece did its job if you stop thinking of your estate plan as the transfer, and start treating it as the final installment of a drip you began decades earlier, on purpose.