global financial crisis, economic freedom, and the demand for … · 2015. 6. 25. · updated data...

1

Global financial crisis, economic freedom, and the demand for

non-life insurance: An empirical investigation

Tam Trinh, Pasquale Sgro and Xuan Nguyen*

ABSTRACT

This paper investigates empirically the determinants of the demand for non-life

insurance in developed and developing countries before and during the global financial

crisis (GFC) based on a unique panel dataset covering 36 developed and 31

developing countries over the period 2000-2011. Results of our instrumental variable

analysis indicate that economic freedom, income, bank development, culture, and law

system are the key drivers of the demand for non-life insurance. In the presence of the

GFC, however, many of these factors become insignificant in explaining the demand

for non-life insurance in developed countries while they remain the determining factors

explaining the demand for non-life insurance in developing countries. The paper yields

insightful policy implications for governments around the world with regards to the

development of the non-life insurance sector, an important engine for long-run

economic growth and prosperity.

JEL classifications: G22, G01, O16.

Key words: culture, economic freedom, global financial crisis, non-life insurance.

Version date: 08th June, 2015.

* Department of Economics, Faculty of Business and Law, Deakin University, Burwood, VIC 3125,

Australia. Email addresses: [email protected] (T Trinh), [email protected] (P Sgro),

[email protected] (X Nguyen).

mailto:[email protected]



2

1. INTRODUCTION

The importance of insurance in the trade and development matrix was recognised at the first session in 1964 of the United Nations Conference on Trade and Development (UNCTAD). UNCTAD stated that “insurance market is an essential characteristic of economic growth” (United Nations 1964). Subsequently, the role and contributions of the insurance sector, which includes life and non-life insurance, on economic growth and development has been analysed and discussed extensively in the literature1. Previous authors have come up with a consensus in which both life and non-life insurance are important for GDP and social welfare (Outreville 1990; Brown 2000; Park & Lemaire 2012). The reason is that both life and non-life insurance have the characteristics of risk transfer and indemnification and pooling of losses and payment of fortuitous losses which help expand aggregate economic activities in a sustainable manner. Furthermore, non-life insurance provides the insureds with risk management and, at the same time, functions as an effective financial intermediary for the economy (Rejda 2011).2

Despite the undoubtedly paramount impact that life and non-life insurance have on the economy, little has been known about the socio-economic factors that may affect the demand for non-life insurance. This also contrasts sharply to a rich body of research on the determinants of demand for life insurance (see Chui & Kwok 2008, 2009; Park & Lemaire 2011 for instance, for a survey on determinants of demand for life insurance). Nevertheless, studies on this topic have identified economic freedom, income, education, culture and law the as the potential factors determining the demand for non-life insurance (Outreville 1990; Browne 2000; Esho et al. 2004; Elanggo & Jones 2011; Lee & Chiu 2012; Park & Lemaire 2012). These conclusions, however, are based on results from cross sectional data analyses in which both the heterogeneity between developed and developing countries and the endogeneity associated with the economic freedom index, an explanatory variable, have often been neglected. It is in this regard that a comprehensive study into the determinants of demand for non-life insurance in developed and developing countries is lacking.

The recent global financial crisis (GFC) that has severely dampened economic activities in many developed countries also raises the need for such a study that can demonstrate a rich set of socio economic factors that influence the demand for non-life insurance. As noted by Mckibbin & Stoeckel (2009), the GFC is captured by three main shocks: the bursting of the housing bubble in the United States, a sharp rise in the equity risk premium, and a reappraisal of risk by households causing them to discount their future labour income and increase savings and decrease consumption. As such, the GFC may change the way in which socio-economic factors impact the demand for non-life insurance. Such a study into this topic can help policy makers around the world find the correct pathway to develop the non-life insurance sector serving the purpose of economic growth and prosperity as well as coping with future

1 According to Organization for Economic Co-operation and Development (OECD), life insurance

includes life assurance, annuities, supplementary insurance, and permanent health insurance, whereas non-life insurance includes insurance against following risks: accident, sickness, land vehicles, railway rolling stock, aircraft, ships, goods in transit, fire and natural forces, other damage to property, motor vehicle liability, aircraft liability, liability for ships, general liability, credit, suretyship, miscellaneous financial loss and legal expenses (https://stats.oecd.org/glossary/detail.asp?ID=3040). 2 With regard to non-life insurance, data from Swiss Re (2014) reveal that worldwide non-life

insurance premium reached US$1997 billion in 2012 and occupied 2.68% of world GDP, 83.2% of which belonged to the OECD countries.

https://stats.oecd.org/glossary/detail.asp?ID=3040

3

financial crises in a more active and effective manner. The present paper aims to fulfil this objective.

To fill these gaps in the literature, the paper explores a unique panel dataset covering 36 developed and 31 developing countries in the period 2000-2011. The purpose of this paper is to address the following two key research questions: “How do the various socio-economic factors influence the demand for non-life insurance in developed and developing countries?”, and “How does the GFC change the way in which socio-economic factors impact the demand for non-life insurance in developed and developing countries?”. We contribute to the literature in the following ways. First, we incorporate the latest available datasets from the Fraser Institute, that contains updated data on the economic freedom index, Hofstede et al. (2010) and Minkov (2011), that contains new, unexplored cultural variables (long-term orientation, indulgence, and hypometropia), into our empirical study. Second, following Nelson and Singh (1998), Carlsson and Lundstrom (2002), and Rode and Coll (2012), we choose trade openness, indices of human development and life expectancy as instrumental variables for economic freedom index, which allow for the treatment of the endogeneity problem associated with this variable. Third, we perform the analysis for three groups of countries: all sample countries, developed countries, and developing countries. Finally, to capture the impact the GFC, we have separated the dataset into no crisis and crisis sub-periods.

Our main results can be summarized as follows. First, in line with the literature, we have found that economic freedom, income, bank development, urbanization, and a number of cultural and law variables are the drivers of the demand for non-life insurance. However, the impacts that these socio-economic variables would have on the demand for non-life insurance vary greatly between developed and developing countries. Some contrasting effects have even been identified between these two groups of countries. For example, bank development, education, urbanization and hypometropia (a new cultural variable) increase the demand for non-life insurance in developing countries while it decreases the demand for non-life insurance in developed countries. In contrast, long-term orientation and indulgence (two new cultural variables), uncertainty avoidance, and maintaining common law system increase the demand for non-life insurance in developed countries while they decrease the demand for non-life insurance in developing countries. These findings not only complement earlier results in the literature but also uncover the connection between the levels of development and the determinants of the demand for non-life insurance.

More importantly, a number of interesting results have also been discovered in regards to how the GFC influences the way socio-economic variables affect the demand for non-life insurance. In particular, we have found that indulgence, long-term orientation and power distance become insignificant in explaining the demand for non-life insurance in developed countries during GFC, while they remain the key drivers of the demand for non-life insurance in developing countries. Furthermore, the impact of bank development on demand for non-life insurance in developing countries during the GFC is unchanged while it becomes negative in developed countries. Our findings perhaps add further evidence for the profound effects of the GFC on developed rather than developing countries.

Our paper is the first that attempts to layout a comprehensive framework based on an instrumental variable analysis to examine factors that influence the demand for non-life insurance in both developed and developing economies before and during the

4

GFC. The results of the paper, as summarised above, are useful for not only domestic but also multinational non-life insurance companies in developing their existing markets and/or choosing newly potential markets. For policy makers, understanding the impact of various socio-economic factors on demand for non-life insurance may help adjust their regulations and policies to develop the non-life insurance sector which inturns foster economic development and social welfare and cope with future financial crisis in a stronger and more effective manner.

The paper will proceed as follows. Section 2 provides a survey of the literature on the determinants of demand for non-life insurance. Section 3 introduces the research methodology and data, followed by empirical results and discussions of results in Section 4. Finally, Section 5 offers some concluding remarks.

2. LITERATURE REVIEW

Nearly all the empirical work on the demand for non-life insurance take Sherden (1984) as a starting point. He pointed out that the increasing automobile purchase prices lead to decreases in the average price elasticities for physical damage coverages. Conversely, when the value of the car is reduced as a result of accidents, the price elasticities for coverages will be increased. The demand for physical damage coverages will depend on the net impact of these two trends in term of increasing the price of automobile insurance. Sherden showed that suburban consumers would be an ideal target market if non-life insurance companies want to maximize the amount of coverage. He also concluded that the need for automobile insurance coverage and the income risk are directly related. Concerning the theory, a paper by Beenstock et al. (1988) has been recognized as the first original research on demand for non-life insurance. Beenstock et al. developed the theoretical model based on the link between income and demand for non-life insurance. They considered both the supply and demand for insurance and found that demand for non-life insurance is affected by income, probability of a loss, price of non-life insurance, interest rate, financial development, and price of insurance. They showed that the price of non-life insurance, the value of property at risk, and the real interest rate have positive effects on the demand for non-life insurance while the impact of probability of loss on non-life insurance’s demand is negative and the impact of wealth on demand for non-life insurance is ambiguous. Subsequently, the results of demand-supply analysis proposed by Outreville (1990) suggested that financial development has a positive effect on quantity demanded for non-life insurance. This was followed by studies of Browne et al. (2000), Esho et al. (2004), Feyen et al. (2011), Elango and Jones (2011), and Park and Lemaire (2012). Altogether, determinants of demand for non-life insurance, based on previous studies, can be grouped into economic factors (in particular economic freedom), cultural factors, and other factors. These will be discussed in turn.

Economic freedom

Park et al. (2002) examined the determinants of insurance (sum of life and non-life insurance) pervasiveness using a cross-national data analysis. Park et al. used the index of economic freedom published by Heritage Foundations and the Wall Street Journal as a proxy for degree of regulation. Evidence of a negative and significant impact of the degree of regulation was found. The paper claimed that more regulation in a country leads to more difficulties in attracting and retaining insurers to provide insurance services. In the same line, previous papers have found that regulations have a significant impact on the insurance industry, as they might alter the supply of

5

insurance products. Requirements for entry of new firms and limits on foreign investment can lead to reducing the competitive ability of foreign insurers in comparison with domestic insurers. Therefore, decreasing the restrictions on entry into markets will increase the economic freedom and increase the competitive ability of the insurance market.

Elango and Jones (2011) analysed the drivers of non-life insurance demand in emerging markets from 1998-2008. They found out that financial freedom have positive impact on demand for non-life insurance while business freedom has a negative impact. The authors used 6 indicators of Heritage rather than the summary of economic freedom index as proxy for the degree of economic freedom.

The Fraser institute’s economic freedom index

The economic freedom index of Fraser institute (EFW), which includes five major areas: size of government, legal system and property rights, sound money, freedom to trade internationally and regulation, is used by a voluminous literature because of following reasons. First, Berggren (2003) stated that a large number of variables of EFW were derived from secondary sources, therefore, it was easy to verify them. Second, according to Gwartney and Lawson (2003), the index of economic freedom of Heritage Foundation/Wall Street Journal is of less value to researchers in case of analysing the effect of changing economic freedom across time periods. In addition, Indexes of Heritage Foundation/Wall Street Journal are both less precise and less transparent than those of Fraser Institute. Third, Doucouliagos and Ulubasoglu (2006) claimed that almost all researchers (33 out of 52 papers) used the economic freedom index of the Fraser Institute because it was the most comprehensive in terms of time span. Haan et al. (2006) critically discussed the economic freedom index of the Fraser institute and concluded that indicators of EFW are both useful and reliable because of their clear description. Finally, Justesen (2008) used EFW of Fraser Institute because it covered a large number of countries over a long-term period. Similarly, the benefits of economic freedom was also comprehensively analysed by Rode and Coll (2012) and Berggren (2003).

Cultural factors

Basically, there are five major research projects that have included cultural dimensions as an explanatory variable for the demand for non-life insurance. These are research projects of Hofstede (1983), Hofstede and Bond (1984), Hofstede et al. (2010), Schwartz (1994), Trompenaars and Hampden-Turner (1998), House et al. (2004), and Minkov (2011). While researchers such as Schwartz (1994), Trompenaars and Hampden-Turner (1998) focused on a richer reflection at a theoretical level of the national culture score, Hofstede’s (1983) work is considered as one of the most important studies that would provide fundamental values on the recent cross-cultural research (Park et al. 2002). The GLOBEL project of House et al. (2004), which identified cultural dimensions using a theory-based approach, is related conceptually and correlated empirically with aspects of Hofstede’s definition. Although some important refinement and clarification has been found in subsequent work, the GLOBAL project suggests that the Hofstede dimensions are robust (Leung et al. 2005).

