The wealth transfer canon presents itself as a body of universal truths about families and money. It is actually a set of regional case studies wearing a universal costume.
The wealth transfer canon presents itself as a body of universal truths about families and money. It is actually a set of regional case studies wearing a universal costume.
Run the audit yourself. Take the statistics that anchor nearly every article, book, and seminar on generational wealth, and check where the data came from.
The finding that wealth advantage compounds across generations rests heavily on the PSID, the Panel Study of Income Dynamics: a longitudinal survey of American households, running since 1968, sampled to represent the United States. The headline numbers on the great wealth transfer, $124 trillion moving by 2048, come from Cerulli Associates, a Boston research firm modeling American households. The famous claim that 70 percent of wealth transfers "fail," and that failures trace to trust and communication breakdowns rather than bad legal work, comes from Williams and Preisser, who surveyed families in the United States and Canada through their post-transition research. The "shirtsleeves to shirtsleeves in three generations" evidence base, the 30-13-3 percent survival rates cited in almost every family business talk ever given, comes from John Ward's 1980s study of manufacturing firms in Illinois.
The United States, Canada, and Illinois. That is the empirical foundation under claims routinely presented as laws of family nature, quoted in Lagos boardrooms and Kampala seminars as if they were measured there.
This is not an accusation of bad faith. Researchers study the data they can get, and rich countries fund panel studies. But a reader in Accra or Nairobi consuming these numbers unlabeled is being handed American measurements as if they were human constants, and nobody in the supply chain has much incentive to add the disclaimer.
The honest position is not "Africa has no data." It has data: thinner, younger, and pointing somewhere interesting.
PwC's Africa Family Business Survey is the closest thing to a continental baseline. Its African samples consistently show family firms that are younger than their global peers, far more likely to be first or second generation, growing faster, and dramatically less prepared on paper: minorities have a documented succession plan, a will covering the business, or formal governance of any kind. The survey's limitation is built into its method: PwC surveys firms large and formal enough to be PwC's audience, so the informal enterprises that dominate African economies are invisible to it.
The Henley Africa Wealth Report, built on New World Wealth's tracking, counts roughly 122,000 dollar millionaires on the continent, concentrated in the "Big 5" markets of South Africa, Egypt, Nigeria, Kenya, and Morocco. Its critique is sharper: New World Wealth's methodology is proprietary, built on databases of known individuals plus estimation models, and academics have questioned its precision for years. Treat the 122,000 as a rough census, not a count. Even as a rough census it establishes the point that matters here: Africa has a substantial and growing private wealth class whose transfer behavior almost nobody is studying longitudinally.
A five-country academic study of 185 African family firms examined founder-to-successor transitions directly. Its findings rhyme with the Western literature: successions succeed where founders plan early, communicate, and prepare successors deliberately, and fail where they do not. The study also surfaces distinctly African mechanics, including extended-family claims on the estate and the pull of communal obligation on business assets. Its limits are the usual ones for the region: a modest sample, cross-sectional rather than longitudinal, and firms that agreed to be studied.
Set the two shelves side by side. The Western shelf has fifty-year panel studies and trillion-dollar projections. The African shelf has practitioner surveys, a contested millionaire census, and early academic work. That asymmetry is the missing data of this article's title, and it should permanently change how you read everything sitting on the first shelf.
Here is the discipline this brief argues for: apply Western findings to non-Western families as hypotheses to test, never as laws to obey. Some will survive the journey. Some will not. You can often predict which by looking at what the finding depends on.
Findings that depend on institutional context travel badly. The American estate planning canon is shaped by an estate tax that most African jurisdictions simply do not have; Kenya, Nigeria, and others levy no inheritance tax at all, and Uganda's estate duty has been dormant for decades. Strategies built to dodge a 40 percent federal estate tax are answers to a question many African families are never asked. Likewise, three-generation survival statistics measured on Illinois manufacturers describe firms embedded in deep capital markets, formal titles, and exit options. A Lagos trading firm faces different mortality causes: currency shocks, informal competition, succession contests widened by polygamous and extended-family structures that the Western data never sampled.
Findings that depend on generational stage need translating. Most of the Western canon studies wealth that is old: families managing, preserving, and losing fortunes built generations ago. African private wealth is overwhelmingly first generation; the founders are alive, often still in the founder's chair. That means the continent's dominant transfer problem is not the third-generation decay the proverb warns about but the first handover ever attempted, which is a different problem with different failure modes, closer to what Ward's firms faced in generation one than to anything in the dynasty literature.
Findings about human behavior probably do travel. The Williams-Preisser claim, whatever its methodological softness, locates transfer failure in trust breakdown, poor communication, and unprepared heirs rather than in documents. The 185-firm African study found essentially the same thing in a different hemisphere. Founders who plan early, talk openly, and prepare successors succeed more often on both continents. This convergence is worth noticing precisely because the contexts differ so much: when the same behavioral core shows up under an estate tax and without one, in fifth-generation firms and first-generation ones, it is probably close to a real human constant. The mechanics change. The psychology appears not to.
So the practical reading rule: the further a statistic sits from raw human behavior and the closer to institutions, taxes, and market structure, the less it transfers. Communication findings travel. Tax strategies do not. Survival percentages are somewhere in between and should be quoted with their birthplace attached.
Because families act on these numbers. A Ugandan founder who hears "70 percent of transfers fail" as a universal law may fatalistically accept decline, when the honest statement is "70 percent failed in a North American sample, for behavioral reasons you can address, and your context differs in ways that cut both directions." A Kenyan family that imports American trust structures wholesale may buy complexity designed for a tax it will never pay, while leaving unaddressed the extended-family claims that actually sink estates in its jurisdiction, the risk its local data flags and the imported canon never mentions.
And because the gap is closable. Africa's wealth is young enough that the continent could build its transfer evidence base in real time, watching first handovers as they happen instead of reconstructing them decades later. PwC's surveys, the Henley census, and the emerging academic work are the start of that shelf. Every African family office, SACCO federation, and business school that starts tracking successions adds to it. The next generation of wealth statistics does not have to be Western by default.
Until then, provenance checking is a reader's own job. It takes about ten seconds. Who was sampled, where, and when. If the answer is American households in 1989 and the audience is a family in Kampala in 2026, the number is not disqualified. It is demoted, from law to hypothesis, which is where it should have been all along.
Adopt one habit: when you read a wealth statistic, ask whose data it was. Country, sample, year, and who paid for the study. Quote the answer alongside the number whenever you repeat it, and treat any finding measured far from your context as a hypothesis to test against your own family's reality rather than a verdict on it.
This piece did its job if the next time someone tells you 70 percent of wealth transfers fail, your first question is not "how do we avoid that fate" but "measured where, on whom," and only then, with the number properly labeled, do you put it to work.