1
Institutional framework effects on financial
development: differences between developing
and developed countries
Financial development plays a pivotal role in supporting economic growth. However, the channels through which financial development is achieved are not adequately identified in the literature. This paper attempts to contribute to the understanding of how formal and informal institutions, and the interaction effect between both, matter in explaining financial development levels and thereby distinguishes the effects between developing and developed countries. Based on a sample of 77 countries, the Ordinary Least Square regression results find a positive significant effect of formal and informal institutions on financial development. With respect to the interaction effect between them, the results provide no support for a significant effect on financial development and no support for a significant difference in interaction effect between developing and developed countries. This indicates that there is no evidence that the unsatisfying success rates of policy initiatives in developing countries, that are implemented in order to improve financial development, are explained by differences in interaction effect between developing and developed countries.
Field Key Words: Institutional framework, Financial development, Developing countries JEL codes: O11, O19, O43, O57
Joris J. A. Wolbers, s2107481
Supervisor: prof. Dr. B.W. (Robert) Lensink MSc Finance – Faculty of Economics and Business University of Groningen, January 12th, 2017
2 1. Introduction
Over a century ago, as one of the pioneers investigating the causal association between financial development and economic growth, Schumpeter (1911) found that financial development of the banking system played a pivotal role in economic growth by improving efficiency, productivity, and stimulating technical change. The findings of Goldsmith (1969) and Levine (1997) support this notion and provided the theoretical justification for many policy changes in the institutional framework of developing countries in order to improve financial development, which was expected to increase economic growth. Changes in the institutional framework, that acts as the core of an economy and promotes the dissemination and co-ordination of information, should provide access through financial products and services that enable individuals and firms to smooth consumption, manage risk, and invest in education, health or enterprises (Worldbank.org, 2016).
During the last decades, developing countries, that are in general characterized by low financial development levels relative to developed countries, were subject to many policy changes in their institutional framework with the intention to improve their financial development levels. Some examples are the financial liberalization in developing countries in the 1980’s based on the findings of McKinnon (1973) and Shaw (1973) with the intention to stimulate financial development (De Gregorio and Guidotti, 1995), the imposition of capitalist institutions in the former communist countries after the demise of the Sovjet Union, and more recently the export of the Western Democracy by the USA to countries in the Middle East, such as Iraq and Israel (Diamond, 2007). There are competing views on the effectiveness of particular policy changes, imposed by the Western World or supranational institutions such as the World Bank or the United Nations, in order to improve the institutional framework to stimulate financial development. However, there is consensus that the results of most policy changes in developing countries have been disappointing, falling well short of expectations. In most cases, the failure to know how to deal with different, imperfect and underdeveloped institutions in these developing countries, and the fact that policy changes cannot change the way people in a country think, feel, and act leaded to unsatisfying outcomes (Moran, Harris, and Moran, 2007).
3 effect of the institutional framework consisting of formal and informal institutions on financial development a compelling research topic (North, 1993).
The effect of formal institutions on financial development is explained by the Law and Finance Theory (La Porta, Lopez-de-Silanes, Shleifer, and Vishny, 1997, 1998, 2000; Beck, Demirgüç-Kunt and Levine, 2003). This theory consists of two contradicting views regarding the effect of strong formal institutions on financial development (Beck and Levine, 2008). Following the Political Economy view, strong formal institutions have a negative effect on financial development by the disappearance of the regulatory capture effect. Moreover, strong formal institutions give low-grade borrowers access to the credit market. On the other hand, the Public Interest Theory suggests a positive relationship, as formal institutions reduce uncertainty in the outcome of transactions. Furthermore, stronger informal institutions indicate stronger social norms in society, which reduce uncertainty, cost of monitoring, possibility of opportunistic behaviour and help people better understand the benefits and consequences of their actions (North, 1990). Therefore, stronger informal institutions result in higher financial development (Williamson, 2009; Williamson and Kerekes, 2011). On top of that, recent literature about New Institutional Economics discusses the existence of an interaction effect between formal and informal institutions and suggests that when there is ‘the right mix’ between both, there could be a complementary effect that positively influences financial development. However, if both types of institutions are not in ‘the right mix’ these could act as substitutes that counteract each other and thereby negatively influence financial development levels (Boettke, Coyne and Leeson, 2008; Williamson, 2009).
4 following research question: do the effects of formal institutions, informal institutions and
their interaction on financial development differ between developed and developing countries?
The results of this research show significant positive effects of formal and informal institutions on financial development. This suggests that financial development levels of countries that improve their formal and informal institutions will increase. Regarding the interaction effect between formal and informal institutions the results do not provide evidence for a significant interaction effect on financial development. Moreover, there is no significant difference between the interaction effect of developing and developed countries. Therefore, the results provide no evidence that the unsatisfying success rates of policy initiatives in developing countries, that were implemented in order to improve financial development levels, are explained by differences in interaction effect between developing and developed countries.
This research contributes to the current literature in three ways. First, this study uses a data set consisting of relatively many developing countries in comparison with data sets used in comparable studies. This enables to distinguish between the effects of developing and developed countries and aids the investigation of the disappointing results of policy implications in developing countries. Second, besides analysing the effect of formal institutions on financial development, the effect of informal institutions is also investigated, because this exert most influences on day to day lives of people and their interaction with others. Third, this research investigates the interaction between both. According to the literature, the effect of the institutional framework on financial development is not simply adding the effect of formal and informal institutions to each other, though depends on the ‘right mix’ between them. Therefore, the interaction between both is expected to affect the level of financial development; the direction of this effect potentially differs between developing and developed countries.
5 2. Literature review
In the recent literature, various academics have noticed the importance of the institutional framework for financial development and economic growth. Researchers in development economics mentioned that countries should first ‘get the institutions right’ in order to achieve economic growth (Williamson, 1994). This phrase suggests an active role of policy makers for selecting the right institutions and replace ‘wrong’ institutions for ‘good’ institutions. However, as history illustrates, numerous institutional policy changes aiming to improve financial development in developing countries failed and never achieved the intended benefits of the ‘good’ institutions.
The importance of the institutional framework for economic outcomes is confirmed in the paper of North (1990). The author notices that institutional theory must be based on human behaviour, since all institutions are created by humans and institutions exist in a society to reduce transaction costs and information asymmetry by establishing a stable structure for human interaction. The differences in economic outcomes between countries lie in the differences with respect to the rules of the game (institutions) and their interaction with the players of the game (organizations). The current literature concerning the institutional framework is still incomplete and there is no consistent evidence about which institutions matter, how institutions interact with each other, and if institutions matter for developing and developed countries in the same way.
