The invisible hand of God? : an update of the relation between religion and economic attitudes

(1)

1

The Invisible Hand of God?

An update of the relation between religion and economic attitudes

Nathalie van Baren

Abstract

The relationship between culture and economic development has been tested before, even specifically when it comes to religion. But no consensus has been reached partly because different definitions of religion are being used. Guiso et al. (2002) used the WVS to establish a relationship between religion and economic attitudes. When replicating their results, I found they had some coding errors which gave wrong significance levels and thus different results. I extended the dataset and found that differences in behavior between religious people and non-religious people are becoming smaller. The correlation between religion or atheism and other economic values is undeniable but more ambiguous than presented by Guiso et al.

Thesis submitted to obtain the degree of:

MASTER OF SCIENCE IN ECONOMICS Development Economics

15 ECTS

University of Amsterdam 2016-2017

Student number: 10840842 Supervisor: Prof. Dr. E.J.S. Plug

(2)

2

Statement of Originality

This document is written by Nathalie van Baren, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Acknowledgements

First of all, my parents and grandparents have done everything in their power to guarantee I was be able to pursue my academic career as I preferred, which calls for a very special thanks. Secondly, I would like to thank my supervisor Erik Plug for his understanding of my personal situation, as I balanced working and studying. He gave me the liberty to achieve the last step of my Master degree on my own terms. Since working fulltime and studying is not an easy combination, I want to thank my colleague and friend Nicolas for believing in my capacities and his much appreciated support. Furthermore, I would like to thank Jesse, as he has been my personal empirical advisor and the one who understands my statistical humor like no one else does. In the same breath, I also thank Stef and Tessa, who completed my experience of being an economics student in every way. Additionally I would like to mention Maarten, whose ideas made an achievement on a more personal level possible and who should have been acknowledged earlier. Last but not least my partner Bavo, who has been there with me through the late evenings of worrying, doubts and frustration, even though technically he had no other choice. He deserves special gratitude for motivating me by example but also because of the upgrade I experience in all aspects of life because of his presence.

(3)

3

1 Introduction

As refugees are entering Europe during the European refugee crisis, cultural diversity is growing and the percentage of religious attendance is on the rise again. This is the result of growing fundamentalism in the Middle-East while in the meantime, the western world was experiencing a decline in religiosity. Religious demographics have always been changing but this time, due to the asylum crisis, Europe is importing religion instead of spreading it. What could be the economic effect of an increase in religious people? Previous research found a definite relationship between culture and economic development, but linked with institutions inherent to these cultures. So what might happen if you remove the people from developing countries with bad institutions and insert them in developed countries with good institutions? Will these religious immigrants thrive or will they experience less prosperity than natives? Economic behavior holds the answer to this question. Imagine a country where discrimination and coincidence is absent and everybody starts with equal opportunities, one’s behavior is the only explanatory variable for their prosperity. But what could determine one’s behavior? Keeping in mind this latest wave of migration, it is excellent timing to verify if religious people have different economic behavior, and whether that is good or bad news for the economy?

If someone spends time on practicing religion, they can’t spend that same time working, which probably has a negative effect on the economy. So hypothetically, countries where religiosity has a high level of practice, will experience less economic development. The same result should be found on micro-level. Additionally, as religiosity is on the decline in mostly western countries and fundamentalism on the rise, the hypothesis here is that the difference in attitude is becoming larger than before.

Guiso, Sapienza and Zingales (2002) found a correlation between religious attendance and values proven to have a positive correlation on the economy (see 2.2). These values include trust, women, the government, the law, the market and its fairness, and thrift. They used household level data constructed by the first three waves of the World Value Surveys, so between 1980 and 2000. Although the focus has since shifted to randomized experiments, the use of micro-level data is consistent with more recent movements within development economics, Which is why this article, titled “People’s Opium? Religion and economic

(5)

5

attitudes” forms the basis of this thesis. After a literature review on the research between culture and economic development and on the values proven to be positively related to economic growth, Guiso et al. ‘s research will be updated. First, their original correlations will be replicated. Afterwards, their approach will be optimized and recent data is added. Finally, a comparison between both timeframes will be given, in order to determine whether the correlation between religion and economic attitudes has changed. In the results section it becomes clear that the optimized version mostly changes significance levels, but that a correlation between religion and economic attitudes stays indisputable. This correlation is again confirmed after enlarging the dataset, but is ambiguous, dependent on the definition of religion and admissive to change.