Study of Park and Lemaire (2012) examined the impact of four cultural variables of Hofstede on demand for non-life insurance. They showed that in emerging countries with low in power distance and high in individualism and uncertainty avoidance,

6

demand for non-life insurance had a higher growth than other developing countries that had equal levels of income. However, Park and Lemaire (2012) did not include in their analyses the fifth and sixth cultural dimension which are introduced in a later stage by Hofstede et al. (2010): long-term orientation and indulgence.

Economic factor

All previous papers, such as Beenstock et al. (1988), Outreville (1990), Browne et al. (2000), Esho et al. (2004), Elanggo and Jones (2011), and Park and Lemaire (2012), used income as the main determinant of the demand for non-life insurance. These papers showed that income has a significantly positive influence on non-life insurance’s demand.

Institutional factors

Elango and Jones (2011) studied the drivers of insurance demand in emerging markets from 1998-2008. They examined the impact of institutional factors, economic factors and demographic factors on demand for non-life insurance. Results showed that financial freedom has a positive impact on demand for non-life insurance while business freedom has a negative impact on non-life insurance.

Demographic factors

Previous papers used urbanization, measured as percentage of population living in urban areas, as a proxy for loss probability. These papers stated that concentration of assets in urban areas leads to increasing loss probability. Browne et al. (2000) and Park and Lemaire (2012) found an ambiguous impact of this variable while Esho et al. (2004) found a positive impact. Feyen et al. (2011) examined the impact of population size on non-life insurance consumption. They found a significantly negative effect of population size on non-life insurance consumption. However, Nakata and Sawada (2007) found a positive but insignificant influence.

Social factors

Previous papers used education as a proxy for risk aversion. Browne et al. (2000) used enrolment in tertiary education while Esho et al. (2004) used secondary school enrolments. Browne et al. (2000) found an ambiguous impact of risk aversion on non-life insurance while Esho et al. (2004) found a significant positive impact. Furthermore, Esho et al. (2004) used the uncertainty avoidance index (UAI) of Hofstede as a proxy for risk aversion. They found a significant positive impact of UAI on demand for non-life insurance. Park and Lemaire (2012) examined the impact of four cultural variables of Hofstede, including power distance, individualism, masculinity, and uncertainty avoidance index on demand for non-life insurance. They found that in emerging countries with low power distance and high individualism and uncertainty avoidance, demand for non-life insurance had a higher growth than other developing countries that had equal levels of income.

Structural factors

Porta et al. (1998), and Park and Lemaire (2012) stated that common law (or English law) countries provide the highest protection for creditors’ rights or shareholders while French civil law countries provide the lowest protection. Park and Lemaire found that common law has a positive impact on demand for non-life insurance while Islam law has a negative impact on demand for non-life insurance.

7

Feyen et al. (2011) studied the impact of financial development, using the ratio of private credit to GDP and private bonds to GDP, on insurance market development. They found a positive effect. Besides, banks are an effective distribution channel to sell non-life insurance for insurers. Customers of banks will be advised to choose appropriate products for properties financed by banks.

Park and Lemaire (2012) used a political risk index as a potential explanatory variable in studying the impact of political risk on demand for non-life insurance. A positive impact was found.

Effects of the global financial crisis

Mckibbin & Stoeckel (2009) mentioned that there are three main shocks that capture during the GFC as follows: the bursting of the housing bubble in the United States, a sharp rise in the equity risk premium, and a reappraisal of risk by households causing them to discount their future labour income and increase savings and decrease consumption.

Harrington (2009) examined the role of the insurance sector in the global financial crisis. Harrington stated that insurance sector played a remarkably important role on the periphery of the crisis. Financial rating agencies publicized downgrades of leading “monoline” mortgage and bond insurers because of their significant losses. He also found that the systemic risk in the non-life insurance is relatively low compared with banking because of the non-life insurers’ great amounts of capital. Therefore, the need for broad government guarantees to prevent potential spread in the non-life insurance sector is less than banking.

The effect of the financial crisis on trade flows has been examined by Chor and Manova (2012). They found that the US imports is reduced by 2.5% more and 5.5% less under these respective scenarios via the effect of credit conditions, such as interbank rates. Findings show that through the easing of the credit crunch, the real economy is largely affected by the financial market disturbances. Costs in developing and manufacturing a product of domestic producers may be equally hurt by credit conditions. However, Chor and Manova (2012) also argued that exporters may be influenced more because of shipping’s expenses.

3. RESEARCH METHODOLOGY AND DATA

3.1 Hypothesis

Based on previous papers’ theoretical model and results, it is clear that economic freedom impacts the demand for non-life insurance. Similarly, cultural factors and legal system also impact on the perceived risks and thus the demand for non-life insurance. Combining with earlier developments in the literature, we propose a number of testable hypotheses, as presented in what follows.

i). Economic freedom and regulation

Hypothesis 1a: the larger the government, the higher the degree of freedom to exchange, the higher the degree of access to sound money, the higher the degree of regulation of credit, labour and business and the higher the degree of economic freedom in a country the higher the demand for non-life insurance.

Hypothesis 1b: the more developed the legal structure and security of property rights in a country, the lower the demand for non-life insurance.

8

ii). Cultural variables

Hypothesis 2a: the higher the degree of indulgence, long-term orientation, masculinity, and power distance in a country, the lower the demand for non-life insurance.

Hypothesis 2b: the higher the degree of individualism, uncertainty avoidance, and hypometropia in a country, the higher the demand for non-life insurance.

iii) Other factors

Hypothesis 3: the higher the level of income, banking development, education and urbanization in a country, the higher demand for non-life insurance.

iv). Legal system

Hypothesis 4: Islamic Law has a negative effect on demand for non-life insurance whereas Common Law has a positive effect on demand for non-life insurance.

3.2 Regression equation and variables

The basic empirical model is as follows:

INSit = α + β1EFIit + β2INCit + β3BSDit + β4EDUit + β5URBit + β6CULi + β7LSYi + Dyear + it

Where: INSit is the non-life insurance demand (density or penetration) for country i in year t; EFIit is index of economic freedom (including chain-index summary or 5 areas) for country i in year t; INCit is GDP per capita for country i in year t; BSDit is the bank development for country i in year t; EDUit is school enrolment, tertiary (% gross); URBit is urbanization for country i in year t; CULi is an array of cultural variables (Hofstede) that are time invariant, only vary across countries; LSYi is an array of dummy variables (Common law, Islamic law and international law) that vary across countries. Dyear is an time invariant array of annual dummy variables used to control the effect of time on

non-life insurance consumption; α is a constant term; β1 to β10 and are vectors of

coefficients, and it is the error term for country i in year t.

Results are presented as follows: First, we examine the determinants of insurance premium by cross-sectional and time series for 67 countries (see Table 19, Appendix 4, p.37) from 2000 to 2011 (12 years). Second, we examine the effect of determinants on non-life insurance’s demand before (2000-2006), during (2007-2011) the GFC by using panel dataset made up of 67 countries for every year in the period 2000-2011. Definitions and sources of data are described in Table 2 (see Appendix 2, p.20).

Dependent variable: this research uses density and penetration of non-life insurance provided by Sigma, Swiss Re as a proxy for non-life insurance’s demand3. This data has been available for the world annually since 1980. The non-life insurance density is defined as direct domestic premiums per capita in constant US dollar. This variable is our primary variable because it reflects the average amount or demand that an individual spends on non-life insurance. The main disadvantage of these proxies is aggregating premiums. Premiums across various lines of non-life insurance including non-compulsory and compulsory insurance, and differ across insurance lines and countries. For example, in Thailand, motor vehicle insurance’s premium is the greatest; meanwhile, non-life insurance in other countries focuses on liability

3 Beck, Kunt and Levine (2000, 2009) introduced a new dataset of indicators of financial structure and financial development which includes indicators of the insurance sector such as private credit, total assets, penetration, and density. However, this dataset only focuses on life insurance sector. To the best of our knowledge, there is no comparable dataset in the non-life insurance industry.

9

insurance. Unfortunately, premiums of each of the lines are unavailable for almost countries, with the exception of OECD and Vietnam. Furthermore, a limitation of the non-life insurance density is that we cannot separate the factors of demand and supply because of its aggregation across all non-life insurance policies (Beck & Webb 2003). To reduce these disadvantages, we will include explanatory variables on the demand side, such as urbanization, income and education, in the regression model to control for the bias caused by the missing variables. The second dependent variable is the non-life insurance penetration. It is defined as the ratio of direct domestic premiums per GDP. This variable shows the degree of non-life insurance activities relative to the size of the economy and is only used for robustness checks. Finally, following Park and Lemaire (2012), the non-life insurance density adjusted by purchasing power parity (PPP) factors is used to replace the non-life density, which is measured at US dollar. This variable is used to reflect the differences in prices between a country and the US. Similar to the second dependent variable, density at PPP is also used for purpose of robustness check.

Independent variables: economic freedom, income per capita, bank development, education, urbanization, culture and legal system described as in the following Table 2 (see Appendix 2, p.20).

3.3 Research methodology

Estimation strategy

Following Park and Lemaire (2012), this research will apply panel data analysis because of its many advantages. Cameron (2008), for instance, pointed out that panel data are repeated measures on individuals (i) over time (t) and using panel data has three major advantages. Firstly, using panel data leads to increasing precision estimation because of combining or pooling several time periods of data for each individual. The second reason is the ability of consistent of the fixed effects model. Finally, understanding more about the dynamics of individual behaviour is better than from a single cross country (Cameron & Trivedi 2005). In addition, this approach reduces collinearity and increases degree of freedom of the tests and the significance of results (Park and Lemaire 2012).

To examine both time variant and time in-variant variables in the regression equation presented in Section 3.2, we use following regression models:

- The pooled ordinary least squares (OLS) estimator (and with robust option):

corresponds to running OLS on the observations pooled across country i and year t.

- Pooled generalized least squares (GLS) estimation: GLS estimation has a special interest in connection with time series and cross-section observations. It will turn out to be asymptotically more efficient than system OLS because requires stronger assumptions than system OLS in order to be consistent. (Wooldridge 2002).

- To examine the sensitivity of results, another panel-regression estimation is used, that is random-effect model. While fixed-effect models cannot be used because of existing time in-variant variables (cultural and dummy variables), random-effect models is used to decrease the problem of autocorrelation caused by country-specific effects.

10

Statistical tests

While economic freedom index includes several economic factors, such as economic growth and total investment, using OLS and GLS may ignore the effects of other directions. Reverse causality, omitted variable bias and measurement error between economic freedom and demand for non-life insurance are possible. Therefore, we perform a Durbin-Wu-Hausman test for endogeneity.

Durbin-Wu-Hausman test: there are 4 steps to carry out the test as follow: (1) run a regression for the economic freedom variable using, as regressors, all explanatory variables plus instrumental variables that relevant in explaining the economic freedom variable; (2) extract the predicted values of the economic freedom variable (residuals); (3) run the main regression including residuals as explanatory variables; (4) using a F test, if the residual is significantly different from zero, null hypothesis is that the regressors are exogenous is rejected (Maddala & Lahiri 2009; Wooldridge 2002).

Endogeneity tests

To solve the endogeneity problems, we can apply two-stage least squares (2SLS) estimators using instrumental variables (IV) to account for endogeneity (Wooldridge 2002). The test of heteroskedasticity for instrumental variables (IV) estimation should be also used to check whether the sample is heteroskedastic or not. In term of heteroskedasticity, the IV_2SLS estimation must be combined with robust option (Baum & Schaffer 2007). Baum and Schaffer (2007) also stated that IV_2SLS is the efficient generalized method of moments (GMM) in term of assuming conditionally homoskedastic and independent errors.

Regarding to instrumental variables, we can use the lagged values of the economic freedom index as possible instruments since lagged values are less likely to be impacted by current shocks. Alternatively, we can identify some instruments appropriate for usage in examining the impact of economic freedom on non-life insurance demand by noting the following rule: a good instrument should correlate with the key independent variable (economic freedom index), but not with the main equation dependent variable (the non-life insurance density or penetration).

After that, Anderson canon. corr. LM statistic or Kleibergen-Paap rk LM statistic is carry out to test for underidentification. To reject the null of underidentication, the p-value of this test must be lower than 10%. As the number of instruments exceeds the number of endogenous variables, we also check the validity of instruments by applying the overidentification test. Sargan statistic or Hansen J statistic is test for overidentification. The null hypothesis of Sargan or Hansen J statistic test is that all instruments are valid. To reject the null hypothesis, the p-values for this test must be less than 10%.