2.1 Formal institutions and financial development
Formal institutions are known as the by the government defined rules and enforced constraints to ensure compliance to these rules (North, 1993). The Law and Finance Theory focuses on the role of formal institutions in explaining cross-country differences in financial development. The theory provides two contrasting views, one positive and one negative, describing the relation between strong formal institutions and high financial development (La Porta, Lopez-de-Silanes, Shleifer, and Vishny, 1997, 1999, 2000; Beck et al., 2003).
6 However, regulatory capture is much more difficult to execute in countries with strong formal institutions where powerful organizations have less influence to adjust the regulatory system for their own benefits. The stronger the formal institutions, the more likely the positive influences of the regulatory capture effect will disappear. The political economy view is in particular relevant for developing economies, because government authorities have more influence in the financial markets and the scope for rent-seeking activities is broad (Khan and Sundaram, 2000).
The political economy view is proven in the paper of Stigler (1971) by concluding that if specific economic groups have enough political power to demand regulation, it could be beneficial for their economic status instead of being subject to regulation that is developed by the political process that defies rational explanation. However, more recent work of Lensink and Meesters (2014) investigates the impact of institutions on bank efficiency and found no support for the political economy view in their research. In addition to the positive influence of weak institutions on regulatory capture, Bianco, Jappelli and Pagano (2005) reveal that stronger formal institutions decrease the cost of financial intermediation (i.e. transaction costs) and thereby improve the access of low-grade borrowers to the credit market as well. This raises the average rate of default and causes a deterioration of the credit quality, resulting in lower financial development. Therefore, as the Political Economy View suggests and Bianco et al. (2005) argue in their paper, stonger formal institutions could be negatively related to financial development in some specific situations.
Second, the positive view is represented by the Public Interest Theory of Pigou (2013). This theory concerning regulation, explains in general terms that regulation is supplied in response to the demand of the public for the protection and benefit of the society as a whole. Regulation and the enforcement of this regulation constrain the behaviour of individuals and enforce certain behaviour by punishment. Strong formal institutions reduce uncertainty by establishing a stable formal structure for human interaction and improve efficient allocation of resources between peoples in society (Hantke-Domas, 2003). This argues for a positive relation between formal institutions and financial development. However, since there are always loopholes in the regulation and the enforcement of these rules, individuals must still be monitored to ensure certain desired behaviour.
7 financial markets and resources are more efficiently distributed amongst them. On the contrary, formal institutions that neglect to do this inhibit financial development. Garretsen et al. (2004) and Allen et al. (2014) also conducted extensive research regarding the relation between formal institutions and financial development and made use the ‘depth’ financial development measure, defined as the ratio of private credit by deposit money banks divided by GDP, to analyse the effect of formal institutions on financial development. Both of them found a significant positive effect for formal institutional development.
As illustrated above, the two theories are not consistent regarding the relation between formal institutional development and financial development. Moreover, several researchers such as Cull and Effron (2008) found no significant effect at all for the effect of formal institutions on financial development. However, this might be due to the fact that the authors used money supply as percentage of Gross Domestic Product (M2/GDP) as proxy for financial development. This does not capture the broad access to bank finance by individuals and banks, the quality of bank services and the efficiency of providing banking services into account, and therefore is not an appropriate determinant to measure financial development. The majority of researchers investigating the relation between formal institutions and financial development found results in line with the Public Interest Theory. This leads to the construction of the following hypothesis;
H1: Stronger formal institutions result in higher financial development levels in both developing and developed countries.
2.2 Informal institutions and financial development
Besides the analysis of formal institutions, many researchers stress the importance of informal institutions in determining the level of financial development. Informal institutions consist of comprising codes of conduct, norms, values, and behavioural conventions in a country; together labelled as culture (North, 1993). These informal institutions influence the way people think, act, and respond to changes in their environment. For example, citizens in a country characterized by high trust in - and respect for - other people are more willing to invest their money, whereas citizens in a country characterized by uncertainty avoidance are reluctant to do this. As a result of the informal institutional influence, the formal institutional literature on financial development might take a too narrow view of the role of institutions, and hence informal institutions are perhaps unduly neglected in many papers.
8 behaviour of individuals in a society and their human interaction. This structure helps individuals to make choices based on the uncertainties they face due to incomplete information about decoding the environment, rules, and procedures (North, 1990). However, at the same time it constrains economic choices. Humans that are subject to particular informal institutions will not engage in behaviour that is in opposition with the dominant informal institutions; potentially leading to sanctions such as expulsion from the community or loss of reputation (Pejovich, 1999). Tabellini (2008), for example, identifies in her paper four distinct categories of culture that influence human behaviour by means of the prohibition of certain human behaviour. As a result, informal institutions fundamentally influence human behaviour while not being directly amenable to regulation.
Besides influencing human behaviour, informal institutions help by enforcing formal institutional if these become internalized by economic actors (North, 1990). Lack of informal institutions will increase asymmetric information regarding the behaviour of others, since the perception of people differs as to how binding formal institutions are. This perception varies from place to place, e.g. the binding of traffic rules in Rome versus that in the Netherlands. Informal institutions have a joint role with formal institutions to reduce transaction costs and facilitate economic exchange (Raiser, 1997). However, where formal institutions focus on the jure institutions, informal institutions make formal institutions effective, produce cooperative outcomes, and help concentrating on de facto institutions by closing the gap between the extensiveness and effectiveness of formal institutions (Casson, Della Giusta and Kambhampati, 2010).
9 Williamson (2009) who uses the culture categories defined by Tabellini (2008) to measure informal institutions and compares the economic development levels from countries with strong informal institutions to countries with weak informal institutions. The results indicate that informal institutions have a significant positive effect on economic development and that their added value is extensive.
In contrast to these findings, Garretsen et al. (2004) found that a country’s informal institutions are not a significant determinant of banking sector development. Although the authors did find a significant effect for stock market development, they noticed that the impact of these countries’ informal institutions is very similar to that of their formal institutions. Therefore, they conclude that the value added of informal institutions seems to be rather small when it comes to understanding the role of institutions at large for financial development. The authors used the four Hofstede (1980) dimensions to describe a country’s culture as a measure for informal institutions and used both stock market development and banking sector development as variables to measure financial development in their paper.