(6)

6

2 Literature review

2.1 From Smith and Weber to Guiso, Sapienza and Zingales

One of the founding fathers of economics, Adam Smith, was a pioneer in exploring the relation between economics and religion. According to Smith claimed that religious beliefs and activities are rational choices, observably responding to benefits and costs. But it is doubtable that everybody constructed their own cost-benefit-analysis before turning to a religion. Anderson (1988) claims that Smith’s insights was that he viewed religious participation as a rational way to increase human capital of the individual. For some religions this is more true than others, which is where Max Weber (1904) comes in the picture.

One of the founding fathers of sociology, Max Weber, also had a large impact on further research about the relationship between religion and economics. In a nutshell, he argues in “The Protestant Ethic and the Spirit of Capitalism” that the Protestant ethic encourages its followers to work hard and invest their money in order to grow even more. This is seen as the ideology of the capitalistic system. It is almost unnecessary to mention that this hypothesis received a lot of critique and has been put to the test more than once. Not all papers come to the same reasoning why but empirical research, shows that Weber was indeed correct when observing Protestant regions to be more prosperous than Catholic regions. Grier (1997) was one of the first to empirically test his Weber’s hypothesis. British ex-colonies perform better than their Spanish and French colleagues and Grier wanted to know whether religion could help explain this situation. Although he found out that GDP growth and per capita income levels are positively and significantly correlated with Protestantism’s growth rate, Grier points out that it is only one of many factors. According to Fanfani (1936), the separation of church and state is the real driving force behind economic growth, because all religion has a negative effect on development. This is also tested later in this paper, in form of values such as trust in the law and the government.

According to Blum and Dudley (2001), an explanatory variable of Weber’s observation might be that the willingness to cooperate with strangers appears larger in the Protestant culture, another one of the values tested later. More openness to strangers could generate a profound enlargement of trading networks and therefore more economic prosperity. An alternative hypothesis by Becker and Woessmann (2009)proposes that education is the underlying cause

(7)

7

of blossoming/booming protestant cultures. Protestant followers were more literate (in order to read the Gospel themselves), which suggests that the Protestant ethic might not be of such an influence after all and Weber’s outcome was right but the channel was wrong. Tabellini (2010) also found that literacy rates and the quality of political institutions strongly correlated regional economic development when searching for a correlation between culture and economic development. Literacy rate was one of the instruments used in order to isolate the exogenous variation of culture. Tabellini found that culture has a causal effect on economic development. He used values and beliefs, such as trust and respect for others, and confidence in individual self-determination, as indicators of culture. A vision that is shared by Gorodnichenko and Roland (2011) who argued that individualism is the central cultural dimension that matters for economic growth.

Education and the separation of church and state are only two possible reasons why some countries experienced more economic development, more explanatory variables are possible. David Landes suggested in his book ‘The Wealth and Poverty of Nations’ (1998) that the success of some nations is mostly dependent on cultural factors above anything else. It is, among others, hard work, thrift, tolerance and honesty that can make the difference. A theory partly supported by Granata et al (1996) who found that a growth theory with an empirical model that incorporates both cultural and economic variables is superior to a theory that incorporates only one of these variables. Complimentary to this theory, Barro (1996) tested what determinants are correlated with economic growth. He found that higher growth rate is correlated by higher initial schooling, next to lower infertility, high life expectancy, low inflation, lower government consumption, better maintenance of rule of law, and improvements in terms of trade. It should be noted that this is a correlation and does not consist of a causation.

