In a recent intervention at Bloomberg, Luigi Zingales warned about the low sustainability of organizational rigidity and the more personal than professional management of family businesses. These companies tend to follow corporate policies that reflect the faithful representation of shareholder preferences, the conservation of the family legacy and a strong incrustation in local social norms (evidence here). Its particularities and its documented prevalence around the world, including among large companies, made the interest in its activities and economic impact increase exponentially.
With the severity of the effects of the recent financial crisis still in the spotlight, a question that emerges is whether there were negative effects of different intensity depending on the ownership structure and the control of the companies. Some would argue that, following more conservative strategies, family businesses would reduce investment more during the crisis, thus affecting economic recovery but ensuring that the family would maintain control of the company. Others would be expected that, due to having fewer internal conflicts of interest and a greater ability to contract financing in a relational manner, family businesses would have an advantage during a banking crisis.
In a recent paper -Amore and Epure 2018- we studied the mechanisms underlying the advantages and disadvantages of the family organizational model. To reveal which models of property and corporate governance have the worst effects in a financial crisis, we go one step beyond the traditional approach based on the formal characteristics of the company and the markets: we explain how a cultural attribute in the region where the headquarters is located of the company, the generalized confidence, shapes the effect of the family organizational model on the performance and financing of the company during a crisis.
Kenneth Arrow already indicated in 1972 that there is an element of confidence in any transaction, and economists used abundantly the differences in this cultural attribute to explain individual and company behavior. To measure the level of generalized confidence, surveys are used asking: “Would you say that most people can be trusted? (with the counterpart: “Or that one is never prudent enough when dealing with others?”). The implications of this trust are key to business behavior. Regions with a higher level of generalized confidence are more oriented to impersonal exchanges in the market, while those with lower levels are characterized by personal exchanges in more closed networks; In other words, generalized trust helps lower transaction costs.
There is a lot of evidence about the beneficial effects of generalized trust on aspects such as market participation, decentralization or innovation. Highlights a recent work that shows how widespread confidence in the corporate headquarters area can mitigate the negative effects of the crisis. But would this effect change in the case of family businesses? Its competitive advantages are based on its embedding in local networks, a fact demonstrated both for developed economies and for others in development. In good times these advantages can perfectly materialize in better results, but they are incongruent with the mechanisms of trust in impersonal exchanges in the market. During a crisis, areas with greater generalized trust require greater transparency of transactions and an organizational design based on professional management and responsibility for strategic decisions. We would therefore observe an inversion of the beneficial effect of generalized confidence on the results of family businesses during the financial crisis. While non-family businesses would benefit from widespread confidence to mitigate the negative effects of the crisis, high levels of generalized trust would aggravate the effects of the crisis on family businesses.
In order to perform an analysis that would allow causal conclusions, we would like to have an environment where there would be an unexpected crisis, and cultural values would fall “from the sky” on a random distribution of business organization. In our article we propose that there is an excellent laboratory with these characteristics: Italy. First, economists would love to say that the financial crisis was predictable, but it was not the case. Second, cultural values in Italy did not exactly fall from the sky, but they were formed a long time ago due to the different institutional organization of the Italian regions in the Middle Ages and they are very persistent over time; it is then reasonable to assume that its distribution is exogenous to the results of the companies in the present. The map below is revealing of the heterogeneity of the Italian panorama: even within geographical areas with a similar level of development, there are important differences in the generalized confidence. These differences persist when we recreate the map with previous (2005) or later (2012) values to the crisis.
Differences in generalized confidence
The following graph shows our main result, derived from an analysis of 14,742 unique companies (of which 65% are family businesses) followed during 2004-2010. All these companies are relatively large, with revenues of at least 20 million euros in 2007. The financial crisis had a symmetrical effect on Italian companies in the different regions: a drop in operating return on assets (ROA) between 14% and 18%, equivalent to 1.3-1.6 percentage points (from the average of 9%). While generalized confidence does not have differential impacts in good times, our results show that this cultural attribute helps non-family companies to mitigate the negative effect of the crisis by 0.4 percentage points (statistically significant at 90%); but it aggravates the effect of the crisis on family businesses by 0.3 percentage points (significant at 95%). We obtain qualitatively similar effects on income returns or total factor productivity.
Effect of the crisis on the operating performance on assets (ROA)
These economically important effects survive an extensive series of controls at company level, quality of institutions and regional economic development. They also have several levels of heterogeneity. The increase in the negative effect of the crisis on family businesses located in areas with high levels of generalized trust disappears for those who have professionals on the board of directors (compared to those who have boards of directors totally controlled by families). In contrast, the negative effect is exacerbated for family manufacturing companies that are dependent on external financing. This result suggests that, in areas with high generalized confidence, companies controlled by families do not benefit from credit from suppliers, which seems to be the main source of financial resources for other companies in times of crisis.
Our results have several implications. First, we show that the congruence between business organization and cultural values in the area of location of the company’s headquarters is key to understanding the effects of an economic crisis. Second, we emphasize that it is not simply a matter of property rights of the company. It is essential to consider the importance of corporate governance and professional management, factors that recently attracted the attention of economists in their attempts to explain the persistent differences in productivity at company and country level (examples here and here). Third, financial crises tend to be mostly banking. When this traditional channel is blocked, companies turn to alternative sources of financing as their suppliers (see international or US evidence). And when these providers also suffer the effects of the crisis, their resources flow mainly to companies compatible with the local configuration of cultural values.