Although the theories in the literature are not consistent on the positive significant effect of informal institutions on financial development, the majority of the literature suggests a positive relationship between informal institutions and financial development. Moreover, most of the conducted empirical research confirms this positive relationship. Therefore, the following hypothesis is developed;
H2: Stronger informal institutions result in higher financial development levels in both developing and developed countries
2.3 Interaction effect of formal and informal institutions
The New Institutional Economics provides the basis for the development of the Interaction Thesis, which identifies the interplay of formal and informal institutions as principal factor affecting financial development. Furthermore, the New Institutional Economics sheds light on the importance of the method of choosing and implementing formal rules, since formal and informal institutions must be considered together in order to provide positive influence on financial development.
10 findings can help countries identify lagging areas where a policy reform is necessary or institutional areas that are not related to financial development at all. On the other hand, areas that performed particularly well could be the result of sound policies from which other countries wish to learn and emulate (Beck, Feyen, Ize, and Moizeszowicz, 2008).
Regarding the interplay between formal and informal institutions, the latter one mainly determine which formal institutions are likely to be accepted as legitimate in a country (Licht et al., 2001). This suggests that for policy makers aiming to improve financial development, it is not sufficient to randomly duplicate successful formal institutions of high financial developed countries, since these might not be in the ‘right’ institutional mix with the present informal institutions. Furthermore, this ‘right’ institutional mix may not be transportable from country to country, due to the dynamics and complexity of the institutional framework, since every country has its own informal institutions that are difficult to change and if these are subject to change, will change very slowly over time (Garretsen et al., 2004; Williamson, 2009). It is well-known that informal institutions have survival ability and will not change immediately in reaction to changes in the formal institutions. As a result, tension could arise between changed or imposed formal rules and the persisting informal constraints. Unsatisfactory outcomes can be the consequence of certain policy changes (North, 1990).
11 The importance of a ‘right mix’ between formal and informal institutions is confirmed by the findings of the papers of Boettke et al. (2008) and Williamson (2009), who find that the success of stronger formal institutions in successfully improving financial development depends on the ability to map into the country’s informal institutions. Formal and informal institutions might complement another if there is a minimum level of informal constraints present. Pejovich (1999) mentions that differences in the interaction effect between countries could help explain cross-country differences in financial development and are contributing to the explanation of the disappointing success rates of policy reforms. Subsequently, something that works well in one country could lead to a conflict between formal and informal institutions in another country. For example, a worldwide banking crisis can force the government, under international pressure, to adapt regulations that restrict the bonuses of bankers. However, dependent on the existing informal rules, this could be in conflict with the national culture of incentive based payments in place. In countries whose culture is characterized by uncertainty avoidance, such regulation might be appreciated and restores confidence in the banking sector, since bonuses give managers tendency for risk taking behaviour (Tosi and Greckhamer, 2004). This indicates that the government should consider the present informal institution when imposing or changing formal institutions. Nevertheless, governments often choose to formalize in an inefficient and suboptimal way that is not in the interest of the public, though benefits the state (McChesney, 1990). This is especially true if the controlling elite of a country tries to centralize political power in order to embark on economic planning for growth by doing so ignores the traditional values in favour of formal rules; a common phenomenon in developing countries. In these countries, new formal rules are in direct opposition to the prevailing informal institutions and ended up replacing the old ethos. This results in political corruption, social instability, and economic failures (Bauer, 1988). These findings in combination with the Interaction Thesis lead to the construction of the following hypotheses:
H3: The interaction effect between formal and informal institutions on financial development is positive for developed countries. This results in an increase in the marginal effect of stronger formal institutions with respect to higher levels of informal institutions.
12 3. Data
Based on the data availability of countries’ informal institutional values in the World and European Value Survey, 82 countries were selected. However, for five of these 82 countries there is no data available on their formal institutions in the World Governance Indicators database. Therefore, the sample size of the final data set used in this research consists of 77 countries. The database distinguishes between countries that are classified as ‘developing countries’ and ‘developed countries’ according to the World Economic Situation and Prospects (WESP) for 2014 of the United Nations (United Nations, 2016), resulting in 31 developing and 46 developed countries. This distinction is illustrated in Appendix 1.
Because of limited longitudinal data availability, for example our measure of informal institutions is available at only one point in time, this research focusses on performing cross-sectional data analyses only. However, this is not considered to be problematic as no apparent temporary outliers in the cultural dimension scores are prevalent in the cultural dimension scores since informal institutions only change very slow over time (Gleaser, La Porta, Lopez-de-Silanes and Shleifer, 2004; Acemoglu, Johnson, and Robinson 2005). The observations of financial development, the formal institutional indicators of financial development, and the control variables are averaged over multiple years (2010-2014). This is customary in the financial development literature to reduce the influence of outliers (Allen et al. 2014).
Putting this together, the data set is characterized as a balanced panel and will provide one observation per country for all the variables in our regression analysis. Table 1, represents the most important descriptive statistics of the data set. Developed countries are, as expected, characterized by a on average higher financial development level for the ‘depth’, ‘size’, and ‘broad access’ proxies, higher formal institutions and higher informal institutions in comparison to developing countries. Furthermore, the variable ‘population density’ has a high standard deviation caused by several outliers in the data. However, this is not expected to cause any problems in determining countries’ financial development levels.
13
Table 1: Descriptive statistics of the dependent and independent variables for the two developed and developing subsamples and the total sample.
Developed countries Developing countries Total sample Total sample Total sample
Financial development ‘depth’ 90.19% 44.25% 71.70% 58.09% 49.02 Financial development ‘size’ 97.68% 64.75% 84.53% 67.35% 63.43 Financial development ‘broad access’ 81.63% 40.55% 65.09% 72.65% 29.31
Formal institutional index 7.01 4.17 5.76 5.31 2.64
Informal institutional index 5.62 2.55 3.82 3.27 2.25
Respect for other people 71.02% 64.41% 68.27% 67.00% 12.03
People can be trusted 32.80% 17.22% 26.32% 22.00% 17.44
Control of your life 67.22% 71.94% 69.18% 69.00% 7.01
Obedience is important 27.40% 46.50% 35.34% 34.00% 16.64 Population size 3.32 4.83 3.95 2.35 1.72 Population density 260.54 394.14 316.06 87.23 1138.11 Geography 0.25 0.13 0.20 0.22 0.09 Government consumption 18.92% 14.82% 17.22% 18.07% 4.39 Inflation 3.21% 5.27% 4.07% 2.84% 4.13 Educational enrolment 95.72% 91.19% 93.84% 95.46% 6.02 Number of countries 46 31 77 77 77
Note. Columns 1, 2, and 3 provide averages for the subsamples and the total sample, column 4
provides the median of the total sample, and column 5 provides the standard deviation of the total sample. The data represents the countries’ values for the variables in the period 2010-2014.