Easterly and Levine (1997) did search for a causation of these negative characteristics. They showed that Sub-Saharan Africa suffers from low economic growth as well as low schooling, political instability, high government deficits, etc. partly because of ethnic diversity. The argued that high fragmentation reduces the consensus for public goods, thus creating long-run growth tragedies. Easterly and Levine do not believe that ethnic fragmentation is the only variable but it is only one in a set of many variables. As opposed to ethnic fragmentation, religious fragmentation is not of such explanatory power. According to Montalvo and Reynal-Querol (2002), it is religious polarization that has a significantly negative effect on GDP

(8)

8

growth rate, while religious fragmentation cannot explain such differences. Freedom of religion on the other hand, has a promising effect on prosperity, as estimated by Alon and Chase ( 2005).

Many studies about religion and economic growth have been conducted in different ways, with confusing results, and no consensus. Available literature include studies with macro-level data and micro-macro-level data, sometimes through the form of cross-country observations, or qualitative research. One of the reasons no consensus has been reached yet, is probably due to different definitions of religion. What is religion? When is someone considered religious? How can and should religiousness be measured? All mentioned studies have different approaches of the definition of religion, and are, for clarity, in bolt.

When micro-level research about religion is conducted, the World Value Surveys (further on written as WVS and explained in 3.1) have been used a couple of times, for example by Barro and McCleary In 2003, they published a paper estimating the effect of church attendance and religious beliefs on economic growth, while in 2006, they estimated the other direction. Religious beliefs – belief in hell and heaven – are found to be positively correlated while church attendance is negatively correlated with economic growth. Beliefs are considered an output and church attendance an input of the religious sector. These findings stimulate the idea of a model where beliefs influence personal characteristics that enhance economic performance.

Durlauf et al. (2012) criticized Barro en McCleary’s findings by putting their results to the test. It appears a.s if the findings are replicable but not robust to changes in the baseline model specifications. Therefore, the evidence of religious beliefs as an explanatory variable for cross-country income differences is no longer applicable. Additionally, the original results disappear when replicating Barro’s research with the more recent data, according to Eum (2011).

In 2006, Barro and McCleary published an additional paper, this time about the effect of the economy on religion. According to their research, economic development (per capita GDP) tends to reduce religiosity, depending on the dimension of the development. E.g. Religious measurements were negatively related to urbanization and positively related to education. Participation in religious activities was negatively related to life expectancy but positively related to the fraction of the population under 15 years of age. When keeping these variables

(9)

9

fixed, religion and GDP per capita are basically unrelated, they find, contradicting their previous work.

An even more micro-level approach is to calculate the impact of religion on economic attitudes toward others, women, the government, legal rules, thriftiness, the market and its fairness. While Barro and McCleary don’t use country-fixed effects, Guiso, Sapienza and Zingales (2002) do so in their micro-level research, also using the WVS. Their research finds religion on average to be related with economic attitudes sometimes beneficial to higher per capita income and growth and sometimes detrimental. The intensity of attending religious services serves as the independent religious variable, a completely different approach when comparing to Barro’s believe in heaven and hell. Additionally to this independent variable, Guiso et al. introduce control variables such as atheism and being raised religiously.

Guiso et al. received criticism from Keely (2002), who stated to disagree on some critical concepts of the research. First, Guiso et al. explore “the correlation between the cultural traits of religious belief and practice, and a set of other beliefs that may correspond to certain economic decisions. The results are at best, suggestive of causality, and certainly do not establish causality.” Additionally it is pointed out how religious beliefs depend on religious teachings and that religious practices is measured by religious attendance to determine the intensity of beliefs. This is an abstract way of looking at religion, one should keep in mind that religious institutions could also be important, especially when considering possible policy actions. Also, more attention for country differences is welcome. To conclude, according to Keely, further research should focus more on institutions rather than beliefs, in order to work towards policy prescriptions to improve countries’ economic performance.

