VoxEU Column Industrial organisation Labour Markets Microeconomic regulation

Managing the family firm: Evidence from CEOs at work

The hypothesis that family firms are good for growth has come under scrutiny in recent years. This paper presents novel evidence on fundamental differences in behaviour between family and professional CEOs. Family managers tend to work at least 9% less than non-family ones, which is driven by their preferences for leisure and work. Family CEOs are typically wealthier and thus increase their consumption of leisure, which is a normal good. However, this behaviour may have adverse effects on family owned firms since hours worked by CEOs are strongly related with productivity. Given the ubiquity of family-run firms, this can impact the entire economy.

The exceptional economic success of many European countries in the post-War period was characterised by the wide presence of family firms across the Continent. Particularly, in countries like Germany and Italy, family ownership came to be seen as the best guarantee of economic and social development. However, the consensus that family firms are good for growth has come under scrutiny in recent years.

An emerging body of evidence indicates that family management is actually detrimental for performance (Bloom and Van Reenen 2007). Exploiting a remarkable natural experiment, Bennedsen et al. (2007) estimate a 4% profitability loss for Danish firms due to having a family manager rather than a professional one. Lippi and Schivardi (2014) find that family firms have worse executive selection (because they prefer to hire a less qualified family manager rather than an external professional manager) and this accounts for a 6% productivity loss as compared to conglomerate-owned firms. Given the magnitude of the estimated effect, family ownership might be a serious obstacle to productivity growth in Europe.

What does economic theory predict? Should firms be led by their owners or by professional managers? The argument in favour of owners is that they have more ‘skin in the game’. This is the standard prediction of the principal-agent model, as owners are residual claimants over the income generated by the business and hence are motivated to succeed, other things equal. The argument against is that, put simply, other things are not equal. In particular, firm owners are typically wealthier because they own the firm. Therefore, if leisure is a normal good, they might demand more leisure than professional managers. Given the ubiquity of family firms, understanding which of these effects prevails has important implications for aggregate income and growth.

Family managers work shorter hours

Our research (Bandiera et al. 2015) investigates the issue by conducting a time-use analysis of about 1,100 family and professional (i.e. non-family affiliated) CEOs in six developed and developing economies (Brazil, France, Germany, India, the US, and the UK). About 41% of the CEOs in our sample are family CEOs; 16% are professional CEOs working in family firms; and the rest are professional CEOs running non- family owned businesses.

Our methodology builds on the managerial ethnographic studies conducted by Henry Mintzberg in the 1970s, which we extend to cover large and random samples of managers. Every day of a randomly selected week we record every activity the CEOs undertake through daily interviews with the CEOs themselves or their PAs. This allows us to build an estimate of the total number of hours worked by the CEOs by summing up the duration of all the activities undertaken during the survey week. Compared to the standard recall measure, which we also collect, the bottom-up measure exhibits less bunching at round numbers and more variation.

Using this bottom up measure of CEOs labour supply, we find that family CEOs dedicate systematically fewer hours to work activities compared to professional CEOs (Figure 1).

The difference (at least 9% of total hours worked) is due to two factors: Family CEOs start work later in the day and are more likely to interrupt it to devote time to personal activities.

Figure 1. Total number of hours worked by family and non-family CEOs

Why do family CEOs work fewer hours?

Our results suggest that the fewer hours of work put in by family CEOs are driven by differences in their taste for leisure versus work, rather than by systematic ‘technological’ differences between family and non-family owned firms.

This conclusion is supported by two sets of findings.

  • First, observable CEO, firm, and organisational differences explain very little of the difference in hours worked between family and professional CEOs.

For example, one may argue that family CEOs may afford to work fewer hours because they can delegate their responsibility to other family members more easily. However, differences in the number of family members in management explain a small part of the difference in hours worked. Furthermore, professional CEOs in family owned firms work just as hard as their colleagues in non-family owned organisations, which rules out that family firms simply require less work.

  • Second, we show that the difference between family and professional CEOs depends on the opportunity cost of leisure.

For example, family CEOs working in larger firms and in more competitive industries – where the opportunity cost of CEO leisure is higher because slacking off would presumably have serious negative implications for the firm – work as many hours as professional managers. Symmetrically, family CEOs respond more strongly to exogenous ‘shocks’ increasing the marginal cost of their effort. We show this using data from the largest country in our sample, India, where we are able to exploit two proxies for shocks to the cost of providing effort common to all CEOs: Instances of extreme monsoon rainfall and the broadcasting of popular sport events (Indian Premier League cricket matches). During the monsoon season severe rains and floods often snarl traffic and make it particularly costly to go to work, while the Indian Premier League is an international sporting event that draws superstar cricket players and the eyes of an enthusiastic nation. The results depict a consistent picture. The difference between family and professional CEOs is significantly larger on days when torrential rains hit the region in which the CEO is located, and in days in which cricket matches are broadcasted.

The fact that CEOs seem to value leisure more than professional managers is consistent with the idea that family CEOs tend to be wealthier, as they own the firms they lead, and leisure is a normal good. In support of this, we find that the difference between family and professional CEOs is larger in countries where more permissive hereditary laws favour the concentration of wealth in the hands of the individual designated to inherit the control of the family business. However, the difference between family CEOs and professional managers is unaffected by other country characteristics, such as the level of development, rule of law and trust, which may inhibit the ability to nominate a hard working non-family manager in their role.

From CEOs hours worked to firm performance

  • Our results suggest that the differences in hours worked may have adverse implications for the firms led by family CEOs.

Indeed, hours worked by the CEO are strongly correlated with firm productivity. The elasticity of revenues per employee with respect to CEO hours is 0.36, comparable to the labour (0.48) and capital (0.26) elasticities, and similar results are found when we look at alternative measures of growth and profitability. While no causal inference can be made, combining this correlation with the effect of ownership on hours translates into a 2.6% productivity difference between family and professional CEOs. Given the ubiquity of family-run firms, when family firms are less productive the impact is felt beyond the family’s own profit margins and wealth — it affects the entire economy. The dampened productivity can add up to a lot — in reduced profits, in slower growth, and in lagging wages, all of which flow into the larger economy.

Overall, our findings provide novel evidence on a fundamental difference in behaviour between family and professional CEOs. This difference can be easily reconciled with the predictions of standard models of labour supply in the presence of wealth differentials across individuals. Our findings raise a public finance question. Would an increase in taxation that affects the owners of family firms bring about an increase in productive efficiency? Such taxation might include an inheritance tax, a wealth tax, or a reduction in the various forms of exemptions that family firms enjoy in many parts of the world.

References

Bandiera, O, A Prat and R Sadun (2015), “Managing the Family Firm: Evidence from CEOs at Work”, CEPR Discussion Paper 10379, January.

Bennedsen, M, K M Nielsen, F Perez-Gonzalez and D Wolfenzon (2007), “Inside the Family Firm: The Role of Families in Succession Decisions and Performance”, The Quarterly Journal of Economics, Volume 122( 2), May.

Bloom, N and J Van Reenen (2007), “Measuring and Explaining Management Practices across Firms and Countries,” The Quarterly Journal of Economics, 122 (4): 1351-1408.

Holtz-Eakin, D, D Joulfaian and H S Rosen (1993), “The Carnegie Conjecture: Some Empirical Evidence” The Quarterly Journal of Economics, Vol. 108( 2) (May): 413-435.

Lippi F and F Schivardi (2014), “Corporate Control and Executive Selection”, Quantitative Economics, Vol. 5: 417-456.

1,155 Reads