14 To investigate the suggested correlation between formal and informal institutions, table 2 provides a correlation matrix of all the independent variables. What stands out by analysing the correlation coefficients is that the formal and informal institutional indexes are indeed highly correlated (0.71). This could be due to the fact that both the formal and informal institutional indicators measure the same underlying principles that affect financial development. The high correlation between formal and informal institutions could lead to multicollinearity problems. To deal with these problems a principal component analysis (PCA) is conducted and will be explained in Section 4. On top of that, there is a high negative correlation between several variables, for example inflation and formal and informal institutions. This means that between these variables there is an inverse relationship - when one variable decreases, the other increases – and vice versa.
Table 2: Correlation table between the independent variables of the Ordinairy Least Square regression analysis model
1 2 3 4 5 6 7 8 Informal institutions 1 Formal institutions 0.71 1 Population size -0.15 -0.28 1 Population density 0.07 0.21 -0.09 1 Geography 0.46 0.34 -0.30 -0.26 1 Government consumption 0.43 0.45 -0.24 -0.29 0.57 1 Inflation -0.35 -0.51 0.18 -0.03 -0.10 -0.33 1 Educational enrolment 0.36 0.47 -0.17 -0.08 0.31 0.46 -0.34 1
3.1 Financial development measures
15 banking sector development is used in the paper of Garretsen et al. (2004) and Allen et al. (2014) to measure financial development. On top of that, this variable is associated with long-term economic growth (Levine, Loayza and Beck, 2000; Levine, 2005).
A limitation for the ‘depth’ measurement is that this ratio does not control for the ‘size’ of the banking sector by for instance taking the total liquid liabilities into account. Besides, it does not capture the ‘broad access’ to bank finance by individuals. To include these two characteristics of banking sector development, two other proxies of financial development are included as well. The second proxy to measure financial development is the ‘size’ of the banking sector, measured by the ratio of liquid liabilities in the banking system divided by GDP. The third proxy is the ‘broad access’ to bank finance by individuals, measured by the % of people with an account at a formal institution. The World Bank Financial Development and Structure Dataset constructed according to the instructions of Beck, Demirgüç-Kunt and Levine (2009) and available at Data.worldbank.org is used to determine the financial development levels for the three proxies of the countries from the data set.
3.2 Formal institutional measures
16 component model these data sources are rescaled and combined to create the six indicators. A key feature of the methodology is that it generates margins of error for each governance estimate. These margins of error need to be taken into account when making comparisons over time and across countries (Info.worldbank.org, 2016). The data reflect the views on governance of survey respondents and public, private, and NGO sector experts worldwide and are scaled per indicator in a range of -2.5 – 2.5, where a higher score indicates stronger formal institutions. In their paper, the authors explicitly report margins of error accompanying each country estimate. These reflect the inherent difficulties in measuring governance using any kind of data. However, they conclude that even after taking margins of error into account, the WGI permits meaningful cross-country comparisons. The aggregate indicators, together with the disaggregated underlying source data, are obtained from their website (Info.worldbank.org, 2016). The six WGI are briefly explained in Table 3.
17
Table 3: Description of the six Worldwide Governance Indicator (WGI)
WGI Description of indicator
1. Voice and Accountability Captures perceptions of the extent to which country's citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media.
2. Political Stability and Absence of Violence
Captures perceptions of the likelihood of political instability and/or politically-motivated violence including terrorism. 3. Government Effectiveness Captures perceptions of the quality of public services, the
quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies.
4. Regulatory Quality Captures perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development.
5. Rule of Law Captures perceptions of the extent to which agents have confidence in and abide by the rules of society. In particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. 6. Control of Corruption 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 ‘capture’ of the state by elites and private interests.
18
Table 4: Correlation table of the six World Governance Indicators (WGI)
1 2 3 4 5 6
Voice and accountability 1
Political stability 0.81 1
Government effectiveness 0.84 0.84 1
Regulatory Quality 0.84 0.82 0.95 1
Rule of Law 0.87 0.85 0.97 0.94 1
Control of corruption 0.82 0.82 0.95 0.90 0.96 1
Figure 2: Scatterplot of the association between the formal institutions constructed of the six WGI and financial development measured by the ‘depth’ of banking sector development
3.3 Informal institutional measures
To quantify informal institutions in a reliable way, the proxy needs to be persistent over time and has to show depth and durability. Therefore, a previously established measure of culture developed by Tabellini (2008) and expanded by Tabellini (2010) and Williamson and Kerekes (2011) is relied on to proxy for informal institutions. The reason to use this proxy for informal institutions is that it represents a broad measure of national culture with respect to economic behaviour and social interaction. The latter fits the research objective as the aim is to analyse norms and conducts in society that are specific to financial markets to analyse effects on financial development. Other proxies for informal institutions such as the cultural values of Hofstede and Schwartz emphasize general dimensions of national culture and do not focus on norms that are specific to financial markets. Moreover, the availability of data on developing countries is rather limited for other proxies of informal institutions (e.g. the countries’ cultural dimensions of Hofstede (1980) to measure the informal institutions) and
19 thereby putting heavy restrictions on the number of developing countries that could be analysed.
The measure of culture is constructed by identifying four key traits that are relevant for economic interaction and transactions in a society – in other words, economic culture. Economic culture is defined by Porter (2000) as ‘the beliefs, attitudes, and values that bear on economic activities of individuals, organizations, and other institutions.’ The concept of culture is narrowed in order to focus on how economic cultural traits can effect financial development and therefore the paper relies on the four distinct categories of culture identified by Tabellini (2008). The four components are: trust, respect, individual self-determination (called control), and obedience. These components serve as rules governing financial interaction between individuals in society. In general, trust, respect and individual self-determination are thought to promote social and financial interaction, whereas obedience decreases the urge for financial interaction between individuals. To maximize our sample size, the most recent values from two value surveys, the World Value Survey (Worldvaluessurvey.org, 2016) and the European Value Survey (Atlasofeuropeanvalues.eu, 2016), are used to quantify each component. These surveys capture individual beliefs and values that reflect social norms and customs. Each section of culture has a corresponding question in the survey that is discussed in more detail below.