2.2 Values with an economic impact?

Trust, cooperation, tolerance, openness to strangers, … These are values mentioned before and each time linked to more economic development. First Knack and Keefer (1997) used trust and civic norms to measure social capital, using the WVS, and found that it mattered for economic performance. Trust and civic norms are stronger in nations that experience higher and more equal incomes, higher levels of education and lower levels of ethnic fragmentation. Knack and Zak (2001) found that low levels of trust reduce the rate of investment, which is obviously not wanted when one aims for economic growth. Beugelsdijk et al. (2004) did a

(10)

10

robustness check for both articles and found the latter to be highly robust in terms of significance.

Guiso et al (2002) only looked at economic theory to determine the value about women and believed that a country’s higher endowment of labor will benefit economic growth and development. Actually, theory predicts that the correlation of female employment and economic growth is U-shaped. This means that not in all cases more female employment is correlated with better economic growth, due to income and substitution effects. Lechman and Kaur (2015) established the U-shaped correlation, except for low income countries. Tsani et al. (2013) also empirically confirmed this theory but added some other insights. They found that lowering region-specific barriers has a considerable positive impact on growth. In general, higher female labor participation rates have a positive impact on growth. When a lower participation rate is a result of income developments, it could result in marginally lower economic growth.

When it comes to government, it has already been mentioned that high government consumption and high deficits are correlated with low economic growth. The same can be said for political instability, as it is detrimental for investments according to Alesina and Perotti (1995). Additionally, ‘bad’ governments are linked to bad institutions such as the educational system, which has also been proven to be essential for economic development. To stay in theme of religion, La porta et al. (1999) found evidence that a correlation is present between inferior government performance and high proportions of Catholics or Muslims. In order to measure government performance they used measurements of government intervention, efficiency of the public sector, the size of the government, public good provision and political freedom. Additionally, countries that are poor, close to the equator, experience ethnolinguistic fragmentation and use French law, also have lower government performance. Countries where the protection of investors is lower, experience smaller capital markets. La Porta et al. (1997) used the character of legal rules and the quality of law enforcement as indicators of protection, and found that it is applicable in equity as well as debt markets. It is of no surprise that corruption lowers investment and thus economic growth, but was confirmed in Mauro’s (1995) article ‘Corruption and Growth’. He used cross-country data with indicators of corruption, bureaucracy, efficiency of judicial system and political instability. Mauro claims with caution that bureaucratic efficiency could cause high investment and growth. But he does believe that red tape could be of the same explanatory

(11)

11

value as political instability when it comes to economic growth. Acemoglu and Johnson (2005) found that property rights institutions, that protects citizens from the government and elites, have a positive effect on long-run economic growth, investments and financial development. When it comes to labor, Botero et al (2004) found that heavier regulation of labor is associated with lower labor force participation and higher unemployment.

Easton and Walker (1997) provide evidence that pro-market policies favor growth. Pro-market policies is broad concept. A lot of studies rather mention economic freedom and measure e.g. inflation, taxation, economic structure and international trade. This is the specific case for Gwartney et al. (1999) who found empirical evidence that economic freedom is a significant determinant of economic growth. Critics of this approach, De Haan et al. (2006), find indications that liberalization stimulates economic growth. It is suggested to be a causation because less evidence is found that higher economic growth causes liberalization of the market.

Thrift is an important factor in the modern theory of growth, claimed Guiso et al (2002). Since evidence is better, Granato et al. (1996) empirically found that cultural attitudes towards achievement and thrift have a positive effect on economic growth.

Besides these values, many other are possible to have an effect on economic growth. Perhaps more focus could go to institutions, as Acemoglu et al (2001) found them to have explanatory power when it comes to differences in income per capita. Furthermore, Alesina and Giuliano (2010) found strict family ties to have an influence on economic outcomes. Many other forms of values with an impact on economic behavior and therefore economic growth could be tested, but where should the line be drawn? For simplicity purposes, this empirical research will keep the values Guiso et al. (2002) used, update them where necessary and compare timeframes.