Then, we check whether or not the instruments are weak by applying the weak identification test. One rule of thumb is that in term of a single endogenous regressor, instruments are not weak if a Wald F-statistic of Cragg-Donald is larger than 10. (Staiger & Stock 1997). The summary of statistic tests for endogeneity is described in Table 3 (see Appendix 4, p.21).

3.4 Instruments

A good instrument should correlate with the key independent variable but not with the main dependent variable. Following Nelson and Singh (1998), Carlsson and Lundstrom (2002), and Rode and Coll (2012), we choose the following variables as

11

instruments: One dummy variable measuring the degree of political right and civil freedom; A variable measuring the degree of trade openness (the sum of export and import as a share of GDP); A variable measuring human development of a country; An economic development variable (life expectancy). In addition, we also choose the gross capital formation (%); credit to government and state owned enterprises to GDP (%); financial system deposits to GDP (%); and 5 years lagged economic freedom index as other instruments.

Instruments are divided into 4 groups as follows: Group 1 includes lagged 1 period economic freedom; Group 2 includes dummy variable measuring the degree of political right and civil freedom, variable measuring the degree of trade openness, variable measuring human development of a country, and variable measuring economic development variable (life expectancy). Group 3 includes instruments: lagged 1 period economic freedom variable, the gross capital formation and group 2. Group 4 includes the gross capital formation and variables in group 2. Finally, group 5 includes credit to government and state owned enterprises to GDP (%), financial system deposits to GDP (%), the gross capital formation (%), and variables in group 2.

4. RESULTS AND DISCUSSION

4.1 Statistic test for heteroskedasticity.

Result in Table 4 (see Appendix 4, p.21) shows that disturbance of sample of all groups is heteroskedastic. Therefore, IV_2SLS should be carried out with robust option in term of all sample countries.

4.2 Statistic tests for endogeneity

As can be seen from Table 4 (see Appendix 4, p.21), results of the Durbin-Wu-Hausman tests show that endogeneity hypothesis of group 3 and 5 cannot be rejected.

Although overidentification tests of group 3 and 5 show that overidentification hypothesis of these groups cannot be rejected, however, tests for underidentification show that the equation is identified. In other words, the excluded instruments are relevant and correlated with the endogenous regressors. In addition, tests for weak instruments of group 3 and 5 show that instruments are not weak because their 1st F value are greater than 10. As mentioned in previous section, the lagged value of economic freedom may be correlated with the demand for non-life insurance rather than other instruments. Therefore, we can choose group 5 as the best instruments to solve the endogeneity problem.

4.3 Results and discussion

Determinants of demand for non-life insurance: all sample economies

In Table 9a (see Appendix 4, p.25), we investigate the impact on the demand for non-life insurance of the chosen variables based on pooled OLS, OLS with robustness, pooled FGLS, random-effects GLS and IV. Results show that economic freedom, income, bank development sector, education, and urbanization have a highly significant effect. Cultural variables, with exception of indulgence, individualism, uncertainty avoidance and hypometropia, have a significant influence. The two legal system variables have a highly significant effect.

12

The significant positive coefficient of economic freedom variable (consisting of components such as regulations, freedom to trade internationally, access to sound money, property rights, and legal system) is consistent with previous papers. The significantly positive coefficient of income confirms our prediction: a higher income per capita leads to increases demand for non-life insurance, consistent with all previous literature.

The significant positive impact of bank development sector on non-life insurance’s demand emphasis that bank development fosters non-life insurance. While Park and Lemaire (2012), and Elango and Jones (2011) found an ambiguous impact of education on demand for non-life insurance as well as unexpected negative impact of education found by Park et al. (2002) and Outreville (1990), a significant positive impact of education on demand for non-life insurance in our research supports the hypothesis that the more people are educated, the more risk averse they become. This result is consistent with result of Esho et al. (2004).

The significant positive impact of urbanization on demand for non-life insurance supports our hypothesis that higher urbanization leads to greater concentration of risks due to increasing assets and higher demand for non-life insurance.

For the former four cultural variables of Hofstede, the individualism and uncertainty avoidance variables have an insignificant effect on non-life insurance’ demand, similar to the finding of Park et al (2002). Likewise, masculinity variable has highly significant impact on demand for non-life insurance as expected. The interesting result is the positive impact of power distance variable. This result is not consistent with our hypothesis. It may be due to the fact that the bosses may be willing to help their subordinates and their family by choosing the non-life insurance to hedge the risks rather than using their finance directly. Another reason is that in short-term relationship, high powerful people choose the non-life insurance’s products that have yearly covering period, then, they will choose other ways rather than the non-life insurance when the relationship is developed to high level, such as life insurance’s products.

For the new cultural variables of Hofstede and Minkov, the fifth cultural variable (long-term orientation) has a significant negative impact on demand for non-life insurance as expected. This result conflicts with positive effect in life insurance sector of Park and Lemaire (2011). The reason may be due to the provision of financial support of parents and ancestors in societies with high long-term orientation. They can cover risks for their children without needing non-life insurance. The sixth cultural variable (indulgence) has a negative but insignificant impact on demand for non-life insurance. Finally, the new cultural variable of Monkov (hypometropia) has an insignificant effect on demand for non-life insurance. It should also be noted that Islamic law has a highly significant negative impact on demand for non-life insurance while common law’s impact is positive as expected.

In order to test the stability of determinants and examine how much variation is explained by economic, demographic, institutional and legal variables, we ran 7 regressions in Table 10 (see Appendix 4, p.28). As can be seen from our results in Table 10, the inclusion of the income variable increases the adjusted R-square coefficient from 0.888 in column (3) to 0.952 in column (7). This increasing of the adjusted R-square shows that income is a key determinant in explaining its impact on demand for non-life insurance. The bank development variable is next key determinants.

13

As discussed in the literature review, the areas of economic freedom that affect demand for non-life insurance are size of government, property rights and legal system, sound money, freedom to trade internationally, and regulation. In order to check the impacts of these variables on the demand for non-life insurance, we run several regressions in Table 13 (see Appendix 4, p.31). Results show that area of size of government, property rights and legal systems, access of sound money, and area of regulation have a significantly positive impact on demand for non-life insurance. This clearly supports the result of positive impact of summary economic freedom index on non-life insurance’s demand in Table 9a (see Appendix 4, p.25).

As mentioned in section 3.2, for robustness check, the non-life insurance density is replaced by penetration and density adjusted at PPP. In term of using penetration, Table 9b (see Appendix 4, p.26) shows results that almost determinants are still unchanged. Education’s variable becomes insignificant. This result is similar to study of Park and Lemaire (2012). In case of using PPP factor, results in Table 9c (see Appendix 4, p.27) show that in comparison with Density (USD per capita), almost determinants are still unchanged, exception for Common law. Common law becomes significant when using PPP factors. In Table 9a, 9b, and 9c, almost signs of the variables are the same, however, the value of the adjusted R-square in term of penetration and PPP factor are lower than the case of density. This implies that using non-life insurance density as a proxy for demand for non-life insurance is preferable.

Determinants of demand for non-life insurance in developed economies

Panel data analysis with 36 developed countries4 within 2000 – 2011 are shown in Table 11 (see Appendix 4, p.29). In comparison with all sample countries, results in developed economies have differences as follows:

First, urbanization’s impact becomes negative. It may be due to the fact that risks caused by increasing assets and vehicles are not high in developed countries. Specifically, this is because urbanization in developed countries has been already finished for many years. Citizens and councils have many experiences to manage and control the risks.

Second, indulgence variable’s impact becomes significantly positive. It is possible that developed countries with low indulgence, such as Latvia, Estonia, Hong Kong, Slovakia, Poland, South Korea and Italy, have high non-life insurance demand. As analysed, people in low indulgence or restraint stands may not accept risks due to their pessimism, therefore, they will buy non-life insurance as a way to hedge risks.

Third, individualism variable’s impact becomes significant. As can be seen from Table 5, 6 and 7 (see Appendix 4, p22-23), the mean of individualism in developed countries is higher than that in developing countries and all sample countries. Therefore, this result supports our hypothesis 2b.

Forth, long-term orientation (LTO)’s effect becomes positive. Developed countries have LTO higher than developing countries and all sample countries, this means that individuals in those countries may not depend on finance of their parents or ancestors. Therefore, demand for non-life insurance will be high as they normally hedge their risks.

Fifth, uncertainty avoidance (UAI) variable becomes significantly positive. This result is consistent with our hypothesis 6e. Although the mean of UAI of developed countries

4 The sub-sampling issue is based on the classification of World Bank and IMF.

14

is low than developing countries and all sample countries, its positive impact shows that individuals in developed countries choose buying non-life insurance as a way to hedge risks.

Sixth, power distance’s impact is significantly positive. This finding shows that individuals in developed countries may think that buying non-life insurance to avoid risks is better than depending on their boss. Seventh, hypometropia (HPM) variable’s impact becomes negative. The mean of HPM shows that developed countries have HPM lower than developing countries and all sample countries. Therefore, lower violence leads to decreases in demand for non-life insurance.

Finally, the Common Law dummy variable’s impact becomes significantly positive in the regression results.

Determinants of demand for non-life insurance in developing economies

Panel data analysis is carried out for the 31 developing countries within 2000 – 2011. Results are reported in Table 12 (see Appendix 4, p.30). In comparison with all sample countries, a number of differences are identified and discussed in turn.

First, the masculinity variable’s impact is significantly negative. This result supports our hypothesis 2a.

Second, Islamic law dummy variable’s impact is still significant. This is because Islamic Law is more prevalent in developing countries with Islam law are larger than developed countries.

Similarly, the Common law dummy variable’s impact becomes significantly negative in the regression results.

Determinants of demand for non-life insurance before, during the global financial crisis

For all sample countries, panel data analysis is performed for 67 countries annually in the period 2000 – 2011, and is reported in Table 15 (see Appendix 4, p.33). Results show that in the GFC 2007-2011, economic freedom variable, income variable, bank development variable, indulgence variable, long-term orientation variable, power distance variable, and legal system variables are still significant. Education variable, urbanization variable, and hypometropia variable become insignificant while indulgence variable and masculinity variable become significant.

For developed countries, a similar analysis is performed for 36 countries. Results, which are reported in Table 16 (see Appendix 4, p.34), show that in period 2007-2011, only economic freedom, income, urbanization, individualism, and masculinity, and hypometropia variables are still significant as before the GFC. Indulgence, long-term orientation, and power distance variables become insignificant. Education, bank development, and common law variables’ signs are changed, especially bank development variable become significant negative.

For developing countries, results, which as reported in Table 17 (see Appendix 4, p.35) based on 31 countries, show that in period 2007-2011, almost variables are still significant during the GFC as before the GFC. The effect of education, and individualism variables become significantly negative while power distance and common law variables become significantly positive. Finally, masculinity variable becomes significantly negative.

15

5. CONCLUSION

This paper fills this gap in the literature by undertaking a comprehensive empirical analysis of the determinants of demand for nonlife insurance using a panel dataset covering 36 developed and 31 developing countries in the period 2000-2011. Our instrumental variables analysis, where trade openness and indices of human development and life expectancy are used as instrumental variables for economic freedom index, reveals that economic freedom, income, urbanization, education, culture and bank development all have significant effects on the demand for non-life insurance. Furthermore, new cultural variables of Hofstede and Minkov are examined. Results show that for all sample countries, long-term orientation has significantly negative effect on non-life insurance’s demand while the impacts of hypometropia and indulgence are insignificant.

We have also examined the determinants of demand for non-life insurance in developed and developing economies independently. Generally, the effects of the economic freedom and income are similar to the case of all sample economies. The impacts of bank development and education in term of developing countries are different with developed countries and similar to the case of all sample countries. For new cultural variables, the effect of indulgence, long-term orientation, power distance, and hypometropia variable are different and highly significant in developed and developing economies. Similarly, the legal system dummy variables’ effects are also different.

Finally, effects of determinants on non-life insurance are investigated during the global financial crisis (2007-2011). For all sample countries, we have found that the impacts of education, urbanization, and hypometropia become insignificant. For developed countries, the effects of indulgence, long-term orientation, and power distance on demand for non-life insurance also become insignificant. The interesting feature is that the impacts of these variables in developing economies are mostly unchanged during the global financial crisis.

The findings of this study are useful for not only domestic but also multinational non-

life insurance companies in developing their existing markets and/or choosing newly

potential markets. For policy makers, understanding the impact of various socio-

economic factors on demand for non-life insurance may help adjust their regulations

and policies to develop the non-life insurance sector which inturns foster economic

development and social welfare.