The first component, trust, is argued to reduce transaction cost resulting in more efficient outcomes more quickly. Therefore, Knack and Keefer (1995) argued that higher trust societies will experience higher levels of economic growth. Trust reduces the cost of monitoring and lowers transaction costs; thus promoting mutual trust in individual leads to a quicker and more efficient allocation of products and services (Fukuyama, 1996; La Porta et al. 1997; Woolcock, 1998; Zak and Knack, 2001; Avinash, 2004; Francois and Zabojnik, 2005). A lack of trust between individuals increases the cost of monitoring and uncertainty in the behaviour of others, resulting in sand in the wheels of allocation processes that lead to inefficient use of resources. The component trust is measured with the following question from the survey: “Generally speaking, would you say that most people can be trusted or that you can’t be too careful in dealing with people?” The level of trust is measured in each country by counting the number of respondents that answered ‘Most people can be trusted’ as opposed to ‘Can’t be too careful’ and ‘Don’t know’.
20 (Platteau, 2000). Respecting others also reduces uncertainty in transactions and enhances social and economic interaction. The following survey question is analysed to determine the importance of respect in a society: “Here is a list of qualities that children can be encouraged to learn at home. Which, if any, do you consider to be especially important? Please choose up to five.” Respect is defined as the percentage of respondents in each country who mentioned that the quality “tolerance and respect for other people” is important.
The third cultural trait is defined as control and captures how determined individuals are in their efforts to succeed. Individual motivation depends on the level of self-control that individuals believe they have over their choices and reap the benefits or consequences of their actions. If individuals believe they can make their own choices, the more likely these individuals are willing to work harder, invest in the future, and engage in entrepreneurial activities (Banfield, 1958). When individuals believe that they have control over their lives, they will be more likely to put effort in order to improve their economic welfare. However, if individuals view the likelihood of succeeding as a product of luck or political connections, they will not put this effort. To identify and capture control, the following survey question is used: “Some people feel they have completely free choice and control over their lives, while other people feel that what we do has no real effect on what happens to them. Please use this scale (from 1 to 10) where 1 means ‘none at all’ and 10 means ‘a great deal’ to indicate how much freedom of choice and control in life you have over the way your life turns out.” An aggregate control component is determined by averaging all the individual responses and multiplying by 10.
21 By combining all four components, one comprehensive measure for culture for each country is achieved by summing trust, control, and respect and then subtracting the obedience score. Thereafter, this comprehensive variable is converted to a relative scale ranging from 0 to 10, with 0 representing the country with the culture least supportive for financial development and 10 representing the country with the culture most conducive to financial development. This aggregation is relevant, because the aim is to investigate the effect of informal institutions in general on financial development and the goal is not to argue that a particular cultural dimension is more relevant than another one. However, since the correlation between the four components is low, as illustrated in table 5 below, it is also important to look at the effect of each component separate. Figure 3 provides a scatterplot of the comprehensive measure for culture and financial development suggesting a positive association between informal institutions and financial development.
Table 5: Correlation table of informal institutional indicators
1 2 3 4
Respect 1
Trust 0.37 1
Control 0.25 -0.05 1
Obedience -0.01 -0.47 0.22 1
22 3.4 Control variables
In determining which control variables for financial development to include, previous literature is followed that analyses determinants of financial development and in particular those who take developing countries into account (Acemoglu, Johnson and Robinson, 2001, 2002; Glaeser et al., 2004; Garresten et al., 2004; Acemoglu and Johnson, 2005; Williamson, 2009; Tabellini, 2010; Allen et al., 2014). Exogenous variables are included that describe the environment in term of latitude, population size and population density. Subsequently, macro-economic variables such as consumer price inflation and other country characteristics as for instance the primary school enrolment of the country’s citizens are included that could influence the level of financial development. The variables are explained in detail below and the measurement and data sources are described in Appendix 3.
Population: large populations should spur financial development due to scale and
networking effects that make the provision of financial services more efficient in larger economies (Allen et al., 2014).
Population density: high population density should have a positive influence on financial
development, because it is easier for financial institutions to accumulate savings when a high number of potential depositors have easy access to them (Allen et al., 2014).
Geography: include (the scaled absolute value of) latitude, because temperature zones
further away from the equator have more productive agriculture and healthier climates, which has enabled them to develop their economy and institutions as well (Landes, 1998).
Consumer price inflation: high inflation should decrease financial development because it
makes loan contracting over extended periods more difficult (Allen et al., 2014).
Government consumption: high government consumption should improve financial
development since it provides services and products that are accessible for everyone. Access to more services and products will increase the life standards of citizens and thereby enables their access to financial products and services (Williamson, 2009).
Primary school enrolment: higher primary enrolment provides people with the basic of
reading, writing and mathematical skills, which has a positive effect on financial management and is thereby likely to have an indirect positive impact on financial development1 (Allen et al., 2014).
1In this paper the causality direction described in the paper of Williamson (2009) and Allen et al. (2014) is
23 4. Methodology
As Section 2 has described, the exact impact of the institutional framework on financial development remains unclear and the theory is inconsistent and still incomplete. A cross country regression analysis is used in this study to examine the effect of the formal and informal institutional framework on the level of financial development. On top of that, an interaction term is used to discover the interaction effect of formal and informal institutions on financial development and the results are controlled for several variables. The latter helps to identify what the effect of formal and informal institutions is on financial development. A dummy variable is used to investigate if this effect differs between developing and developed countries. In line with hypothesis 1 and 2 that are developed in the literature overview, a positive effect of stronger formal and informal institutions on financial development in both developing and developed countries is expected.
To measure the interplay between the formal and informal institutions an interaction term is included in the regression. A positive coefficient implies that formal institutions are built off informal rules and are codifying pre-existing practices. This supports the idea that formal and informal institutions are complementary (North, 1990). A negative coefficient suggests a substitution effect between formal and informal constraints and indicates a mismatch between institutions. Another interesting implication regarding the interaction of formal and informal institutions from the prior literature is that formal and informal institutions do not necessarily interact in every country in the same way. More specific, the interplay between these institutions could be complementary in some countries and be substitutes in others. To test for this difference in interaction effect, a developing country dummy variable is included in the interaction term. Following the literature, the interaction effect is suggested to be negative for developing countries and positive for developed countries.
