When Does the Family Govern the Family Firm?
There is extensive evidence on how the family firm’s performance correlates with the controlling family’s participation in the firm’s governance. In contrast, there is very little evidence on what determines this participation in the first place. We characterize the settings where the family chooses to participate vs. not participate in governance, trying also to disentangle the possibly two-way causal relationship between participation and performance.
We use proprietary data from about 70,000 private Norwegian family firms. Most family firms in the world are private, and private firms constitute a much larger part of the economy than public firms do. For instance, 99.7% of all Norwegian firms are private, and private firms are four times larger than public firms by aggregate sales, employment, and assets. Moreover, about 75% of all firms in the world are family firms. Thus, representative samples of family firms must include private firms. The family firms we study constitute 71% of all firms in the Norwegian economy. We define a family firm as one that is majority-owned by individuals related by blood or marriage.
We find that the average family participates very actively, holding both the CEO and chair positions in 79% of the firms. The family participates more intensively the more equity the family owns in the firm, the smaller the firm, the more profitable the firm, and the less risky the firm. There is no evidence of a life-cycle effect, as the relationship between family participation and family firm characteristics is very similar in young and old firms. These relationships hold in both single-owner and multiple-owner family firms, when we use alternative definitions of the family firm, and when we measure participation in alternative ways.
We propose four economic rationales for these results. First, the positive relationship between participation intensity and ownership concentration supports the idea that a larger investment gives stronger incentives to be active in governance because more wealth is at stake. The relationship may also reflect that the controlling family tries to mitigate conflicts with minority shareholders, as intensive governance by the controlling family is less threatening to minority shareholders the more equity the family owns. This concern for shareholder conflicts is also supported by our finding that participation is less intensive in multiple-owner firms than in single-owner firms, where shareholder conflicts do not exist.
Second, our finding that the family is less active in larger firms supports the idea that governing larger firms requires skills that are harder to find inside the family than in the much greater pool of candidates outside the family. Thus, the family’s desire to recruit the most talented officers and directors seems stronger than the temptation to use these positions to extract private benefits more easily.
Third, the inverse relationship between participation and risk supports the notion that the family bears a cost for being undiversified. Taking employment at the same firm that produces most of the employee’s wealth adds more to the costs of being undiversified the riskier the firm. Hence, higher risk in the family’s financial investment induces the family to reduce the risk in its human capital investment.
Finally, the positive association between participation intensity and performance may reflect that the family tailors its participation to the firm’s economic health. This result supports the idea that participation is not exogenous to the firm’s prospects, and that causation also runs from performance to participation because the family uses past performance to self-select. We find that causation does run both ways, and that the positive effect from performance to participation is twice as large as the positive effect from participation to performance. We also show that the widespread practice in the literature of ignoring the family’s self-selection produces biased estimates of how governance affects performance.
Accounting for self-selction in corporate governance is rare in studies of how the controlling family’s participation interacts with the firm’s performance. Our evidence may partially explain why this voluminous literature has produced very ambiguous results.