ACKNOWLEDGEMENTS

We thank participants of seminar at Deakin University for many helpful comments.

16

References

Baum, C & Schaffer, M, 2007. ‘Enhanced routines for instrumental variables/GMM

estimation and testing’, Boston College Economics Working Paper No. 667.

Beenstock, M, Dickinson, G & Khajuria, S 1988, ‘The Relationship Between Property-

Liability Insurance Premiums and Income: An International Analysis’, Journal of Risk

and Insurance, vol. 55, no. 2, pp. 259-272.

Beck, T, & Webb, I 2003, ‘Economic, demographic, and institutional determinants of

life insurance consumption across countries’, World Bank Economic Review, vol. 17,

no. 1, pp. 51–88.

Beck. T, Kunt, A & Levine, R 2000, ‘A new database on the structure and development

of the financial sector’, The World Bank Economic Review, vol. 14, no. 3, pp. 597-605.

Beck. T, Kunt, A & Levine, R 2009, ‘Financial Institutions and markets across

Countries and over time: Data and Analysis’, Policy Research Working Paper.

Berggren, N 2003, ‘The benefits of economic freedom: A survey’, The independent

review, vol. 8, no. 2, pp. 193-211.

Browne, M, Chung, J, & Frees, E 2000, ‘International Property-Liability Insurance

Consumption’, Journal of Risk and Insurance, vol. 67, no. 1, pp. 73-90.

Cameron, A & Trivedi, P 2005, Microeconometrics: Methods and Applications,

Cambridge University Press, Cambridge, UK.

Carlsson, F & Lundstrom, S 2002, ‘Economic freedom and growth: Decomposing the

effects’, Public Choice, vol. 112, pp. 335–344.

Chor and Manova (2012), ‘Off the cliff and back? Credit conditions and international

trade during the global financial crisis’, Journal of International Economics, vol. 87, pp.

117-133.

Chui, A & Kwok, C 2008, ‘National Culture and Life Insurance Consumption’, Journal

of International Business Studies, vol. 39, pp. 88-101.

Doucouliagos, C & Ulubasoglu, M 2006,’ Economic freedom and economic growth:

Does specification make a difference?’, European Journal of Political Economy, vol.

22, pp. 60– 81.

Elanggo, B & Jones, J 2011, ‘Drivers of Insurance Demand in Emerging Markets’,

Journal of Service Science Research, vol. 3: 185-204

Esho, N, Kirievsky, A, Ward, D & Zurbruegg, R 2004, ‘Law and the Determinants of

Property-Casualty Insurance’, Journal of Risk and Insurance, vol. 71, no. 2, pp. 265-

283.

Feyen, E, Lester, R & Rocha, R 2011, ‘What Drives the Development of the Insurance

Sector’, Policy Research Working Paper No. 5572, World Bank.

Freedom in the world 2015, ‘Individual country ratings and status, FIW 1973-2015’,

Freedom House.

Global Financial Development Database (GFDD) 2014, the World Bank

17

Gwartney, J, Lawson, R & Block, W 1996, Economic Freedom in the World, 1975–

1995, Fraser Institute, Vancouver.

Gwartney, J & Lawson, R 2003, ‘The concept and measurement of economic

freedom’, European Journal of Political Economy, vol. 19, pp. 405–430.

Gwartney, J, Lawson, R & Hall, J 2014, Economic Freedom of the World: 2014 Annual

Report, Fraser Institute, Vancouver.

Haan, J, Lundstro¨m, S & Sturm, J 2006, ‘Market-oriented institutions and policies and

economic growth: A critical survey’, Journal of Economic Surveys, vol. 20, pp. 157–

191.

Harrington, S 2009, ‘The financial crisis, systemic risk, and the future of insurance

regulation’, The journal of risk and insurance, vol. 76, no. 14, pp. 785-819.

Hofstede, G 1983, ‘The cultural relativity of organizational practices and theories’,

Journal of International Business Studies, vol. 14, no. 2, pp. 75–89.

Hofstede, G 1995, ‘Insurance as a Product of National Values’, The Geneva Papers,

vol. 20, pp. 423-429.

Hofstede, G & Bond, M 1988, ‘The Confucius connection: From cultural roots to

economic growth’, Organizational Dynamics, vol. 15, no. 1, pp. 4–21.

Hofstede, G 2006, ‘What did GLOBE really measure? Researchers’ minds versus

respondents’ minds’, Journal of International Business Studies, vol. 37, pp. 882-889.

Hofstede, G, Hofstede, G.J & Minkov, M 2010, Cultures and Organizations: software

of the mind: intercultural cooperation and its importance for survival. 3rd edn, Mc Graw

Hill.

House, R, Hanges, P, Javidan, M, Dorfman, P & Gupta, V 2004, Culture, leadership,

and organizations: The GLOBE study of 62 societies, Thousand Oaks, CA: Sage.

Human development index 2014, Human development report, United Nations

Development Programme.

International Association of Insurance Supervisors 2012, Global Insurance Market

Report, Switzerland.

Lee, C & Chiu, Y 2012, ‘The impact of real income on insurance premiums: Evidence

from panel data’, International Review of Economics and Finance, vol. 21, pp. 246–

260.

Leung, K & Bhagat, R, Buchan, N, Erez, M & Gibson, C 2005, ‘Culture and

international business: Recent advances and their implications for future research’,

Journal of International Business Studies, vol. 36, no. 4, pp.357–378.

Maddala, G & Lahiri, K 2009, Introduction to econometrics, 4th edn, Wiley, England.

Mckibbin & Stoeckel (2009), ‘Modelling the global financial crisis’, Oxford Review of

Economic Policy, vol. 25, no. 4, pp. 581-607.

Messick, R 1996, World Survey of Economic Freedom, 1995-1996, Freedom House

and Transaction Publishers.

18

Minkov, M 2011, Cultural differences in a globalizing world. 1st edn, Bingley, UK :

Emerald.

Nakata, H, & Sawada, Y 2007, ‘Demand for Non-life Insurance: A Cross-Country

Analysis’, CIRJE Working Paper F-461.

Nelson, M & Singh, R 1998, ‘Democracy, economic freedom, fiscal policy and growth

in LDCs: a fresh look’, Economic Development and Cultural Change, vol. 46, pp. 677–

696.

Outreville, J 1990, ‘The Economic Significance of Insurance Markets in Developing

Countries’, Journal of Risk and Insurance, vol. 18, no. 3, pp. 487-498.

Park, H, Borde, S & Choi, Y 2002, ‘Determinants of Insurance Pervasiveness: A

Cross-National Analysis’, International Business Review, vol. 11, pp. 79-96.

Park, S & Lemaire, J 2012, ‘The Impact of Culture on the Demand for Non-Life

Insurance’, ASTIN Bulletin, vol. 42, pp. 501-527.

Porta, R, Lopez-de-Silanes, F, Shleifer, A & Vishny, R 1999, ‘The quality of

Government’, Journal of Law, Economics and Organization, vol.15, pp. 222-279.

Rejda, G 2011, Principle of Risk Management and Insurance, Pearson education, Inc.

Rode and Coll (2012), ‘Economic freedom and growth. Which policies matter the

most?’, Constitutional Political Economy, vol. 23, pp. 95–133.

Schwartz, S 1994, ‘Beyond individualism Collectivism: New cultural dimensions of

values’ in U. Kim, H. C. Triandis, C. Kagitcibasi, S. C. Choi & G. Yoon (Eds),

individualism and collectivism: Theory, method and applications, Thousand Oaks, CA:

Sage, 85–99.

Sherden 1984, ‘An Analysis of the Determinants of the Demand for Automobile

Insurance’, The Journal of Risk and Insurance, vol. 51, no. 1, pp. 49-62.

Staiger, D & Stock, J 1997, ‘Instrumental Variables Regression with Weak

Instruments’, Econometrica, vol. 65, pp. 557–86.

Swiss Reinsurance Company 2014, Various Years, Sigma, Zurich.

Swiss Reinsurance Company 2013, World insurance in 2012: Progressing on the long

and winding road to recovery, Sigma, Zurich.

The World Factbook, 2014, CIA

Trompenaars, F & Hampden-Turner, C 1998, Riding the waves of culture:

Understanding diversity in global business, New York: McGraw-Hill.

United Nations 1964, Proceedings of the United Nations Conference on TRADE AND

DEVELOPMENT: Final Act and Report, vol.1, New York.

World Development Indicators 2014, the World Bank

Wooldridge, J., 2002, Econometric Analysis of Cross Section and Panel Data, The

MIT Press, Cambridge, Massachusetts, London, England.

19

Appendix 1: Economic freedom

Table1: Indicators of economic freedom used in three sets Fraser Institute Heritage Foundation Freedom House

Area 1: Size of Government:

1. Government consumption

2. Transfers and subsidies

3. Government enterprises and investment

4. Top marginal tax rate

4.1 Top marginal income tax rate

4.2 Top marginal income and payroll tax rate

Area 2: Legal System and Property Rights:

1. Judicial independence

2. Impartial courts

3. Protection of property rights

4. Military interference in rule of law and politics

5. Integrity of the legal system

6. Legal enforcement of contracts

7. Regulatory restrictions on the sale of real

property

8. Reliability of police

9. Business costs of crime

Area 3: Sound Money:

1. Money growth

2. Standard deviation of inflation

3. Inflation: most recent year

4. Freedom to own foreign currency bank

accounts

Area 4: Freedom to Trade Internationally:

1. Tariffs

1.1 Revenue from trade taxes (% of trade sector)

1.2 Mean tariff rate

1.3 Standard deviation of tariff rates

2. Regulatory trade barriers

2.1 Non-tariff trade barriers

2.2 Compliance cost of importing and exporting

3. Black-market exchange rates

4. Controls of the movement of capital and people

4.1 Foreign ownership/investment restrictions

4.2 Capital controls

4.3 Freedom of foreigners to visit

Area 5: Regulation:

1. Credit market regulations

1.1 Ownership of banks

1.2 Private sector credit

1.3 Interest rate controls/negative real interest

rates

2. Labour market regulations

2.1 Hiring regulations and minimum wage

2.2 Hiring and firing regulations

2.3 Centralized collective bargaining

2.4 Hours regulations

2.5 Mandated cost of worker dismissal

2.6 Conscription

3. Business regulations

3.1 Administrative requirements

3.2 Bureaucracy costs

3.3 Starting a business

3.4 Extra payments/bribes/favouritism

3.5 Licensing restrictions

3.6 Cost of tax compliance

Area 1: Business Freedom

1. Starting a business—procedures (number);

2. Starting a business—time (days);

3. Starting a business—cost (% of income per

capita);

4. Starting a business—minimum capital (% of

income per capita);

5. Obtaining a license—procedures (number);

6. Obtaining a license—time (days);

7. Obtaining a license—cost (% of income per

capita);

8. Closing a business—time (years);

9. Closing a business—cost (% of estate); and

10. Closing a business—recovery rate (cents on

the dollar)

Area 2: Trade Freedom

1. The trade-weighted average tariff rate and

2. Non-tariff barriers

Area 3: Fiscal Freedom

1. The top tax rate on individual income,

2. The top tax rate on corporate income, and

3. Total tax revenue as a percentage of GDP.

Area 4: Government Size

1. Government expenditures as a percentage of

GDP3.

Area 5: Monetary Freedom

1. The weighted average inflation rate for the most

recent three years and

2. Price controls.

Area 6: Investment Freedom

1. This factor scrutinizes each country’s policies

toward foreign investment, as well as its policies

toward capital flows internally.

Area 7: Financial Freedom

1. The extent of government regulation of financial

services.

2. The extent of state intervention in banks and

other financial services.

3. The difficulty of opening and operating financial

services firms (for both domestic and foreign

individuals).

4. Government influence on the allocation of credit.

Area 8: Property Rights

1. A country’s laws protect private property rights

and the degree to which its government enforces

those laws.

Area 9: Freedom from Corruption

1. Transparency International’s Corruption

Perceptions Index (CPI).

Area 10: Labour Freedom

1. Regulations concerning minimum wages, laws

inhibiting layoffs, severance requirements, and

measurable regulatory burdens on hiring, hours,

and so on.

