What Is a Family Firm and Why Does It Matter?

To illustrate why defining a family firm matters, consider two iconic companies: Hermès and Suzuki. Hermès is a luxury brand with deep family ownership and succession across generations. Suzuki has long been controlled by the founding family for despite minimal ownership. Both are family firms by some definitions—but not by others. In contrast, Microsoft under Bill Gates had strong founder control but no succession intent or any evidence of family involvement. Should Microsoft be classified a family firm? 

These cases highlight the definitional ambiguity that plagues family firm research. Our goal is to clarify this landscape and offer a workable definition. We take inspiration from Donnelley (1964), who defined a family firm as one where at least two generations are closely involved and influence both the company’s direction and family goals. This intergenerational embeddedness is central to identifying what makes a firm truly "family."

Clearly the spatial and temporal embeddedness of the family in the business is a central theme that not only identifies common features within the class of firms called family businesses, but also distinguishes this class of firms from non-family firms. For e.g., the temporal embeddedness is a common feature of succession practices in family firms, but clearly not in non-family firms. Spatial embeddedness can be thought of as the involvement of multiple family members in the business. In the parlance of Williamson (1979, 1983), the family and the business represent a bilateral monopoly, with each side making specialized investments in the other, wherein the value of each is diminished in the absence of the other. 

Why Definitions Matter

Definitions shape empirical outcomes. Depending on the criteria used, the number of family firms in a dataset can vary dramatically. For example, in the U.S., using a 5% ownership threshold, 37% of Fortune 500 firms qualify as family firms. Add succession and spatial embeddedness, and the number drops to 7%. 

This variance affects:

1. Prevalence: Counting founder-managed firms like Microsoft inflates the family firm economic footprint.
2. Performance: Villalonga and Amit (2006) shows that family firms outperform only under the broadest definition, but that this result flips when succession and embeddedness are added. 
3. Innovation: Mixed findings on R&D investment and efficiency also track with definitional differences.

In short, definitional choices affect everything from policy recommendations to theoretical conclusions.

The Five Conceptual Dimensions

We identify five key dimensions that define family firms, which often interact in complex ways:

1. Ownership: Most common criterion, with thresholds from 5% to 50%. Some distinguish between cash flow and voting rights.
2. Control: Often proxied by board presence or voting rights; for example, the Suzuki family exerts control despite minimal equity.
3. Management: Is a family member the CEO or in top management? This dimension distinguishes active family involvement from passive ownership.
4. Embeddedness: Shared values and multi-member involvement; a proxy for long-term commitment and human capital investment.
5. Succession: Key for dynastic firms; distinguishes from founder-led startups without generational intent.

Our Survey: 122 Papers, 29 Definitions

We reviewed 122 empirical papers from 1964 to 2022 and identified 149 unique definitions. Results show:

• Ownership used in over 80% of definitions, often as a sole criterion. 
• Management appears in ~20%.
• Embeddedness and succession appear in only 14% and 8%, respectively.

About 44% of definitions rely on a single criterion. This oversimplifies the complexity of family firms and complicates cross-study comparisons.

What can we learn?

Early research on family firms offered a remarkably clear and conceptually rich definition of what constitutes a family business. In his seminal 1964 article, Donnelley proposed the idea of embeddedness as a key identification feature of a family business. He notes that the embeddedness feature bestows a mutual reciprocal reliance of the family and the business in setting firm policy and family objectives. 

This definition emphasized two foundational dimensions—succession and spatial embeddedness—that distinguish family firms from other organizational forms. Yet, despite the clarity and depth of this early work, the field has since struggled to converge on a consistent definition. Over the decades, empirical studies have increasingly relied on narrow, data-driven criteria—most commonly ownership thresholds—while often neglecting the intergenerational and relational aspects that Donnelley considered essential. As a result, the definitional landscape remains fragmented, with significant implications for how we count, compare, and understand family firms.

Our findings suggest:

• Broad definitions may misclassify firms like Microsoft, skewing data.
• Narrow definitions may exclude dynastic firms like Suzuki.
• A multi-dimensional approach, particularly one that incorporates succession and embeddedness, captures the essence of what defines a truly 'family' firm.

• Researchers should test alternative definitions to check the stability of their findings.

The field must move beyond convenience-based definitions toward conceptually grounded, empirically feasible frameworks. While we do not prescribe a single definition, we strongly advocate for incorporating Donnelley’s (1964) twin pillars of succession and embeddedness. These dimensions best capture what makes a family firm distinct.

______________

Morten Bennedsen is Professor at the University of Copenhagen, Visiting Professor at INSEAD, and an ECGI Research Member.

Yi-Chun Lu is an Assistant Professor at National Sun Yat-sen University.

Vikas Mehrotra is the Stanley A. Milner Professor and Dean at the Alberta School of Business, University of Alberta.

The ECGI does not, consistent with its constitutional purpose, have a view or opinion. If you wish to respond to this article, you can submit a blog article or 'letter to the editor' by clicking here.