(12)

12

3 Data and Methodology

3.1 Data Collection

As mentioned before, data is originated from the World Value Surveys, commonly known as trustworthy and representative. This project is led by Ronald Ingelhart of the Institute for Social Research of the University of Michigan. One of the aims of the institute is to conduct a cultural map of the world. The data currently consists of six waves, (1981-1984, 1990-1993, 1995-1998, 1999-2004, 2005-2009, 2010-2014) and is conducted in almost hundred countries. Each wave contains representative national household surveys with, next to demographical information, very specific questions about economic characteristics, religion, political preferences and attitudes. EVS, European Value Surveys, is a similar project with the same outset but was not incorporated in the WVS until 2008. Therefore it was necessary to integrate the additional EVS data in the WVS data in order to create a more complete dataset. Some countries, such as Spain and old Yugoslavia, have their own regional WVS, which explains the differences in number of interviewees between Guiso et al. and its reproduction. This regional data is not as freely available as the WVS and EVS, and was not included.

3.1.1 Independent variables: religion

As religion is difficult to define and has many aspects, one variable does not provide enough information for a complete image of religion so Guiso et al. used multiple variables. Following questions were used to form dependent variables: “Where you brought up religiously at home?”, “Apart from weddings, funerals, and christenings, about how often do you attend religious services these days?”, “Are you a religious person, not a religious person or an atheist?” The default of Guiso et al. is a person who is not religious and not raised religious, who does not attend religious services and is not an atheist. Attendance of religious services is split into two variables: actively religious and currently religious. Actively religious is defined as attendance of religious service at least once a week, while going at least once a year -outside of special occasions- is considered currently religious.

This use of variables raises some concerns, as someone could be not-religious and still attend services because of e.g. peer pressure. On the other hand, someone could consider themselves very religious and still not attend services, because that is not the way they experience their

(13)

13

religion. Thus the distinction between being religious and attending services is not formed in their approach. An additional challenge in reproducing Guiso et al. is that the question “Were you brought up religiously at home?” was only asked during the second and third wave. In order to compare the two timeframes, it was necessary to change this aspect. More focus was put on the question if someone identifies themselves as religious or not.

In order to try to isolate of the characteristics of being religious, individual characteristics such as perceived health, age, gender, education, income, subjective social status and country-fixed effects are controlled for. Again, they should not be seen as causal but only confirm a correlation.

When replicating Guiso et al,. it dawned that some variables were not correctly identified. For example: social status could have a form between 1 (low) and 5 (high social status), but in their descriptive statistics (TABLE 1) also 0 was observed. This made it clear that for many variables, when values were missing, they were treated as 0 instead. As this is not a proper way to treat data, it was corrected in the optimized replication of Guiso et al. Additionally, the age of graduating was chosen as parameter for education, which was changed to the level of education in the optimized research, because of interpretation issues as incorrect values.

3.1.2 Dependent variables: economic attitudes

The relation between being religious and attitudes towards women, the law, trust and cooperation, the government, thriftiness, the market and its fairness, is being measured. For every one of those subjects some questions are asked, which are treated as the dependent variables. “Do you think that women should have children in order to be fulfilled or is this not necessary?” is an example of such question meant to measure attitude towards women. Most questions are answered with 0 or 1, but some answers are coded 1 to 4, to represent e.g. a higher degree of agreement. All different questions and possible answers are explained in table 1D.

It should be noted that when Guiso et al. did their research, those subjects were found to be positively correlated with economic progress. Since then, additional research may have changed that perceptive. But because the purpose of this research is mainly to compare effects in time, the same assumptions as Guiso et al. were used, so the dependent variables stayed the

(14)

14

same during all research. However, there is one exception for “Poor Are Lazy”, since it is not clear which question was actually asked, as there are multiple similar questions in WVS.