Area 1 Political rights

1. Electoral process

2. Political pluralism and participation

3. Functioning of government

Area 2 Civil liberties

1. Freedom of expression and belief

2. Associational and organizational rights

3. Rule of law

4. Personal autonomy and individual rights

20

Appendix 2: Definition and sources of data

Table 2: Definition and sources of data

Variable Hypothesis Description Time periods

Data of source

Density of non-life insurance at constant prices

Direct domestic premiums per capita, US dollars 2000-2011 Sigma, Swiss Re

Density at purchasing power parity (PPP) prices

The non-life insurance density adjusted by purchasing power parity (PPP) factors

2000-2011 Sigma (Swiss Re), IMF

Penetration of non-life insurance

Direct domestic premiums as a % of GDP 2000-2011 Sigma, Swiss Re

Economic freedom index + Using summary chain-linked index of Fraser Institute, economic freedom index is calculated based on 5 areas: (1) size of government, (2) legal system and property rights, (3) access to sound money, (4) freedom to trade internationally, and (5) regulation. The index ranges from 0 to 10 with higher values indicating more freedom.

2000-2011 Fraser Institute

Income per capita + GDP per capita in constant 2005 US dollars 2000-2011 The Global Financial Development Database (GFDD), the World bank

Bank development sector + Deposit money banks' assets to GDP (%) 2000-2011 GFDD, the World bank

Education + School enrolment, tertiary (% gross) 2000-2011 WDI, World bank

Urbanization + Urban population (% of total) 2000-2011 WDI, World bank

Indulgence - Indulgence stands for a society that people have the right to enjoy life and have fun. The index ranges from 0 to 100 with higher values indicating more happiness, leisure, and enjoying life.

Invariant time

Hofstede et al (2010) and Minkov, M., (2011)

Individualism + Individualism regards to societies that attachment between individuals are loose. The index ranges from 0 to 100 with higher values indicating low developing human relationship, more looking after him or herself.

Long-term orientation - Long-term orientation refers to a positive, dynamic, and future oriented culture linked with four ‘positive’ Confucian values. The index ranges from 0 to 100 with higher values indicating more national pride, the preservation of family values and traditions, and saving.

Masculinity - Masculinity regards to societies that social gender roles are clearly distinct. In these societies, men focus on their material success, such as career and business whereas women focus on improvement the quality of their family. The index ranges from 0 to 100 with higher values indicating more masculinity.

Uncertainty avoidance + Uncertainty avoidance refers to a society that people feel threatened by uncertain situations and they want to avoid these matters. The index ranges from 0 to 100 with higher values indicating more uncertainty avoidance.

Power distance - Power distance index refers to measuring of the interpersonal power or influence between boss and subordinates; or measuring the degree of inequality among people in a society. The index ranges from 0 to 100 with higher values indicating more inequality, power distance.

Hypometropia + Hypometropia stands for a society that have high national murder rates and acceptance of the mortal risks. The index ranges from 0 to 1000 with higher values indicating more serious violence and murder rate.

Islam law - Dummy variables, value =1 for countries with a Common law, Islam law, otherwise value =0

Invariant time

The World Factbook, CIA

Common law +

Instruments: dummy variable of political right and civil freedom: value =1 if freedom or partly freedom, otherwise =0, data source: Freedom House; the degree of trade openness: the sum of export and import as a share of GDP, data source: WDI; human development index: data source: United Nations Development Programme; life expectancy, the gross capital formation: data source: WDI; credit to government and state owned enterprises to GDP (%), financial system deposits to GDP (%): data source: GFDD.

21

Appendix 3: Tables of statistic tests and data description.

Table 3: Summary of statistic tests Test Heteroskedasticity test Endogeneity test Underidentification

test Overidentification test

Weak Instruments test

Pagan-Hall general test statistic

Ho: Disturbance is homoscedastic. Reject Ho: p-value <10%.

Durbin-Wu-Hansman

Ho: EF variable is exogenous. Reject Ho: p-value<10%

Hausman Ho: difference in coefficients is not systematic . Reject Ho: p-value <10%

Anderson canon. corr. LM statistic

Ho: Instruments is underidentified. Reject Ho: p-value<10%

Sargan statistic Ho: all instruments are valid. Reject Ho: p-value<10%

F-statistic Ho: instruments are weak. Reject Ho: F>10

Table 4: Summary of statistic tests for endogeneity (all sample countries)

Tests Group 1

(1 instruments)

Group 2

(4 instruments)

Group 3

(6 instruments)

Group 4

(5 instruments)

Group 5

(7 instruments)

Heteroskedasticity test P-value=0.000 disturbance is heteroskedastic.

P-value=0.000 disturbance is heteroskedastic.




Durbin-Wu-Hausman statistic

P-value = 0.1673 P-value = 0.1673 P-value = 0.011 P-value = 0.216 P-value = 0.002

Conclusion Exogeneity hypothesis cannot be rejected.

Exogeneity hypothesis cannot be rejected.

Exogeneity hypothesis is rejected.

Exogeneity hypothesis cannot be rejected.

Exogeneity hypothesis is rejected.

Underidentification P-value=0.000 instruments are relevant.

P-value=0.000 instruments are relevant.

Overidentification P-value=0.000 instruments are not valid.

P-value=0.000 instruments are not valid.

Weak Instruments 1st F = 152.974 >10 instruments are not weak.

1st F = 18.88 >10 instruments are not weak.

22

Table 5: Descriptive statistics (all sample countries)

Table 6: Descriptive statistics (developed countries)

Variables Observations Mean Std. Dev. Min Max

Log of density of non-life insurance at constant prices (lnINSC) 804 5.05 1.85 -0.51 8.45

Log of density of non-life insurance at PPP prices (lnINSP) 804 5.57 1.47 0.56 8.30

Log of penetration of non-life insurance (lnINSD) 804 0.50 0.65 -1.71 2.25

Log of economic freedom index (lnEFI) 801 1.95 0.13 1.33 2.20

Log of size of government (lnarea1) 801 1.79 0.26 0.94 2.24

Log of property rights & legal system (lnarea2) 801 1.82 0.35 0.37 2.26

Log of access to sound money (lnarea3) 801 2.12 0.19 1.00 2.29

Log of freedom to trade internationally (lnarea4) 801 2.01 0.17 1.05 2.27

Log of the regulation (lnarea5) 802 1.92 0.13 1.44 2.19

Log of income per capita (lnGDPP) 797 9.14 1.31 5.86 11.38

Education (EDU) 649 49.61 23.49 2.61 113.98

Bank development (BANK) 743 83.14 51.93 4.74 245.13

Urbanization (URB) 804 68.28 18.91 10.83 100

Indulgence (INDUL) 732 0.48 0.23 0.00 1.00

Individualism (IDV) 780 0.44 0.25 0.06 0.91

Long-term orientation (LTO) 744 0.48 0.22 0.13 1.00

Masculinity (MAS) 792 0.49 0.19 0.08 1.10

Uncertainty avoidance (UAI) 792 0.67 0.23 0.08 1.12

Power distance (PDI) 792 0.60 0.21 0.11 1.04

Hypometropia (HPM) 660 1.92 1.72 0 10

Islam law (ISL) 804 0.10 0.31 0 1

Common law (CL) 804 0.25 0.44 0 1


Log of density of non-life insurance at constant prices (lnINSC) 432 6.36 0.97 3.18 8.45










Education (EDU) 366 62.37 18.49 3.06 113.98


Urbanization (URB) 432 74.82 16.12 10.83 100




Masculinity (MAS) 432 0.50 0.23 0.08 1.10



Hypometropia (HPM) 384 1.20 0.50 0.00 2.38

Islam law (ISL) 432 0.03 0.16 0.00 1.00

Common law (CL) 432 0.28 0.45 0.00 1.00

23

Table 7: Descriptive statistics (developing countries)


Dependent variable

Log of density of non-life insurance at constant prices (lnINSC) 372 3.53 1.42 -0.51 6.07










Education (EDU) 283 33.12 18.49 2.61 78.63


Urbanization (URB) 372 60.70 19.10 23.59 93.50




Masculinity (MAS) 360 0.48 0.12 0.21 0.73



Hypometropia (HPM) 276 2.92 2.24 0.49 10.00

Islam law (ISL) 372 0.19 0.40 0.00 1.00

Common law (CL) 372 0.23 0.42 0.00 1.00

24

Table 8: Correlations (all sample countries)

lnINSC lnEFI lnGDPP EDU BANK URB INDUL IDV LTO MAS UAI PDI HPM ISL CL

lnINSC 1.00

lnEFI 0.73 1.00

lnGDPP 0.96 0.71 1.00

EDU 0.68 0.49 0.69 1.00

BANK 0.65 0.47 0.60 0.26 1.00

URB 0.60 0.44 0.62 0.45 0.22 1.00

INDUL 0.36 0.37 0.35 -0.00 0.19 0.43 1.00

IDV 0.65 0.60 0.67 0.45 0.38 0.48 0.17 1.00

LTO 0.12 0.01 0.15 0.23 0.17 0.04 -0.52 0.06 1.00

MAS 0.02 0.03 0.05 -0.26 0.13 -0.04 0.12 0.11 -0.00 1.00

UAI -0.00 -0.19 0.01 0.08 -0.20 0.23 -0.10 -0.32 -0.03 -0.13 1.00

PDI -0.61 -0.62 -0.63 -0.51 -0.38 -0.43 -0.33 -0.64 0.03 0.10 0.19 1.00

HPM -0.30 -0.06 -0.33 -0.24 -0.46 0.03 0.43 -0.25 -0.43 -0.12 0.10 0.09 1.00

ISL -0.22 -0.08 -0.23 -0.21 0.00 -0.12 0.03 -0.19 -0.14 0.05 -0.29 0.39 0.03 1.00

CL 0.11 0.29 0.06 0.05 0.24 -0.11 0.27 0.31 -0.34 0.14 -0.57 -0.18 -0.02 0.21 1.00

25

Appendix 4: Tables of results Table 9a: Determinants of non-life insurance demand (all sample countries)

Note: *,**, and *** indicate significance at the 10%, 5% and 1% level, respectively. Numbers in brackets show standard errors. Year fixed effects are included but not reported here.

Variables Pooled OLS (1) Pooled OLS (2) (Robustness)

Pooled FGLS (3)

Random-effects GLS (4)

IV (group 5) (5)

lnEFI 0.941*** (0.264) 0.766*** (0.273)

0.941*** (0.257) 2.064*** (0.197) 3.990*** (0.603)

lnGDPP 1.200*** (0.042) 1.237*** (0.043)

1.200*** (0.041) 1.395*** (0.075) 1.034*** (0.055)

BANK 0.003*** (0.001) 0.003*** (0.001)

0.003*** (0.001) 0.003*** (0.000) 0.003*** (0.001)

EDU 0.002 (0.001) 0.002 (0.001) 0.002 (0.001) 0.003** (0.001) 0.003 (0.002) (p-value=0.113)

URB 0.004** (0.002) 0.002 (0.002) 0.004** (0.002) 0.0001 (0.004) 0.005** (0.002)

INDUL -0.094 (0.160) -0.125 (0.165) -0.094 (0.156) -0.623 (0.456) -0.025 (0.202)

IDV 0.003 (0.137) -0.097 (0.141) 0.003 (0.133) -0.472 (0.414) -0.118 (0.166)

LTO -0.245** (0.109) -0.233** (0.112)

-0.245** (0.106) -0.478 (0.353) -0.350*** (0.128)

MAS -0.287*** (0.088) -0.204** (0.091)

-0.287*** (0.086) -0.090 (0.293) -0.266*** (0.107)

UAI 0.021 (0.121) 0.119 (0.125) 0.021 (0.117) -0.220 (0.385) -0.062 (0.142)

PDI 0.300** (0.136) 0.201 (0.141) 0.300** (0.132) 0.744* (0.432) 0.715*** (0.172)

HPM 0.024 (0.022) 0.030 (0.023) 0.024 (0.021) 0.088 (0.068) -0.037 (0.031)

ISL -0.315*** (0.107) -0.247*** (0.111)

-0.315*** (0.105) -0.462 (0.348) -0.557*** (0.131)

CL 0.112* (0.063) 0.166*** (0.065)

0.112* (0.061) 0.083 (0.195) -0.024 (0.079)

Constant -8.032*** (0.476) -7.936*** (0.493)

-8.032*** (0.464) -11.533*** (0.746)

-12.581*** (0.938)

Observations 504 504 504 504 488

R-squared 0.957 0.954 Within=0.892; between=0.950; overall=0.949

Endogeneity: yes.

Underidentification:no.

Overidentification: yes.

