**3. Methodology**

#### **3.1. Data and variables**

Data for this study were taken from the World Values Survey (WVS) worldwide network of social scientists focused on the study of the changing values. Six waves of the WVS have been published that enquire into the basic values and attitudes of individuals, and thus this database is an excellent proxy for informal institutions. Following Inglehart and Baker [83], who analyzed aggregated nation-level data and carried out three waves of representative national surveys, we used data from the most recent WVS data bases, Wave 5 (2005–2008) and Wave 6 (2010–2012). These databases also contain the greatest number of countries with data in two or more periods of time. Our final sample consists of a balanced panel, with data from 67 observations and 35 countries.

#### *3.1.1. Dependent variable*

This variable was measured with an item in the WVS that represents leadership. This variable collects the degree of self-control and freedom, an important prerequisite for self-leadership [9]. Freedom or autonomy is related to identity and leadership [84] in order to consider that the freedom or the autonomy of the actor is the origin and the destination of their action [85]. This variable is measured by country using a Likert scale (1 = "none at all" to 10 = "a great deal of choice").

#### *3.1.2. Independent variables*

Five independent variables were considered in this study. These variables are in line with the Schwartz dimensions for studying informal institutions. Schwartz [71] used the 'Schwartz Value Inventory' (SVI) for a wide survey of over 60,000 people to identify common values that acted as 'guiding principles for one's life'. Informal institutions were operationalized through tolerance, social capital, creativity, power and responsibility, as follows. *Tolerance*: Percentage of individuals in a country who define tolerance as an important quality. *Creativity*: The respondents were questioned about the importance of coming up with new ideas and being creative, and doing things in one's own way. This variable measures the scale by country using a Likert scale (1 = "not like me" to 10 = "very much like me". *Social capital:* Percentage of respondents who belong to a professional organization by country. *Power:* The respondents were asked about the importance of being rich, having a lot of money and expensive things. This variable measures the scale by country using a Likert scale (1 = "not like me" to 10 means "very much like me"). *Responsibility*: Percentage of individuals who define hard work as an important quality, by country.

#### *3.1.3. Control variables*

restriction on the free flow of information (related to social capital); low value of education and innovation (related to creativity); domination by a restrictive religion, family or clan (related to power) and inability to accept responsibility, and low prestige attached to work (related to responsibility). Leadership development is handicapped by these same national sings [79].

Increasing development and increasing complexity tend to propel societies in the direction of higher income, better education, and more political and economic participation [80], as well as smaller power distances in organizations [24]. These elements of more developed and advanced societies tend to empower subordinates, and thus makes top-down decision making and close supervision in organizations less important and less effective [24, 80]. It has been suggested that some kinds of leadership, such as autocratic, will be seen as less effective and attractive in richer countries [81]. Hofstede [24] consistently tested the effect of economic and social conditions on the structure and functioning of a country's institutions or a country's identity; however, there have been few studies considering the moderating role of contextual factors in

*H6. The level of development of countries will positively moderate the relationship between informal* 

Data for this study were taken from the World Values Survey (WVS) worldwide network of social scientists focused on the study of the changing values. Six waves of the WVS have been published that enquire into the basic values and attitudes of individuals, and thus this database is an excellent proxy for informal institutions. Following Inglehart and Baker [83], who analyzed aggregated nation-level data and carried out three waves of representative national surveys, we used data from the most recent WVS data bases, Wave 5 (2005–2008) and Wave 6 (2010–2012). These databases also contain the greatest number of countries with data in two or more periods of time. Our final sample consists of a balanced panel, with data from 67 observations and 35 countries.

This variable was measured with an item in the WVS that represents leadership. This variable collects the degree of self-control and freedom, an important prerequisite for self-leadership [9]. Freedom or autonomy is related to identity and leadership [84] in order to consider that the freedom or the autonomy of the actor is the origin and the destination of their action [85]. This variable is measured by country using a Likert scale (1 = "none at all" to 10 = "a great deal of choice").

Five independent variables were considered in this study. These variables are in line with the Schwartz dimensions for studying informal institutions. Schwartz [71] used the 'Schwartz Value Inventory' (SVI) for a wide survey of over 60,000 people to identify common values that

leadership [82]. Thus, the following hypothesis is formulated:

*institution and leadership behavior.*

**3. Methodology**

46 Leadership

**3.1. Data and variables**

*3.1.1. Dependent variable*

*3.1.2. Independent variables*

Although we were interested in developing an institutional model, other factors may also influence leadership behaviors. Control variables were included to ensure that the results were not unjustifiably influenced by such factors: education level, the gross domestic product (GDP) at purchasing power parity (PPP), labor force and control of corruption. The data was obtained from the WVS. *Education*: While the level of education and the leadership have been positively associated [86], there are few studies that have used education as a demographic variable in their examination of leadership. Vecchio and Boatwright [87] found that persons with higher levels of education and greater job tenure expressed less preference for leadership structuring (task-oriented behaviors). This control variable was obtained from WVS and was controlled through elementary education. *Gross domestic product (GDP) at purchasing power parity (PPP) per capita* was a measure of the development of countries. Leadership is strongly correlated with wealth and other indices of socioeconomic status [88]. The data source used for the GDP-PPP variable was the International Monetary Fund World Economic Outlook database. *The labor force participation rate* is the proportion of the population aged 15–64 that is economically active: all people who supply labor for the production of goods and services during a specified period. The source of this variable was the International Labour Organization's Key Indicators of the Labour Market Database. *Control of corruption:* This indicator captures perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as the "capture" of the state by the elite and private interests. Values were between −2.5 and 2.5 with higher scores corresponding to better outcomes of institutions [89].

#### **3.2. Statistical procedures**

In this study, given the availability of data, we started with the simplest approach to analyzing panel data, a pooled regression, which omits the dimensions of space and time of the data, calculating an ordinary least squares regression. We therefore propose the following general model:

$$Leadership\_{\mathfrak{a}} = \mathfrak{a} + \mathfrak{P}\_1 \amalg \mathfrak{i}\_{\mathfrak{i}-1} + \mathfrak{P}\_2 \amalg \mathfrak{i}\mathfrak{i}\_{\mathfrak{i}-1} + \mathfrak{e}\_{\mathfrak{i}\mathfrak{i}};\tag{1}$$

where *i* is county and *t* is time; *IIit−1*: matrix of informal institutions in country *i* in year *t; CVit−1*: matrix of the control variable in country *i* in wave *t*. Specifically, we estimated random and fixed-effects models and we used the Hausman specification test [*X*<sup>2</sup> (7) = 30.73, Prob > *X*<sup>2</sup> = 0.0003] in order to verify the choice of the fixed- or random-effects model. The test suggested the use of the fixed-effects specification. We have corrected heteroskedasticity, estimating with feasible generalized least squares (FGLS).