3.2 Empirical Strategy

For every single variable, the same structure, Ordinary Least Squares (OLS) is used. As example, the same dependent variable as above is used: “Do you think that women should have children in order to be fulfilled or is this not necessary?” Below the theoretical framework used by Guiso et al.:

𝐶ℎ𝑖𝑙𝑑𝑟𝑒𝑛𝑁𝑒𝑐𝑒𝑠𝑠𝑎𝑟𝑦

= 𝛽₀+ 𝛽₁𝐻𝑒𝑎𝑙𝑡ℎ + 𝛽₂𝑀𝑎𝑙𝑒 + 𝛽₃𝐴𝑔𝑒 + 𝛽₄𝐴𝑔𝑒𝐸𝑑𝑢𝑐𝑎𝑡𝑖𝑜𝑛 + 𝛽₅𝐼𝑛𝑐𝑜𝑚𝑒 + 𝛽6𝑆𝑜𝑐𝑖𝑎𝑙𝐶𝑙𝑎𝑠𝑠 + 𝛽7𝐴𝑡ℎ𝑒𝑖𝑠𝑡 + 𝛽8𝑅𝑎𝑖𝑠𝑒𝑑𝑅𝑒𝑙𝑖𝑔𝑖𝑜𝑢𝑠𝑙𝑦

+ 𝛽9𝐶𝑢𝑟𝑟𝑒𝑛𝑡𝑙𝑦𝑅𝑒𝑙𝑖𝑔𝑖𝑜𝑢𝑠 + 𝛽10𝐴𝑐𝑡𝑖𝑣𝑒𝑙𝑦𝑅𝑒𝑙𝑖𝑔𝑖𝑜𝑢𝑠 + 𝛽11𝐶𝑜𝑢𝑛𝑡𝑟𝑦 + 𝛽₁₂𝑌𝑒𝑎𝑟 + 𝜀

As mentioned before, some variables are later changed in order to better compare between timeframes; “AgeEducation” changes to “Education”, “SocialClass” is left out and “RaisedReligiously” becomes “ReligiousPerson”. “Year” is also later changed to “Wave”.

3.3 Methodology

The first step is to replicate the research Guiso et al. conducted, therefore it is necessary to replicate their descriptive statistics. Please find in the Appendix for the first three waves TABLE 1;

PANEL A: Summary statistics of religious beliefs by country;

PANEL B: Report of the distribution of population by Religious belief and country;

PANEL C: Summary Statistics of the Intensity of Religious Belief by Religious Denomination;

PANEL D: Summary statistics of the dependent variables;

(15)

15

The next step is to replicate the regressions of Guiso et al. Since there are some difference in the available data, it is possible to find some inconsistencies when comparing Guiso’s work with its replication. However, for the most part their empirical outcome is similar so no further attention has been given to these differences.

Please find in the Appendix TABLE 2;

PANEL A: Attitudes toward others and the government

PANEL B: Attitudes towards women

PANEL C: Attitudes towards legal rules

PANEL D: Attitudes toward the market

PANEL E: Attitudes toward thriftiness and market’s fairness

Thirdly, Guiso et al.’s work is optimized, meaning missing values are treated as missing and not as zero. Also, some variables are changed because of misidentified or missing data. Please find in the Appendix TABLE 3, the starting point of comparing between timeframes;

PANEL A’: Attitudes toward others and the government PANEL B’: Attitudes towards women

PANEL C’: Attitudes towards legal rules PANEL D’: Attitudes toward the market

PANEL E’: Attitudes toward thriftiness and market’s fairness

The descriptive statistics of all waves, so all data between 1981 and 2014, can be found in TABLE 4. It is comparable with the first table and gives an interesting overview of how religion has changed over the years.

PANEL A: Summary statistics of religious beliefs by country;

PANEL B: Report of the distribution of population by Religious belief and country;

(16)

16

PANEL D: Summary statistics of the dependent variables;

PANEL E: Demographic characteristics of the respondents.

Finally, in TABLE 6, the same theoretical framework as table 3, but with all waves of data. In the next section, the differences between table 3 and table 6 will be discussed, using table 5.