Instruments: not weak

Adjusted R2 0.955 0.952 Wald chi2=11197.93

Wald chi2=4383.19

26

Table 9b: Determinants of non-life insurance demand (all sample countries, using Penetration (Premium per GDP)


Variables Pooled OLS (1) Pooled OLS (robustness) (2)

Pooled FGLS (3) Random-effects GLS (4) IV (group 5) (5)

Premium per GDP

USD per capita

Premium per GDP

USD per capita

Premium per GDP

USD per capita

Premium per GDP

USD per capita

Premium per GDP

USD per capita

lnEFI 0.293 (0.250)

0.941*** (0.264)

0.263 (0.254)

0.766*** (0.273)

0.293 (0.243)

0.941*** (0.257)

0.284* (0.158)

2.064*** (0.197)

3.482*** (0.542)

3.990*** (0.603)

lnGDPP 0.298*** (0.039)

1.200*** (0.042)

0.324*** (0.040)

1.237*** (0.043)

0.298*** (0.038)

1.200*** (0.041)

0.310*** (0.065)

1.395*** (0.075)

0.126** (0.052)

1.034*** (0.055)

BANK 0.003*** (0.001)

0.003*** (0.001)

0.003*** (0.001)

0.003*** (0.001)

0.003*** (0.001)

0.003*** (0.001)

0.002*** (0.0004)

0.003*** (0.000)

0.003*** (0.001)

0.003*** (0.001)

EDU 0.001 (0.001)

0.002 (0.001)

0.001 (0.001)

0.002 (0.001)

0.001 (0.001)

0.002 (0.001)

0.002 (0.001)

0.003** (0.001)

0.002 (0.002)

0.003 (0.002) (p-value=0.113)

URB 0.002 (0.002)

0.004** (0.002)

0.001 (0.002)

0.002 (0.002)

0.002 (0.002)

0.004** (0.002)

0.009** (0.004)

0.0001 (0.004)

0.004 (0.002) (P-value=0.110)

0.005** (0.002)

INDUL -0.197 (0.151)

-0.094 (0.160)

-0.205 (0.154)

-0.125 (0.165)

-0.197 (0.147)

-0.094 (0.156)

-0.392 (0.425)

-0.623 (0.456)

-0.122 (0.197)

-0.025 (0.202)

IDV 0.102 (0.129)

0.003 (0.137)

-0.015 (0.131)

-0.097 (0.141)

0.102 (0.126)

0.003 (0.133)

-0.156 (0.390)

-0.472 (0.414)

-0.030 (0.165

-0.118 (0.166)

LTO -0.153 (0.103)

-0.245** (0.109)

-0.160 (0.104)

-0.233** (0.112)

-0.153 (0.100)

-0.245** (0.106)

-0.140 (0.338)

-0.478 (0.353)

-0.266** (0.124)

-0.350*** (0.128)

MAS -0.250*** (0.083)

-0.287*** (0.088)

-0.124 (0.085)

-0.204** (0.091)

-0.250*** (0.081)

-0.287*** (0.086)

-0.053 (0.283)

-0.090 (0.293)

-0.229** (0.107)

-0.266*** (0.107)

UAI 0.063 (0.114)

0.021 (0.121)

0.159 (0.116)

0.119 (0.125)

0.063 (0.111)

0.021 (0.117)

-0.129 (0.366)

-0.220 (0.385)

0.024 (0.137)

-0.062 (0.142)

PDI 0.301** (0.129)

0.300** (0.136)

0.223* (0.131)

0.201 (0.141)

0.301** (0.125)

0.300** (0.132)

0.305 (0.416)

0.744* (0.432)

0.734*** (0.163)

0.715*** (0.172)

HPM 0.045** (0.021)

0.024 (0.022)

0.047** (0.021)

0.030 (0.023)

0.045** (0.020)

0.024 (0.021)

0.030 (0.064)

0.088 (0.068)

-0.019 (0.029)

-0.037 (0.031)

ISL -0.344*** (0.102)

-0.315*** (0.107)

-0.287*** (0.103)

-0.247*** (0.111)

-0.344*** (0.099)

-0.315*** (0.105)

-0.421 (0.335)

-0.462 (0.348)

-0.593*** (0.121)

-0.557*** (0.131)

CL 0.173*** (0.060)

0.112* (0.063)

0.220*** (0.061)

0.166*** (0.065)

0.173*** (0.058)

0.112* (0.061)

0.268 (0.189)

0.083 (0.195)

0.027 (0.074)

-0.024 (0.079)

Constant -3.308*** (0.450)

-8.032*** (0.476)

-3.466*** (0.458)

-7.936*** (0.493)

-3.308*** (0.450)

-8.032*** (0.464)

-3.549*** (0.684)

-11.533*** (0.746)

-8.073*** (0.837)

-12.581*** (0.938)

Observations

504 504 504 504 504 504 504 504 504 488

R-squared 0.674 0.957 0.667 0.954 Within=0.225; between=0.662; overall=0.643

Within=0.892; between=0.950; overall=0.949

Instruments are relevant, not weak

Instruments are relevant, not weak

Adjusted R2

0.657 0.955 0.650 0.952

Wald chi2 1040.12 11197.93

27

Table 9c: Determinants of non-life insurance demand (all sample countries, using PPP factors)


Variables Pooled OLS (1) Pooled OLS (robustness) (2)

Pooled FGLS (3) Random-effects GLS (4) IV (group 5) (5)

PPP per capita

USD per capita

PPP per capita

USD per capita

PPP per capita

USD per capita

PPP per capita

USD per capita

PPP per capita

USD per capita

lnEFI -0.122 (0.266)

0.941*** (0.264)

-0.149 (0.277)

0.766*** (0.273)

-0.122 (0.259)

0.941*** (0.257)

0.410*** (0.159)

2.064*** (0.197)

3.496*** (0.613)

3.990*** (0.603)

lnGDPP 0.967*** (0.042)

1.200*** (0.042)

0.980*** (0.044)

1.237*** (0.043)

0.967*** (0.041)

1.200*** (0.041)

1.142*** (0.067)

1.395*** (0.075)

0.771*** (0.057)

1.034*** (0.055)

BANK 0.002*** (0.001)

0.003*** (0.001)

0.001*** (0.001)

0.003*** (0.001)

0.002*** (0.001)

0.003*** (0.001)

0.001*** (0.0004)

0.003*** (0.000)

0.002*** (0.001)

0.003*** (0.001)

EDU 0.0002 (0.001)

0.002 (0.001)

0.001 (0.002)

0.002 (0.001)

0.0002 (0.001)

0.002 (0.001)

0.002 (0.001)

0.003** (0.001)

0.004 (0.002)

0.003 (0.002) (p-value=0.113)

URB 0.003** (0.002)

0.004** (0.002)

0.002 (0.002)

0.002 (0.002)

0.003** (0.002)

0.004** (0.002)

0.011*** (0.004)

0.0001 (0.004)

0.005** (0.002)

0.005** (0.002)

INDUL -0.376** (0.161)

-0.094 (0.160)

-0.380** (0.168)

-0.125 (0.165)

-0.376** (0.157)

-0.094 (0.156)

-1.255*** (0.462)

-0.623 (0.456)

-0.296 (0.203)

-0.025 (0.202)

IDV -0.003 (0.138)

0.003 (0.137)

-0.112 (0.143)

-0.097 (0.141)

-0.003 (0.134)

0.003 (0.133)

-0.766* (0.462)

-0.472 (0.414)

-0.148 (0.186)

-0.118 (0.166)

LTO -0.041 (0.110)

-0.245** (0.109)

-0.074 (0.114)

-0.233** (0.112)

-0.041 (0.107)

-0.245** (0.106)

-0.397 (0.371)

-0.478 (0.353)

-0.156 (0.131)

-0.350*** (0.128)

MAS -0.399*** (0.089)

-0.287*** (0.088)

-0.261*** (0.092)

-0.204** (0.091)

-0.399*** (0.086)

-0.287*** (0.086)

-0.101 (0.312)

-0.090 (0.293)

-0.366*** (0.116)

-0.266*** (0.107)

UAI 0.158 (0.122)

0.021 (0.121)

0.282** (0.126)

0.119 (0.125)

0.158 (0.118)

0.021 (0.117)

-0.385 (0.401)

-0.220 (0.385)

0.055 (0.144)

-0.062 (0.142)

PDI 0.321** (0.137)

0.300** (0.136)

0.226 (0.143) (P-value-0.114)

0.201 (0.141)

0.321** (0.134)

0.300** (0.132)

0.706 (0.459)

0.744* (0.432)

0.814*** (0.184)

0.715*** (0.172)

HPM 0.047** (0.022)

0.024 (0.022)

0.037* (0.023)

0.030 (0.023)

0.047** (0.022)

0.024 (0.021)

0.104 (0.071)

0.088 (0.068)

-0.025 (0.031)

-0.037 (0.031)

ISL -0.120 (0.109)

-0.315*** (0.107)

-0.059 (0.113)

-0.247*** (0.111)

-0.120 (0.106)

-0.315*** (0.105)

-0.368 (0.371)

-0.462 (0.348)

-0.412*** (0.145)

-0.557*** (0.131)

CL 0.348*** (0.064)

0.112* (0.063)

0.397*** (0.066)

0.166*** (0.065)

0.348*** (0.062)

0.112* (0.061)

0.399* (0.209)

0.083 (0.195)

0.197*** (0.086)

-0.024 (0.079)

Constant -3.304*** (0.481)

-8.032*** (0.476)

-3.304*** (0.499)

-7.936*** (0.493)

-3.304*** (0.468)

-8.032*** (0.464)

-5.713*** (0.738)

-11.533*** (0.746)

-8.734*** (0.959)

-12.581*** (0.938)

Observations 504 504 504 504 504 504 504 504 488

R-squared 0.917 0.957 0.910 0.954 Within=0.805; between=0.906; overall=0.904

Within=0.892; between=0.950; overall=0.949

Instruments are relevant, not weak.

Instruments are relevant, not weak.

Adjusted R2 0.913 0.955 0.905 0.952

Wald chi2 5569.69 11197.93

28

Table 10: Determinants of non-life insurance demand (robustness)


Variables (1) (2) (3) (4) (5) (6) (7) full factors

lnEFI 1.094*** (0.276)

3.401*** (0.369)

0.254 (0.237)

0.849*** (0.265)

0.846*** (0.240)

0.766*** (0.273)

lnGDPP 1.305*** (0.038)

1.293*** (0.040)

1.240*** (0.030)

1.192*** (0.041)

1.184*** (0.030)

1.237*** (0.043)

BANK 0.003*** (0.001)

0.010*** (0.001)

0.003*** (0.0004)

0.003*** (0.001)

0.002*** (0.000)

0.003*** (0.001)

EDU -0.002 (0.001)

0.002 (0.001)

0.034*** (0.002)

0.004*** (0.001)

0.004*** (0.001)

0.002 (0.001)

URB 0.003 (0.002)

0.002 (0.002)

-0.008*** (0.003)

0.0002 (0.002)

0.002 (0.001)

0.002 (0.002)

INDUL -0.155 (0.165)

-0.150 (0.165)

2.363*** (0.209)

-0.027 (0.122)

0.040 (0.164)

-0.125 (0.165)

IDV -0.357*** (0.139)

-0.077 (0.140)

2.077*** (0.189)

0.299*** (0.108)

0.057 (0.137)

-0.097 (0.141)

LTO -0.108 (0.110)

-0.194* (0.111)

-0.273* (0.166)

-0.046 (0.103)

-0.269*** (0.103)

-0.233** (0.112)

MAS -0.103 (0.089)

-0.216** (0.091)

0.488*** (0.134)

-0.250*** (0.082)

-0.170* (0.091)

-0.204** (0.091)

UAI 0.327*** (0.122)

0.136 (0.124)

0.794*** (0.183)

0.371*** (0.098)

0.151 (0.106)

0.119 (0.125)

PDI 0.029 (0.138)

0.106 (0.135)

0.596*** (0.207)

-0.002 (0.125)

0.140 (0.121)

0.201 (0.141)

HPM -0.019 (0.022)

0.050** (0.022)

-0.141*** (0.032)

0.050*** (0.012)

0.020 (0.023)

0.030 (0.023)

ISL -0.154 (0.111)

-0.191* (0.109)

0.833*** (0.165)

-0.101 (0.094)

-0.340*** (0.072)

-0.247*** (0.111)

CL 0.340*** (0.063)

0.214*** (0.091)

-0.933*** (0.092)

0.166*** (0.056)

0.001 (0.048)

0.166*** (0.065)

Constant -9.394*** (0.486)

-6.882*** (0.279)

-9.164*** (0.699)

-7.120 -7.737*** (0.490)

-7.666*** (0.387)

-7.936*** (0.493)

Observations 531 504 509 577 504 595 504

Adjusted R2 0.947 0.953 0.888 0.951 0.951 0.952 0.952

F 398.48 421.39 168.67 488.73 425.80 660.17 399.74

29

Table 11: Determinants of non-life insurance demand (developed countries, density)