The relationship between the institutional framework and financial development is investigated by employing univariate and bivariate Ordinary Least Squares (OLS) regressions. The first and second OLS regressions are univariate and are identified as,
1) 𝑌𝑖 = 𝛼 + 𝛽1𝐹𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑠 + 𝜀𝑖, 2) 𝑌𝑖 = 𝛼 + 𝛽2𝐼𝑛𝑓𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑠 + 𝜀𝑖,
24 Build on these initial results, additional control variables are included to test the first two hypotheses. The empirical model is estimated as:
4) 𝑌𝑖 = 𝛼 + 𝛽1𝐹𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑠 + 𝛽2𝐼𝑛𝑓𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑠
+ 𝛽3𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑠𝑖𝑧𝑒 + 𝛽4𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 + 𝛽5𝐺𝑒𝑜𝑔𝑟𝑎𝑝ℎ𝑦 + 𝛽6𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛 + 𝛽7𝐺𝑜𝑣𝑒𝑟𝑛𝑚𝑒𝑛𝑡 𝑐𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛
+ 𝛽8 𝑃𝑟𝑖𝑚𝑎𝑟𝑦 𝑠𝑐ℎ𝑜𝑜𝑙 𝑒𝑛𝑟𝑜𝑙𝑚𝑒𝑛𝑡 + 𝜀𝑖.
To test the significance of the interaction effect between formal and informal institutions, an interaction term is included. Moreover, another interaction term together with a dummy variable for developing countries is included in the regression analysis to distinguish the effect of the interaction term between developing and developed countries. This leads to the following regression, 5) 𝑌𝑖 = 𝛼 + 𝛽1𝐹𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑠 + 𝛽2𝐼𝑛𝑓𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑠 + 𝛽3𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑠𝑖𝑧𝑒 + 𝛽4𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 + 𝛽5𝐺𝑒𝑜𝑔𝑟𝑎𝑝ℎ𝑦 + 𝛽6𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛 + 𝛽7𝐺𝑜𝑣𝑒𝑟𝑛𝑚𝑒𝑛𝑡 𝑐𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 + 𝛽8 𝑃𝑟𝑖𝑚𝑎𝑟𝑦 𝑠𝑐ℎ𝑜𝑜𝑙 𝑒𝑛𝑟𝑜𝑙𝑚𝑒𝑛𝑡 + 𝛽9𝐹𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑎𝑙 𝑖𝑛𝑑𝑒𝑥 ∗ 𝐼𝑛𝑓𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑎𝑙 𝑖𝑛𝑑𝑒𝑥 + 𝛽10𝐷1 + 𝛽11𝐹𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑎𝑙 𝑖𝑛𝑑𝑒𝑥 ∗ 𝐼𝑛𝑓𝑜𝑟𝑚𝑎𝑙 𝑖𝑛𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑎𝑙 𝑖𝑛𝑑𝑒𝑥 ∗ (1 − 𝐷1) + 𝜀𝑖 .
Where 𝐷1 is a dummy variable that is [1] if the country is a developing country and [0] if the country is a developed country.
25 5. Results
The results of the univariate and bivariate regression analyses are presented in Table 6. Columns 1 and 2 present the results of the first two univariate regressions. Both present a strong positive significant effect of formal (p<0.01) and informal (p<0.01) institutions on the ‘depth’ proxy of financial development. Column 3 presents the results of the bivariate regression that analyses formal and informal institutions together. Both relate positive and significant to financial development. Moreover, the adjusted R-square of the bivariate regression increases to 0.48. Column 4 substantiates these results by including the control variables and consequently the adjusted R-square increases further to 0.49. The positive significant effect of formal and informal institutions on financial development confirms hypothesis 1 and 2 and is in line with the findings of Licht et al. (2007) and Williamson (2009). With respect to the control variables, ‘population density’ has a positive significant effect (p<0.10) on financial development. This effect corresponds to the expectations based on the literature. The coefficients of the other control variables are insignificant.
The low correlation between the informal institutional indicators as illustrated in Table 3, indicate that all four indicators measure different aspects of informal institutions. Consequently, it is important to analyse each indicator separately since their effects on financial development could differ. In column 5, the separate indicators are analysed and have, with the exception of obedience (p<0.10), no significant effect on financial development. Moreover, the explanatory power of the model, the adjusted R-square, is reduced to 0.46. These findings are in contrast to the strong significant effect of the informal institutional index on financial development. The latter indicates that the informal institutions only have a significant effect on financial development if the indicators are analysed together instead of separate of each other. Since this research focusses on the effect of informal institutions in general on financial development instead of arguing that a particular cultural dimension is more relevant than another one, the remaining part of this research focusses on the informal institutional index instead of separate informal institutional indicators.
26
Table 6: Ordinary least square regression of the effect of formal and informal institutions, and the interaction between both, on the ‘depth’ proxy of financial development
1 2 3 4 5 6
Formal institutions 12.513*** 9.488*** 6.876*** 8.668** 12.071** (1.502) (2.105) (2.575) (2.767) (5.556)
Informal institutions 13.424*** 5.159* 6.085** 16.598*
(2.012) (2.566) (2.747) (8.642)
Respect for others 0.036
(0.468)
People can be trusted 0.064
(0.360)
Control of your life 0.314
(0.694) Obedience -0.582* (0.341) Formal institutions * Informal institutions -2.044 (2.139) Formal institutions * Informal institutions * (1-developing dummy) 0.676 (1.294) Developing dummy -5.081 (22.669) Population size 0.542 0.993 0.687 (2.575) (2.801) (2.611) Population density 0.008* 0.007 0.007 (0.004) (0.005) (0.004) Geography -31.842 -38.752 -92.093 (58.072) (69.421) (69.586) Government Consumption 0.759 1.093 1.090 (1.302) (1.407) (1.321) Inflation -1.237 -0.993 -0.727 (1.171) (1.261) (1.203) Educational enrolment 0.337 0.470 0.157 (0.807) (0.856) (0.814) Constant 2.409 20.422** -0.542 -26.682 -32.601 -33.604 (9.267) (8.904) (9.203) (72.060) (96.409) (83.969) #Observations 77 77 77 77 77 77 Adjusted R-square 0.47 0.36 0.48 0.49 0.46 0.50
Note. SEs are shown in parentheses.
27 The results of the OLS regression for the ‘size’ and ‘broad access’, proxies of financial development, are presented in Table 7. Columns 1 and 2 use ‘size’ of the banking sector as proxy for financial development. In column 1, the formal institutions have a significant positive effect on financial development yet the other variables are insignificant. When the interaction term is included in column 2, the formal institutions become even more significant and the informal institutions have a positive significant effect as well. Moreover, the interaction effect is significant negative for developing countries and significant positive for developed countries, providing support for hypothesis 3 and 4.