PANEL A’’: Attitudes toward others and the government PANEL B’’: Attitudes towards women

PANEL C’’: Attitudes towards legal rules PANEL D’’: Attitudes toward the market

(17)

17

4 Results

4.1 The optimized research from Guiso, Sapienza, and Zingales

As explained before, Guiso et al. treated missing values as zero, which could under- or over-estimate the correlation. Additionally, the variable “raised religious” was only asked in the second and third wave, so the researched time period does not start in 1981 but in 1990. When replacing this variable and changing the missing values back to missing, mostly significance levels change. A detail per subject is available below.

Attitudes towards others and the government.

The update regression shows that the correlation between atheists and economic attitudes is no longer as significant as in Guiso’s outcome. Atheists do trust other people more than non-religious people, and much more than non-religious people. But non-religious people have more trust in the government and the services they provide (army, police). That relation has not changed, whether checking for being raised religious or identifying as religious.

Attitudes towards women

Again, atheists have less significant results than before, but for the few significant results that remain, it is possible to observe that atheists believe more in equality among genders. Around the same results can be found for religious upbringing as for being religious. The results concerning being currently religious are more significant than before, while the results about actively religious people are less significant. For the remaining significant results, the relation between currently religious and this unequal attitude towards women is stronger. The final relation of someone who attends religious services on a weekly basis stays the same, since the additional effect of being actively religious goes down a bit.

Attitudes towards legal rules

For the third time, the significance of the correlation between atheists and economic attitudes - less ethical behavior in this case - goes down. When a relation is still significant, it is smaller than in Guiso’s research. The results of someone claiming to be religious or being raised religious are comparable; they both behave more ethical than non-religious people. Not much change can be observed when checking the behavior of someone who attends a service at least

(18)

18

once a year. But the relation between actively going to those services and behavior becomes smaller. To conclude, just because you sometimes go to church e.g., it does not mean you are more ethical, while going at least one week does.

Attitudes towards the market

The significance of atheists stays the same on this subject; according to the regression outcome they are more open to a free, competitive market where inequality of income is not necessary considered a bad thing. This could also been said about people who were brought up religious, according to Guiso’s work. However, in this new method of estimating the correlation, being a religious person makes you less open to this competitive and private market. This is the first real change that can be observed when comparing both methods, not bringing significance levels into account. Additionally, currently religious people generate less significant results and active religious people stay insignificant on this matter.

Attitudes towards thriftiness and market’s fairness

For the final time, atheists no longer show significant results in economic attitudes, contradicting Guiso’s findings. When someone is raised religious, on average they believe more in hard work and that thrift should be encouraged than non-religious people, according to Guiso et al. Now, when you identify as being religious, you still believe that hard work improves life but also that wealth can grow for everyone. Other variables di not generated a significant correlation. Similarly, the relation between currently religious people and economic attitudes becomes insignificant. This is also the case for most variables and being actively religious. However, when someone attends services at least weekly, they believe, on average, that thrift should not be encouraged but that hard work improves life.

4.2 The overall correlation between religion and economic attitudes

It is possible to reconfirm the existence of a correlation between religion and economic attitudes, and just like Guiso et al. found, it is ambiguous. However, this relation has changed over the years, which will be discussed after an evaluation of the relation of religion with every specific subject of those economic attitudes.

(19)

19

People who identify themselves as religious appear to have more trust in the government, the army and the police. The more someone goes to religious services, the more trust in the government (and its services) they seem to have. The difference between not attending services and attending at least once a year is larger than the difference whether someone goes once a year or once a week. When it comes to trusting people in general, actively religious and atheists are more trusting than not-religious people and religious people who don’t go to services. No significant data is available about a relation between religion and tolerance towards other races and immigrants.

When observing attitudes towards women, a very clear trend can be observed. Atheists believe men and women to be more equal than all religious people and non-religious people believe. A very easy conclusion can be formed here. Atheists are more open to women on the labor market and thus more receptive to economic growth caused by working women. The more intense someone practices their religion, the less open they are to women and to economic development.

