Shirtsleeves to Shirtsleeves in Three Generations: What the Research Says

"From shirtsleeves to shirtsleeves in three generations" is the most repeated line in the family business world. The first generation works with its sleeves rolled up and builds something. The second enjoys it. The third inherits it, doesn't know how it was made, and loses it. Conference speakers use it to open talks. Advisors use it to sell services. Families use it, half-joking, at their own dinner tables.

It's a vivid story. This article asks whether it's also true. It traces where the proverb comes from, looks at the statistic that is usually stapled to it, and explains why that comparison misleads. Then it covers what research suggests actually helps a family stay entrepreneurial and wealthy across generations. For the wider picture of how ownership and leadership pass between generations, start with the family business succession hub.

The Proverb Is Older Than the Statistic

The phrase is often credited to Andrew Carnegie. The Word Histories site traces it differently: the earliest citation it lists is from Appletons' Journal on June 27, 1874, with later appearances in The Atlantic Monthly in June 1882 and a Pennsylvania newspaper in April 1883. It finds no evidence supporting the Carnegie attribution. A British version, "clogs to clogs in three generations," also circulates.

That timeline matters. The proverb was already common in the 1870s and 1880s, long before anyone surveyed a family firm. It began life as a folk observation about money, not a finding about companies. The measurement came later, and it got glued to a saying that people already believed.

Notice also what the proverb itself describes. It's about a family's wealth and habits. It says nothing about whether a particular company keeps trading. That distinction is the center of this whole debate.

The Statistic Everyone Quotes

The figure usually attached to the proverb says that only a minority of family firms survive into the second generation and a smaller group into the third. It traces to a 1987 book. Family Business Magazine's account of the research says John L. Ward's 1987 study reported that only 30% of family companies survive the second generation and just 13% make it through the third. The same account says Ward's sample was about 200 companies, drawn from a single state (Illinois) and one industry (manufacturing).

Even the primary reference is worded loosely. The Kellogg School's catalog entry for Ward's book, Keeping the Family Business Healthy, says that less than one-third of family-owned businesses survive until the second generation. That's a related claim but not the same sentence as the one Family Business Magazine reports. Anyone quoting "the" three-generation figure is choosing a version.

You'll also see a fourth-generation number, a few percent, attached to it in speeches and slide decks. It usually travels without any study attached, so it deserves even more suspicion than the Ward figures. Find where it was first published before you repeat it.

Three things stand out about the origin:

• It's one study. Not a body of research, a single sample.
• It's regional and sectoral. Illinois manufacturers in the 1980s aren't a stand-in for a Malaysian trading house or a German engineering firm.
• It has been repeated for decades. Repetition made it feel settled, which is different from being replicated.

What the Critics Say

Two lines of criticism are worth knowing, and they attack different things.

The "it doesn't generalize" critique

The researchers Robert Nason, Mattias Nordqvist and Thomas Zellweger led a longevity study for the Family Firm Institute and Goodman. According to Family Business Magazine, they described Ward's survival numbers as "low, out of context and not generalizable." Their point about context is the interesting one. The same article says they set Ward's figures against general business longevity, where 50% to 60% of all new companies fail within five years and only 25% last a decade. Against that background, a family firm that reaches a second generation is beating the usual odds, not falling short of some higher standard.

The "wrong unit of analysis" critique

Zellweger, Nason and Nordqvist's 2012 paper in Family Business Review, "From Longevity of Firms to Transgenerational Entrepreneurship of Families", argues for shifting from the firm to the family as the thing you study. In the University of St. Gallen repository's abstract, the authors say that looking at the family level reveals entrepreneurial activity that you miss when you focus only on one firm. A family that sells its original company and runs three others hasn't failed. The firm ended. The family's enterprise didn't.

The 2021 Harvard Business Review challenge

In July 2021, family business advisors Josh Baron and Rob Lachenauer published "Do Most Family Businesses Really Fail by the Third Generation?" in Harvard Business Review. The article opens by describing the three-generation rule as conventional wisdom, one that says most family businesses don't survive beyond three generations, and goes on to question it. Baron has called much of the failure-rate talk "nonsense" and says the real accomplishment is simply making a business last about 90 years, or three generations. The full text of the HBR piece sits behind a subscription, so we cite it here only for its framing and its title question.

Why the Comparison Misleads

Put the critiques together and the problems with the headline statistic become clear.

It measures firms, but the proverb is about families. A company can be sold, merged or closed while the family's wealth grows. The reverse is also true: a company can limp on for decades while the family around it fractures. A firm-survival percentage can't tell you which one happened.