Pooled FGLS (3)


IV (group 5) (5)

lnEFI 1.799*** (0.596) 1.485*** (0.497)

1.799*** (0.571)

0.661* (0.355) 12.750*** (2.686)

lnGDPP 1.171*** (0.064) 1.136*** (0.053)

1.171*** (0.061)

1.420*** (0.124)

0.803*** (0.124)

BANK 0.001 (0.001) 0.001 (0.001) 0.001 (0.001) 0.001** (0.001) -0.003** (0.001)

EDU -0.002 (0.002) 0.005*** (0.002)

-0.002 (0.002) 0.002 (0.002) -0.005* (0.003)

URB -0.009*** (0.003) -0.016*** (0.002)

-0.009*** (0.003)

-0.009 (0.006) -0.015*** (0.004)

INDUL 0.807*** (0.223) 1.379*** (0.186)

0.807*** (0.213)

0.312 (0.618) 0.312 (0.352)

IDV 0.515*** (0.192) 0.472*** (0.160)

0.515*** (0.184)

0.612 (0.550) 1.126*** (0.307)

LTO 0.553*** (0.171) 0.624*** (0.142)

0.553*** (0.163)

0.728 (0.476) -0.386 (0.367)

MAS -0.684*** (0.123) -0.338*** (0.103)

-0.684*** (0.118)

-0.473 (0.336) -1.227*** (0.212)

UAI 0.429*** (0.169) 1.185*** (0.141)

0.429*** (0.162)

0.599 (0.498) 0.162 (0.261)

PDI 0.445*** (0.142) 0.339*** (0.119)

0.445*** (0.136)

0.548 (0.454) 0.734*** (0.216)

HPM -0.139* (0.078) -0.057 (0.065)

-0.139* (0.075) 0.014 (0.208) -0.988*** (0.225)

ISL

CL 0.366*** (0.105) 0.416*** (0.088)

0.366*** (0.101)

0.461* (0.274) -0.317 (0.253)

Constant -9.864*** (1.016) -9.711*** (0.847)

-9.864*** (0.964)

-10.102*** (1.351)

-24.252*** (3.838)

Observations 309 309 309 309 293


Endogeneity: yes.

Underidentification:no.

Overidentification: yes.

Instruments: weak


Wald chi2=1814.85

30

Table 12: Determinants of non-life insurance demand (developing countries, density)



Pooled FGLS (3)


IV (group 5) (5)

lnEFI 1.753*** (0.291) 1.196*** (0.156)

1.753*** (0.271) 1.753*** (0.291)

lnGDPP 1.764*** (0.092) 1.969*** (0.049)

1.764*** (0.085) 1.764*** (0.092)

BANK 0.007*** (0.001) 0.008*** (0.001)

0.007*** (0.001) 0.007*** (0.001)

EDU 0.007*** (0.002) 0.012*** (0.001)

0.007*** (0.002) 0.007*** (0.002)

URB 0.009*** (0.002) 0.005*** (0.001)

0.009*** (0.002) 0.009*** (0.002)

INDUL -2.085*** (0.282) -1.701*** (0.151)

-2.085*** (0.263) -2.085*** (0.282)

IDV 0.045 (0.245) 0.805*** (0.131)

0.045 (0.228) 0.045 (0.245)

LTO -1.824*** (0.228) -3.310*** (0.122)

-1.824*** (0.213) -1.824*** (0.228)

MAS -0.290 (0.395) -2.553*** (0.211)

-0.290 (0.368) -0.290 (0.395)

UAI -2.636*** (0.488) -4.976*** (0.260)

-2.636*** (0.454) -2.636*** (0.488)

PDI 0.298 (0.310) 2.789*** (0.166)

0.298 (0.289) 0.298 (0.310)

HPM 0.281*** (0.030) 0.368*** (0.016)

0.281*** (0.028) 0.281*** (0.030)

ISL -1.168*** (0.179) -2.253*** (0.096)

-1.168*** (0.167) -1.168*** (0.179)

CL -0.101 (0.097) -0.605*** (0.052)

-0.101 (0.090) -0.101 (0.097)

Constant -12.110*** (0.810) -11.686*** (0.433)

-12.110*** (0.793)

-12.008*** (0.851)

Observations 195 195 195 195


Endogeneity: no


Wald chi2=3630.50

31

Table 13: Determinants of non-life insurance demand (all sample countries, categories of economic freedom index)


Variables Pooled OLS Pooled OLS (Robustness)

Pooled FGLS Random-effects GLS

IV (group 5)

lnarea1 -0.081 (0.070) -0.089 (0.071) -0.081 (0.068) 0.175*** (0.066) 0.108 (0.218)

lnarea2 0.028 (0.128) 0.063 (0.129) 0.028 (0.124) 0.219* (0.122) 1.345 (1.397)

lnarea3 0.372*** (0.126) 0.338*** (0.126) 0.372*** (0.122) 0.585*** (0.067) 0.390 (0.793)

lnarea4 -1.315*** (0.184) -1.444*** (0.185) -1.315*** (0.179) -0.035 (0.174) 3.689 (4.745)

lnarea5 1.897*** (0.223) 1.903*** (0.224) 1.897*** (0.216) 0.627*** (0.174) -3.793 (5.573)

Income 1.193*** (0.043) 1.267*** (0.043) 1.193*** (0.041) 1.404*** (0.075) 1.005*** (0.189)

BANK 0.004*** (0.0005)

0.003*** (0.001) 0.004*** (0.0005) 0.003*** (0.0005) 0.001 (0.002)

EDU 0.004*** (0.001) 0.003** (0.001) 0.004*** (0.001) 0.002* (0.001) 0.000 (0.004)

URB 0.0004 (0.002) 0.0001 (0.002) 0.0004 (0.002) 0.0004 (0.004) 0.014 (0.014)

INDUL -0.006 (0.143) -0.215 (0.144) -0.006 (0.139) -0.580 (0.412) -0.090 (0.321)

IDV 0.088 (0.127) -0.104 (0.128) 0.088 (0.124) -0.423 (0.372) -0.515 (0.664)

LTO -0.297*** (0.103) -0.337*** (0.103) -0.297*** (0.100) -0.445 (0.314) -0.176 (0.236)

MAS -0.311*** (0.089) -0.225*** (0.090) -0.311*** (0.087) -0.143 (0.261) 0.015 (0.456)

UAI 0.210* (0.111) 0.178 (0.111) (P-

value=0.110) 0.210* (0.108) -0.154 (0.346) -0.204 (0.536)

PDI 0.292** (0.122) 0.201* (0.122) 0.292** (0.118) 0.643* (0.382) 0.540** (0.241)

HPM 0.023 (0.021) 0.037* (0.022) 0.023 (0.021) 0.088 (0.061) 0.076 (0.091)

ISL -0.436*** (0.098) -0.404*** (0.098) -0.436*** (0.095) -0.433 (0.308) -0.036 (0.410)

CL 0.022 (0.068) 0.071 (0.068) 0.022 (0.066) 0.117 (0.175) 0.383 (0.322)

Constant -7.962*** (0.421) -8.037*** (0.423) -7.962*** (0.408) -10.682*** (0.687) -8.421*** (1.370)

Observations 504 504 504 504 -Endogeneity: yes.

- Underidentification: yes.

- Overidentification: yes.

Instrument: weak.

R-squared 0.966 0.966 Within=0.896; between=0.956; overall=0.955.

Wald chi2= 4667.24

Adjusted R2 0.964 0.964 Wald chi2= 14397.85

32

Table 14: Summaries of results of determinants of demand for non-life insurance

(*) Park H et al (2002) used summary economic freedom of Heritage in 1997. This index was constructed as follows: the lower scores of economic freedom index, the higher of degree of economic freedom. However, then heritage changed construction of economic freedom’s scores as follows: the higher scores of economic freedom index, the higher of degree of economic freedom. Therefore, negative impact of economic freedom index on consumption of insurance is considered as positive impact for new construction of economic freedom index.

Variables

Hypothesis

Our results (2000-2011)

Non-life insurance Both life & non-life insurance

Life insurance

All sample countries

Developed countries

Developing countries

Park and Lemaire (2012)

Elanggo & Jones (2011)

Esho et al (2004)

Browne et al (2000)

Outreville (1990)

Lee & Chiu (2012)

Park et al (2002)

Park & Lemaire (2011)

Chui & Kwok (2008)

lnEFI + + + + 7 areas + (*)

lnGDPP + + + + + + + + + + + + +

BANK + + Insig. + - +

EDU + + Insig. + -(insig.) -(insig.) + - - + - (Insig.)

URB + + - + insig + - (Insig.) + (Insig.)

INDUL - - (insig.) + -

IDV + Insig. + + + + (Insig.) + +

LTO - - + - +

MAS - - - - +(insig.) + - -

UAI + Insig. + - + + (p-value

= 16%)

- (Insig.) + +

PDI - + + + - + (Insig.) + -

HPM + Insig. - +

ISL - - - - -

CL + + + - - (insig.) + (Insig.)

+

33

Table 15: Determinants of non-life insurance demand before and during the GFC (all sample countries)

Note: *,**, and *** indicate significance at the 10%, 5% and 1% level, respectively. Numbers in brackets show standard errors. Year fixed effects are included but not reported here


Pooled FGLS Random-effects GLS IV (group 5)

2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

lnEFI 1.033*** (0.310)

-0.321 (0.543)

0.911*** (0.319)

-0.423 (0.535)

1.033*** (0.300)

-0.321 (0.515)

2.326*** (0.237)

0.018 (0.434) 3.701*** (0.706)

2.417*** (0.769)

lnGDPP 1.158*** (0.047)

1.420*** (0.089)

1.163*** (0.048)

1.563*** (0.088)

1.158*** (0.045)

1.420*** (0.085)

1.156*** (0.093)

0.157*** (0.109)

1.036*** (0.056)

1.251*** (0.104)

BANK 0.005*** (0.001)

0.001 (0.001)

0.005*** (0.001)

-0.001 (0.001)

0.005*** (0.001)

0.001 (0.001)

0.004*** (0.001)

-0.0003 (0.001)

0.003*** (0.001)

0.002* (0.001)

EDU 0.002 (0.002)

-0.001 (0.003)

0.002 (0.002)

-0.002 (0.002)

0.002 (0.002)

-0.001 (0.002)

0.003* (0.002) -0.001 (0.002) 0.002 (0.002)

0.001 (0.003)

URB 0.006*** (0.002)

0.001 (0.003)

0.006*** (0.002)

0.002 (0.003)

0.006*** (0.002)

0.001 (0.003)

0.002 (0.005) 0.007 (0.006) 0.008*** (0.003)

0.001 (0.003)

INDUL -0.194 (0.197)

-0.305 (0.273)

-0.223 (0.203)

-0.902*** (0.268)

-0.194 (0.190)

-0.305 (0.258)

-0.098 (0.497) -0.983* (0.535) -0.106 (0.247)

-0.162 (0.324)

IDV 0.029 (0.168)

-0.100 (0.230)

-0.0004 (0.174)

-0.715*** (0.227)

0.029 (0.163)

-0.100 (0.218)

-0.036 (0.441) -0.539 (0.472) -0.129 (0.203)

-0.123 (0.263)

LTO -0.353*** (0.135)

-0.173 (0.178)

-0.302** (0.139)

-0.424** (0.175)

-0.353*** (0.131)

-0.173 (0.168)

-0.370 (0.359) -0.434 (0.424) -0.421*** (0.155)

0.231 (0.191)

MAS -0.108 (0.108)

-0.773*** (0.151)

-0.018 (0.111)

-0.377*** (0.149)

-0.108 (0.104)

-0.773*** (0.143)

-0.101 (0.292) -0.634* (0.337) -0.132 (0.122)

-0.690*** (0.170)

UAI 0.058 (0.146)

-0.193 (0.204)

0.165 (0.151)

-0.065 (0.201)

0.058 (0.141)

-0.193 (0.193)

0.055 (0.399) -0.545 (0.454) -0.046 (0.168)

-0.221 (0.230)

PDI 0.301* (0.171)

0.109 (0.215)

0.181 (0.176)

-0.1844 (0.212)

0.301* (0.165)

0.109 (0.204)

0.612 (0.436) 0.343 (0.489) 0.758*** (0.235)

0.378* (0.205)

HPM 0.039 (0.028)

0.033 (0.035)

0.049* (0.029)

0.045 (0.034)

0.039 (0.027)

0.033 (0.033)

0.019 (0.069) 0.060 (0.079) -0.033 (0.040)

-0.002 (0.039)

ISL -0.378*** (0.132)

-0.215 (0.177)

-0.298** (0.136)

-0.183 (0.174)

-0.378*** (0.127)

-0.215 (0.167)

-0.484 (0.345) -0.436 (0.384) -0.628*** (0.161)

-0.326* (0.194)

CL 0.154** (0.076)

0.088 (0.109)

0.182** (0.078)

0.329*** (0.107)

0.154** (0.074)

0.088 (0.103)

0.075 (0.192) 0.198 (0.230) 0.029 (0.086)

-0.008 (0.134)

Constant -8.368*** (0.569)

-6.402*** (0.946)

-8.232*** (0.586)

-6.872*** (0.931)

-8.368*** (0.550)

-6.402*** (0.893)

-10.747*** (0.818)

-8.071*** (1.110)

-12.487*** (1.148)

-10.435*** (1.168)

Observations

319 185 319 185 319 185 319 185 310

R-squared 0.960 0.956 0.958 0.956 Wald chi2=7728.72

Wald chi2=3982.31

Within=0.858; between=0.961; overall=0.957.