The negative interaction effect on financial development implies that in developing countries formal and informal institutions act as substitutes of each other. The governments of developing countries often duplicate or implement new formal institutions, without considering the present informal institutions. This causes friction between formal and informal institutions resulting in a decrease in the marginal effect of stronger formal institutions with respect to higher levels of informal institutions. It is important that governments of developing countries are aware of the fact that there exists no one-size-fits-all institutional framework that works for every country. Imposing regulations and enforcement constraints that worked well in developed countries are definitely no guarantee to enhance financial developed with the goal to stimulate economic growth. The idea that successful institutional arrangements are identifiable and transportable to other countries might not be true, and should at least be taken with serious caution. The positive interaction effect for developed countries suggests that these countries do consider the existing informal institutions when implementing new formal institutions or make changes in the existing ones. This effort pays off in a positive interaction between them and creates a situation whereby formal and informal institutions act as complements. As such, the analysis suggests that to improve financial development, finding the ‘right mix’ between formal and informal institutions is probably even more important than previously believed.
28
Table 7: Ordinary least square regression results for formal and informal institutions, and the interaction between both, for the ‘size’ and ‘broad access’ proxies of financial development.
1 2 3 4 Formal institutions 12.203* 27.654*** 8.907*** 8.560*** (6.439) (8.271) (1.579) (2.144) Informal institutions 14.754 37.889*** 4.573* 4.052 (10.311) (12.864) (2.528) (3.335) Formal institutions * informal institutions -1.639 -9.889*** -0.549* -0.364 (1.226) (3.185) (0.301) (0.826) Formal institutions* Informal institutions * (1 - Developing dummy) 5.362*** 0.121 (1.926) (0.499) Developing dummy -8.812 66.923** -22.002*** -23.705*** (20.972) (33.746) (5.141) (8.748) Population size 0.483 -0.432 1.416 1.437 (4.066) (3.886) (0.997) (1.007) Population density 0.016** 0.012*** 0.002 0.002 (0.007) (0.007) (0.002) (0.002) Geography -18.251 -6.147 -47.218* -47.490* (108.665) (103.59) (26.639) (26.855) Government Consumption -1.787 -2.206 1.155** 1.165** (2.058) (1.966) (0.505) (0.507) Inflation -1.893 -1.431 0.688 0.677 (1.872) (1.790) (0.459) (0.464) Educational enrolment 0.378 0.600 -0.078 -0.083 (1.270) (1.212) (0.311) (0.314) Constant 6.402 -121.199 11.066 13.935 (122.104) (125.002) (29.934) (32.406) #Observations 77 77 77 77 Adjusted R-square 0.25 0.32 0.79 0.79
Note. SEs are shown in parentheses.
* Significant at the 10% level. ** Significant at the 5% level. *** Significant at the 1% level.
The coefficients of columns 1 and 2 present the effect on the ‘size’ proxy for financial development. The coefficients of columns 3 and 4 present the effect on the ‘broad access’ proxy for financial development.
29 4, the effect for formal institutions remains positive significant whereas the effect of the informal institutions becomes insignificant. The increase in standard error of informal institutions that made the coefficient insignificant might be the result of multicollinearity problems that will be analysed hereafter. Regarding the control variables, the results provide a positive significant effect of government consumption on ‘broad access’ to the banking sector as measure of financial development. This is in line with the literature stating that government consumption enables access to financial products and services for everyone. Interesting to ascertain is the high adjusted square (0.79) of column 3 and 4 in comparison to the R-squares for the ‘depth’ and in particular to the ‘size’ of the banking sector models. This indicates that the variables in the OLS regression model better explain the variance in ‘broad access’ to the banking sector compared to the other two proxies of financial development.
However, the results of table 6 and 7 should be interpreted with caution, due to the high correlation between formal and informal institutions and the presence of two interaction terms in de regression model. Interaction terms include the main effects terms and therefore highly correlate with them. The high correlation between formal and informal institutions in addition to the high correlation of the interaction terms could lead to multicollinearity problems in the regression analysis of table 6 and 7. Severe multicollinearity is a problem because it can increase the standard error of the coefficient estimates and make the estimates very sensitive to minor changes in the model. The result is that coefficient estimates become unstable and difficult to interpret. Multicollinearity saps the statistical power of the analysis, can cause the coefficients to switch signs, and makes it more difficult to specify the correct model. Several signs indicating the presence of multicollinearity are displayed in the table 6 and 7. Examples are the increase in standard error of the formal and informal institutions variables when comparing column 4 to column 6 of table 6, and the sign and significance changes in the comparison between column 1 and 2 of table 7.
30 erroneous results could occur that lead to misinterpretations. A centralized VIF value of 5 or greater indicates a reason to be concerned about multicollinearity, because it illustrates that the coefficients are poorly estimated and their p-values should be interpreted with caution. The results of the VIF test are presented in table 8 below.
Table 8: VIFs for all the independent variables in the OLS regression analysis. The VIF of the formal and informal institutions, and for the interaction terms indicate multicollinearity problems (>5) when the interaction terms are included in the OLS regression model.
Centralized VIF Centralized VIF
Formal institutions 3.051 8.155 Informal institutions 2.337 15.309 Population size 1.200 1.211 Population density 1.383 1.419 Geography 1.817 2.341 Government consumption 2.001 2.070 Inflation 1.431 1.520 School enrolment 1.445 1.461
Formal institutions * Informal institutions 70.881
Developing dummy 11.931
Formal institutions * Informal institutions * (1- Developing dummy)
27.677
The results indicate that the VIF for formal, informal institutions, and for the interaction terms are larger than 5 only if the interaction terms are included in the regression analysis. This suggests that multicollinearity only has a severe effect on the results of the regression analyses that include the interaction terms. Consequently, the coefficients of the variables that are subject to multicollinearity are poorly estimated and should be interpret with caution.
31 much variance as the underlying variable. Therefore, it is in line with the original intention of the analysis to reduce the variables down to ‘more substantive’ underlying factors to select principal components with an eigenvalue larger than 1 (Nunnally and Bernstein, 1994).