It has no base rate. "13% survive to the third generation" sounds terrible until you ask what percentage of non-family firms do. Most businesses of any kind don't last 60 or 90 years. Without a comparison group, the number can't be read.

It treats an ending as a failure. Founders sell for good reasons: a strong offer, no interested heir, a market that has shifted. A planned sale to a strategic buyer is a successful exit, and the statistic files it next to bankruptcies.

It ignores who was in the sample. One state, one industry, one decade. The statistic traveled much further than the data did.

None of this means succession is easy or that families never lose companies. It means the dramatic version of the claim isn't supported by the dramatic version of the evidence. The honest version is narrower: passing a business between generations is hard, and many firms don't make it, for reasons that are mostly the same ones that affect all firms plus some that are particular to families.

What Research Suggests Helps Families Last

If survival of a single company is the wrong yardstick, what's the right one? The Family Firm Institute and Goodman team studied it directly. According to the Family Business Magazine summary, they surveyed 541 executives and received 118 usable responses. Among those respondents, 89.4% of families controlled multiple businesses, averaging 3.4 firms. Those families had also divested underperforming assets and acquired an average of 2.7 companies. The authors' reading was that transgenerational wealth rests on the family's entrepreneurial adaptability, not on keeping one legacy company alive at all costs.

A sample of 118 isn't a census, and it comes from families who agreed to respond. But the pattern is consistent with the 2012 paper's thesis, and it suggests a few practical principles:

1. Treat the family as the investor, not just the owner. Families that hold a portfolio can let a weak business go without losing their identity.
2. Keep entrepreneurship alive across generations. The 2012 paper's central idea, family entrepreneurial orientation, is about the family continuing to start, buy and reshape businesses, not guarding one.
3. Separate loyalty to the family from loyalty to a specific company. That's a hard conversation, and it's easier to have before a crisis.

Other disciplines matter too. Families that write down their rules for ownership, employment and decision-making tend to argue less about them when stakes are high, which is the territory of a family constitution. The stage a family has reached also changes the problem: a founder-led firm, a sibling partnership and a cousin consortium need different structures, as the family business lifecycle explains. And for firms that have lasted centuries, the habits they share are covered in long-lived family firms.

Where the Proverb Is Right

It would be a mistake to throw the proverb out entirely. It points at three real risks, even if its arithmetic is shaky.

Founder-era habits don't transfer automatically. The first generation often builds through intuition, risk tolerance and personal relationships. Those traits are hard to teach, which is part of why next-generation leadership deserves deliberate preparation rather than hope.

Ownership spreads faster than understanding. Each generation tends to add shareholders, and many of them never work in the business. Planning how shares move is its own discipline, covered in intergenerational ownership transfer.

Key-person dependence is real. When one person holds the customer relationships, the know-how and the final say, the transition is a risk event whatever the family tree looks like. See key-person risk and the general principles of succession planning.

The right reading of the proverb, then, is as a warning about what happens when a family doesn't plan, not a forecast that it will fail.

Using the Statistic Responsibly

If you write, advise or present on this topic, a few habits keep you out of trouble:

• Say "often cited" rather than "research shows" unless you can name and link the study.
• Name the source and the sample: Ward, 1987, roughly 200 Illinois manufacturing firms.
• Don't print a fourth-generation figure unless you've found where it was first published.
• Pair any survival number with a comparison group, because a percentage with no baseline invites the wrong conclusion.
• Be specific about the unit. "Firms survive" and "families stay wealthy" are different claims.

Key Facts: The Three-Generation Rule

• The phrase "shirtsleeves to shirtsleeves" appears in print by 1874, with no evidence supporting the common Carnegie attribution (Word Histories).
• The oft-quoted figures come from John L. Ward's 1987 study of about 200 Illinois manufacturing companies, reported as 30% surviving the second generation and 13% the third (Family Business Magazine).
• Nason, Nordqvist and Zellweger called those figures "low, out of context and not generalizable" (same source).
• Their research team found 89.4% of the responding families controlled multiple businesses, averaging 3.4 firms (same source; 118 responses).
• A 2012 Family Business Review paper argues for studying the family, not the firm, as the unit of transgenerational entrepreneurship (University of St. Gallen repository).
• Harvard Business Review published a 2021 article questioning the rule (Baron and Lachenauer, HBR).

Related Reading

• Family business succession
• Long-lived family firms
• Family business lifecycle
• Intergenerational ownership transfer