Wald chi2= 2464.55

Within=0.663; between=0.950; overall=0.951.

Wald chi2= 835.43

-Endogeneity: yes.

- Underidentification: no.


Instrument: not weak.

-Endogeneity: yes.

- Underidentification: no.


Instrument: not weak.

Adjusted R2

0.958 0.951 0.955 0.951

34

Table 16: Determinants of non-life insurance demand before and during the GFC (developed countries)




2000-2006

2007-20011 (crisis)

2000-2006

2007-20011 (crisis)

2000-2006

2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

lnEFI

2.139*** (0.782)

3.191** (1.481)

1.817*** (0.597)

4.609*** (0.934)

2.139*** (0.741)

3.191** (1.357)

1.122*** (0.356)

0.106 (0.737)

15.782*** (4.720)

16.670*** (4.257)

lnGDPP

1.123*** (0.075)

1.308*** (0.120)

1.054*** (0.057)

1.386*** (0.076)

1.123*** (0.071)

1.308*** (0.110)

1.050*** (0.149)

1.294*** (0.197)

0.700*** (0.181)

1.290*** (0.157)

BANK

0.001 (0.001)

-0.001 (0.001)

0.001 (0.001)

-0.002*** (0.001)

0.001 (0.001)

-0.001 (0.001)

0.002*** (0.001)

-0.0004 (0.001)

-0.010** (0.004)

-0.003** (0.002)

EDU

-0.004* (0.002)

0.000 (0.003)

0.002 (0.002)

0.012*** (0.002)

-0.004* (0.002)

0.000 (0.003)

0.001 (0.002) -0.001 (0.003)

-0.017*** (0.006)

0.004 (0.005)

URB

-0.004 (0.003)

-0.017*** (0.005)

-0.015*** (0.003)

-0.016*** (0.003)

-0.004 (0.003)

-0.017*** (0.005)

-0.013 (0.008)

-0.002 (0.010)

0.006 (0.007)

-0.042*** (0.009)

INDUL 0.657*** (0.265)

0.438 (0.409)

1.516*** (0.202)

0.400 (0.258)

0.657*** (0.251)

0.438 (0.375)

1.321* (0.738)

0.304 (0.836)

0.458 (0.431)

-0.827 (0.674)

IDV 0.261 (0.225)

0.994*** (0.372)

0.415** (0.172)

0.718*** (0.234)

0.261 (0.213)

0.994*** (0.341)

0.639 (0.642) 0.408 (0.722)

0.081 (0.374)

2.702*** (0.654)

LTO 0.334* (0.210)

0.416 (0.296)

0.569*** (0.161)

-0.061 (0.187)

0.334* (0.199)

0.416 (0.271)

0.701 (0.566) 0.238 (0.679)

-1.141* (0.628)

-0.268 (0.505)

MAS -0.536*** (0.143)

-1.225*** (.0240)

-0.402*** (0.109)

-0.615*** (0.151)

-0.536*** (0.135)

-1.225*** (.0220)

-0.469 (0.388)

-0.807* (0.430)

-0.980 (0.265)

-2.199*** (0.402)

UAI 0.272 (0.197)

0.384 (0.334)

1.050*** (0.150)

0.817*** (0.210

0.272 (0.186)

0.384 (0.306)

0.542 (0.577) -0.027 (0.685)

-0.419 (0.393)

1.100** (0.465)

PDI 0.493*** (0.174)

0.249 (0.231)

0.394*** (0.133)

0.080 (0.146)

0.493*** (0.165)

0.249 (0.212)

0.507 (0.527) 0.222 (0.565)

0.959*** (0.311)

0.351 (0.302)

HPM -0.151* (0.096)

-0.360** (0.153)

-0.147** (0.074)

-0.180* (0.096)

-0.151* (0.091)

-0.360** (0.140)

-0.069 (0.236)

-0.121 (0.263)

-1.218*** (0.385)

-1.315*** (0.337)

ISL

CL 0.401*** (0.128)

0.157 (0.185)

0.421*** (0.098)

-0.217* (0.117)

0.401*** (0.121)

0.157 (0.170)

0.447 (0.313) 0.187 (0.378)

-0.782* (0.475)

-0.258 (0.318)

Constant -9.481*** (1.289)

-11.959*** (2.837)

-8.926*** (0.984)

-16.597*** (1.789)

-9.481*** (1.222)

-11.959*** (2.601)

-7.559*** (1.595)

-6.410*** (2.418)

-28.584*** (6.718)

-35.731*** (7.706)

Observations 197 112 197 112 197 112 197 112 105 R-squared 0.914 0.848 0.948 0.936 Wald

chi2=2102.10

Wald chi2 = 622.41

Within=0.884; between=0.911; overall=0.908. Wald chi2=1390.26

Within= 0.398; between= 0.856; overall= 0.830. Wald chi2= 151.00

-Endogeneity: yes. - Underidentification: no. - Overidentification: yes. Instrument: weak.

-Endogeneity: yes. - Underidentification: no. - Overidentification: no. Instrument: weak.

Adjusted R2 0.905 0.820 0.942 0.924

35

Table 17: Determinants of non-life insurance demand before and during the GFC (developing countries)




2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

2000-2006 2007-20011 (crisis)

lnEFI

2.105*** (0.361)

0.425 (0.573)

1.422*** (0.206)

0.364 (0.326)

2.105*** (0.329)

0.425 (0.493)

2.653*** (0.280)

0.425 (0.573)

1.147* (0.608)

lnGDPP

1.999*** (0.121)

1.488*** (0.128)

2.085*** (0.069)

1.698*** (0.073)

1.999*** (0.110)

1.488*** (0.110)

2.394*** (0.224)

1.488*** (0.128)

1.446*** (0.114)

BANK

0.005*** (0.001)

0.011*** (0.002)

0.007*** (0.001)

0.008*** (0.001)

0.005*** (0.001)

0.011*** (0.002)

0.004*** (0.002)

0.011*** (0.002)

0.012*** (0.002)

EDU

0.010*** (0.003)

-0.009** (0.004)

0.018*** (0.002)

-0.005*** (0.002)

0.010*** (0.003)

-0.009** (0.003)

0.008* (0.004)

-0.009** (0.004)

-0.006** (0.003)

URB

0.010*** (0.003)

0.011*** (0.002)

0.008*** (0.002)

0.009*** (0.001)

0.010*** (0.003)

0.011*** (0.002)

-0.002 (0.010)

0.011*** (0.002)

0.012*** (0.002)

INDUL -2.596*** (0.393)

-2.042*** (0.356)

-1.855*** (0.224)

-2.390*** (0.203)

-2.596*** (0.357)

-2.042*** (0.306)

-3.453*** (1.143)

-2.042*** (0.356)

-1.963*** (0.313)

IDV 0.273 (0.314)

-1.601*** (0.355)

1.361*** (0.179)

-1.706*** (0.202)

0.273 (0.285)

-1.601*** (0.305)

0.205 (1.346)

-1.601*** (0.355)

-1.451*** (0.318)

LTO -2.056*** (0.300)

-1.233*** (0.269)

-3.596*** (0.171)

-2.035*** (0.153)

-2.056*** (0.273)

-1.233*** (0.231)

-2.013** (0.958)

-1.233*** (0.269)

-1.274*** (0.236)

MAS -0.228 (0.527)

-0.436 (0.442)

-2.726*** (0.300)

-0.699*** (0.252)

-0.228 (0.479)

-0.436 (0.380)

1.059 (2.162)

-0.436 (0.442)

-0.285 (0.393)

UAI -3.912*** (0.670)

0.283 (0.583)

-6.335*** (0.381)

-0.870*** (0.332)

-3.912*** (0.609)

0.283 (0.502)

-3.805** (1.688)

0.283 (0.583)

0.135 (0.514)

PDI -0.008 (0.397)

1.009*** (0.394)

2.289*** (0.226)

1.656*** (0.225)

-0.008 (0.361)

1.009*** (0.339)

-0.983 (1.558)

1.009*** (0.394)

1.134*** (0.349)

HPM 0.344*** (0.041)

0.200*** (0.032)

0.394*** (0.023)

0.222*** (0.018)

0.344*** (0.037)

0.200*** (0.028)

0.437*** (0.124)

0.200*** (0.032)

0.205*** (0.028)

ISL -1.400*** (0.236)

-0.713*** (0.224)

-2.482*** (0.135)

-1.118*** (0.128)

-1.400*** (0.215)

-0.713*** (0.193)

-1.157 (0.745)

-0.713*** (0.224)

-0.830*** (0.203)

CL -0.153 (0.125)

0.237* (0.137)

-0.721*** (0.071)

0.155** (0.078)

-0.153 (0.113)

0.237* (0.118)

-0.087 (0.502)

0.237* (0.137)

0.195* (0.121)

Constant -13.327*** (1.083)

-9.152*** (1.222)

-11.697*** (0.617)

-9.543*** (0.696)

-13.327*** (1.003)

-9.090*** (1.054)

-16.574*** (1.892)

-9.090*** (1.226)

-10.451*** (1.253)

Observations

122 73 122 73 122 73 122 73 73

R-squared 0.951 0.981 0.984 0.994 Wald chi2= 2343.35

Wald chi2= 3699.21

Within= 0.918; between= 0.936; overall= 0.934. Wald chi2=1153.63

Within= 0.652; between= 0.971; overall= 0.981. Wald chi2= 2736.41

Endogeneity: no.

-Endogeneity: yes. - Underidentification: no. - Overidentification: yes. Instrument: not weak.

Adjusted R2

0.941 0.974 0.981 0.992

36

Table 18: Summaries of results of determinants of non-life insurance demand before and during the GFC

Variables Hypothesis All sample economies Developed economies Developing economies

2000-2006 2007-2011

(crisis)

2000-2006 2007-2011

(srisis)

2000-2006 2007-2011

(crisis)

lnEFI + + + + + + +

lnGDPP + + + + + + +

BANK + + + + - + +

EDU + + Insig. - + + -

URB + + + (insig.) - - + +

INDUL - - (insig.) - + Insig. - -

IDV + Insig. - + + + -

LTO - - - + Insig. - -

MAS - - (insig.) - - - insig. -

UAI + Insig. -(insig.) + + - -

PDI - + + + + (Insig.) Insig. +

HPM + + Insig. - - + +

ISL - - - - -

CL + + + + - - +

37

Table 19: List of developed and developing economies

Developed countries Developing countries

Australia Switzerland Argentina

Austria Trinidad Bangladesh

Belgium United Arab Brazil

Canada United Kingdom Bulgaria

Czech Rep. United States Chile (*)

Denmark China

Estonia Colombia

Finland Costa Rica

France Croatia (*)

Germany Ecuador

Greece El Salvador

Hong kong Guatemala

Hungary India

Ireland Indonesia

Italy Iran

Japan Malaysia

Latvia Mexico

Lithuania Morocco

Luxembourg Nigeria

Malta Pakistan

Netherlands Panama

New Zealand Peru

Norway Philippines

Poland Romania

Portugal Russia (*)

Singapore South Africa

Slovakia Thailand

Slovenia Turkey

South Korea Uruguay (*)

Spain Venezuela

Sweden Vietnam

Source: Developing regions of the World Bank and World economic outlook 2014 of IMF

Note: (*): Countries, which are not in list of developing regions of World Bank, are in the list of World economic outlook 2014 of IMF.

global financial crisis, economic freedom, and the demand for … · 2015. 6. 25. · updated data...

Documents