Table 9 shows that the largest principal component, PC1, has an eigenvalue of 6.295. PC1 could be regarded as a factor that represents the formal institutional indicators, as illustrated in table 10. The formal indicators that are associated with PC1 create certainty in the outcomes of transactions by constraining undesirable behaviour and enforcing these constraints. Therefore, PC1 is expected to be positive related to financial development. The second largest principal component, PC2, has an eigenvalue of 1.364 and is associated with several informal indicators. As a result of the low correlation between the various informal institutions, these indicators could not be combined into on principal component. Consequently, PC2 is in specific representative for the informal institutional indicator ‘control of your life’ and is not representative for all informal institutional indicators. Other principal components that have less explanatory power will represent the remaining informal institutional indicators. PC3 for example, could be seen as the representative of the informal institutional indicator, ‘people can be trusted’, as table 10 illustrates.However, the eigenvalue of this principal component is smaller than one and therefore will not be taken into account in the further analysis.
PC1 and PC2 together explain 77% of the total variance in the institutional framework. Their eigenvector loadings are used to construct PC1 and PC2 values for each country in the sample. Thereafter, PC1 and PC2 replace the formal and informal institutional variables in the OLS regression analysis to explain the financial development levels of countries.
Table 9: Presents the eigenvalues of the PCA of all formal and informal institutional indicators. The first two components constructed of all institutional indicators have an eigenvalue >1.
Principal components
Eigenvalue Difference Proportion Cumulative
32
Table 10: Presents the eigenvector loadings with the institutional indicators of PC 1 and PC2 with an eigenvalue >1, and PC3.
Institutional indicators PC 1 PC 2 PC3
Voice and accountability 0.35 0.08 0.33
Regulatory quality 0.37 -0.00 0.17
Rule of law 0.39 0.00 0.10
Political stability and absence of violence 0.35 -0.11 0.22
Government effectiveness 0.39 0.02 -0.01
Control of corruption 0.38 0.07 0.10
Control of your life 0.00 0.69 -0.24
Obedience is not important2 0.22 -0.49 -0.41
People can be trusted 0.27 -0.10 -0.72
Respect for other people 0.21 0.50 -0.22
Columns 1 and 2 of Table 11 present the results for ‘depth’ proxy, columns 3 and 4 for the ‘size’ proxy and columns 5 and 6 for the ‘broad access’ proxy of financial development3
. PC1, representing the formal institutions, depicts as expected a positive and significant effect on financial development for all three proxies. The effect of PC2 on financial development is insignificant, except for the ‘size’ proxy of financial development when the interaction terms are not included (P<0.10). Notable is the negative coefficient of PC2, since this is in contrast to the expectations based on the literature that predicts a positive relationship between ‘control of your life’ and financial development. This negative coefficient could be explained by the behaviour of people that believe they have control over their life. In specific, these people might be willing to behave unethical if this increases their chance of success, notwithstanding their behaviour could damage others. Unethical behaviour leads to more uncertainty in transactions between people because, people’s behaviour is less predictable if they do not take behavioural constraints into account in their efforts to achieve success. Increases in uncertainty during transaction could lead to lower financial development due to not taking place of the transaction at all.
2 This variable is defined contrary to the description in section 3.3 to make the interpretation of the results more
straightforward. Therefore, ‘Obedience is not important’ is expected to be positively associated with financial development.
3 Appendix 4 illustrates that the results of table 11 are not subject to multicollinearity problems, since all VIF are
33
Table 11: Ordinary least square regression results of PC 1 and 2, and the interaction between both, on the ’depth’, ‘size’, and ‘broad access’ proxies of financial development.
1 2 3 4 5 6 PC1 11.638*** 8.848*** 10.272*** 8.593** 8.813*** 7.476*** (2.442) (2.877) (3.603) (4.199) (1.060) (1.108) PC2 -2.346 3.629 -11.370* -2.615 -0.552 1.119 (4.198) (5.785) (6.194) (8.442) (1.822) (2.228) PC 1 * PC 2 4.845 4.842 -0.532 (3.468) (5.061) (1.336) PC 1 * PC 2 * (1 – Developing dummy) -4.994 -14.019 -0.807 (5.007) (17.307) (1.928) Developing dummy -23.637 -19.961 -24.656*** (15.471) (22.577) (5.958) Population size 1.322 2.090 -0.388 -1.167 0.980 1.679 (2.664) (2.729) (3.931) (3.983) (1.156) (1.051) Population density 0.007 0.008* 0.013* 0.013* 0.002 0.002 (0.005) (0.005) (0.007) (0.007) (0.002) (0.002) Geography -32.29 -85.444 -56.237 -85.047 11.110 -39.579 (67.511) (75.038) (99.610) (109.502) (29.293) (28.896) Government Consumption 0.848 0.919 -2.127 -1.987 0.994* 0.986 (1.340) (1.338) (1.977) (1.952) (0.581) (0.515) Inflation -1.375 -1.447 -2.324 -1.175 0.142 0.520 (1.187) (1.259) (1.751) (1.837) (0.515) (0.485) Educational enrolment 0.254 0.024 0.505 0.396 0.137 -0.033 (0.829) (0.838) (1.224) (1.222) (0.360) (0.323) Constant 39.739 79.936 91.627 115.493 28.902 62.261* (77.967) (81.285) (115.037) (118.617) (33.830) (31.301) #Observations 77 77 77 77 77 77 Adjusted R-square 0.47 0.47 0.29 0.31 0.71 0.78
Note. SEs are shown in parentheses
* Significant at the 10% level. ** Significant at the 5% level. *** Significant at the 1% level.
The coefficients of columns 1 and 2 present the effect on the ‘depth’ proxy of financial development. The coefficients of columns 3 and 4 present the effect on the ‘size’ proxy of financial development. The coefficients of columns 5 and 6 present the effect on the ‘broad access’ proxy of financial development.
34 informal institutions. In specific, the PCA methodology removes the clear-cut distinction between formal and informal institutions when determining the interaction effect between PC1 and PC2 on financial development. Nevertheless, PC1 is a good approximation of the formal indicators and PC2 of one informal institutional indicator and this makes principal components the best approximation to provide an answer on the research question. Consequently, the coefficients of the interaction terms do not provide significant evidence to confirm hypotheses 3 and 4.
The dummy variable is only significant negative for the ‘broad access’ proxy of financial development. This indicates that developing countries have a significant lower intercept compared to developed countries regarding ‘broad access’ to bank finance. Moreover, the control variables government consumption and population density are positive significant. This is in line with the current literature and indicates that higher government consumption increases the access to bank finance for individuals in a country and higher population density improves the ‘depth’ and ‘size’ of banking sector development.
6. Robustness
