The seventy percent statistic is not evidence
Where the figure came from, what it actually counted, and the three questions that would tell a family where it really stands.
Seventy percent of wealth transfers fail. That figure is not a finding about wealth transfers. It is the inverse of a thirty percent business continuity rate, taken from public records on two hundred manufacturers in one American state.
The research behind it was carried out in the early 1980s and published in 1987. It asked one question: did majority family ownership of the company pass to the next generation. Seventy percent is what remains when thirty percent clears that bar.
The worry the figure is used to express is real. Families do lose enterprises, and they lose them for reasons visible years ahead. The figure measures none of those reasons.
The figure has one source.
Roy Williams and Vic Preisser published Preparing Heirs in 2003, reporting private research into the legacies of 3,250 wealthy families. The book names a seventy percent failure rate in estate transitions and sets out the causes it attributes that rate to.
In 2022 James Grubman published an audit of the claim in the International Family Offices Journal. He followed every footnote, endnote and reference in the Williams and Preisser writings back to the sources cited. One finding was consistent: the seventy percent rule comes only from John Ward's 1987 study.
The other authorities the book invokes hold no separate measurement. Grubman reports that the citation to the Massachusetts Institute of Technology resolves to a 1983 article by two other researchers, and that the citation to The Economist resolves to a 2001 piece naming no such rate.
Ward counted majority ownership of one company.
Ward analysed public records on a cohort of two hundred family businesses. All of them sat in a single industry, manufacturing, and a single region, Illinois. Success meant that majority family ownership passed to the next generation.
That is a business continuity rate. Continuity of majority ownership in one named company is the whole of what it observes.
A family that sells a manufacturer at a strong valuation and puts the proceeds into three new ventures is counted in the seventy percent. So is a family that appoints professional management and keeps every share. So is a family that closes a business it no longer wants.
Two of those three are decisions a governed family takes on purpose.
Four things the figure cannot tell you.
Each is a question families actually have. None of them is what was measured.
- Whether the wealth survived
- The study followed majority ownership of one company. A family's holdings can grow while that ownership ends and shrink while it continues. Ownership of a named firm and the wealth of a family are different quantities, and only the first was counted.
- Whether the family survived
- Nothing in the measurement observes whether relatives still speak to one another, whether the second generation holds a defined role, or whether anyone can state what the enterprise is for. Those are the outcomes most families mean when they say a transition failed.
- Whether the outcome was chosen
- A sale, an appointed chief executive and a wind-down all read as the same event in public records. A decision taken deliberately and an outcome nobody wanted are indistinguishable there, and the difference between them is a family's entire interest.
- Whether it applies to you
- The cohort was one industry in one American state, studied more than forty years ago. It speaks to no holding company in Milan, no family office in Geneva, and no operating group with businesses in four countries.
Counting families gives a different answer.
Grubman records that a well-crafted study in 2011 found essentially opposite results from Ward by asking what happens to business families instead of to individual businesses. Families pursuing several ventures showed significant longevity and success across generations.
The research that makes that case directly is Zellweger, Nason and Nordqvist, published in Family Business Review in 2012 and released online the year before. Close to ninety percent of the families in their sample held business activities beyond a single firm, controlling on average 3.4 companies.
Under that lens the unit of measurement is a family's capacity to keep building, and one company name surviving is incidental to it. A family that closes one business and starts two has failed at nothing.
What would actually tell you where you stand.
A statistic about other families settles nothing about yours. What settles it is a short list of questions with checkable answers, asked of your own records.
Can a material decision taken two years ago be produced today, with the reasoning behind it and the name of whoever approved it? Can two senior people, asked separately, give the same account of who may commit the enterprise at a given threshold? If the founder were unavailable for six months, what would stop?
All three are answerable within a fortnight, and the answers are evidence about your enterprise rather than about a cohort of Illinois manufacturers. They are also the ground the Governance Maturity Curve places an enterprise on, stage by stage.
We publish no figure for how many families fail. We have not measured one. What we publish is the method that places an enterprise on a stage, and the evidence a facilitator scores it against.
Questions this raises.
Is the seventy percent figure simply wrong?
It is an accurate reading of what it measured, which is whether majority family ownership of two hundred Illinois manufacturers passed to the next generation. It is not a measurement of wealth transitions, of family cohesion, or of anything outside that cohort. Wrong is the wrong word. Repurposed is closer.
Where did the figure originally come from?
John Ward's 1987 book Keeping the Family Business Healthy, reporting research carried out in the early 1980s on public records for two hundred family manufacturers in Illinois. Thirty percent passed majority family ownership to the next generation, and seventy percent is the remainder.
If the number is unreliable, is succession planning still worth doing?
Yes, and the case for it never rested on that number. Families lose enterprises for reasons observable in advance: authority nobody wrote down, a founder holding the only complete account of how things are decided, and a second generation with no defined role. Each of those is visible in your own enterprise today.
What should a family measure instead?
Whether a decision can be produced with its approval and its reasoning intact, whether authority is allocated in writing at stated thresholds, and whether the enterprise would keep operating for six months without the person who built it. All three are checkable from your own records, and none of them requires a comparison with anybody else.