Another clear trend can be observed when looking at attitudes towards legal rules. It is obvious that the more religious someone is, the more trust they have in the legal system and the more skeptical they are about not paying what should be paid. Atheists seem to have less ethical problems with cheating on taxes or avoiding a fare on public transport. One bizarre finding within these results is the idea that people who go to a religious service at least once a year but less than once a week, are more willing to claim government benefits they are not entitled to. This correlation becomes negative when looking at actively religious people. Attitudes towards the market

The empirical findings are not so often significant when looking at attitudes towards the market. Thus no significant conclusions can be formed about religious people. On the other hand, it is clear that atheists have a vision about the market that is known to have a less positive relation to economic growth than the default person.

(20)

20

Contesting earlier findings, when checking for values such as hard work and wealth, atheists don’t have an attitude that corresponds with more economic growth. Religious people are more focused on hard work and the idea that when wealth grows, it can grow for everybody than non-religious or atheist people are. One may find this an interesting result, keeping in mind other findings.

4.3 The difference between timeframes and thus change of the correlation

Next, the difference between waves 1-3 and waves 1-6 is shown for all subjects. When a value is negative, it means that there is less agreement with the statement presented than before. The values are always the difference towards the values of the default, a non-atheist and non-religious person. The variable ‘wave’ is added in the table to give an idea about the change in the default, since their values also change over time. The closer a value is to zero, the less it differs from the values shared by the default. When it comes to the difference between the two time periods, it is thus the difference in perception towards the default. E.g. intolerance towards other races has been going down for currently religious people, but it is possible that everybody has become less tolerant. The wave variable confirms this finding, since intolerance is on average, going up per wave. This makes it difficult to correctly measure the real change of economic behavior amongst religious people and service attendees, especially when the wave variable is not significant for all timeframes. However, the perceptive change towards the earlier timeframe and default is clearly measurable and discussed below. Please keep in mind that coefficients for actively religious people should be added to the one for currently religious people when comparing to the default, as it measures the additional change in value and not the whole. This sum of coefficients should than probably be added to the one for “religious person”, when you reading values for a person who considers themselves religious and goes to a service at least once a week.

(21)

21

***= significant at 1% level; **= significant at 5% level, *= significant at 10% level

Trust in people has been going down in general during the years. What can be observed is that trust in others is no longer much more present in atheists and people who attend services than in people who don’t attend services. People who don’t attend services but consider themselves religious seem to have more trust now than before. On a positive note, on average, intolerance has also been going down when checking for difference between waves. Significance levels on intolerance are missing when it comes to differences between levels of religion, so change in them cannot be discussed.

While trust in the army was lowering with each wave last century, this is no longer visible in the extended dataset. The observation that religious people, whether or not they went to services, had more faith in the army than non-religious people is still true but in a smaller extent. When it comes to trust in the police, the same trend is visible but only for people who attend services.

(22)

22

(23)

23

Not all questions about women that are researched by Guiso et al. are equally feminist – leaving the choice to the woman- which makes it difficult to know why someone agrees or disagrees. The question about scarce jobs and university are the most relevant here.

Equality for women has been going up year after year, and the difference between atheist and non-religious people has been going down. Although it cannot be deducted from the numbers here, it is probably due to non-religious people becoming more feminist than atheists become less feminist. It is visible that religious people and service-attendees are also going into the direction of feminists, since the difference towards the default is smaller than before. The exact same trend is observed when looking at values about university education.

Trust towards the legal system is still larger when comparing the values from religious people in all forms towards the default. But the difference between both has become smaller after adding more recent data. Where atheists used to have much less trust in the legal system than non-religious non-atheist people, are significant correlations no longer visible. This could be because atheists regained trust or because the default lowered their trust. Unfortunately, that is not visible from the data. When it comes to ethics on legal rules, religious people and service attendees are still more ethical than the default and atheists, but overall are the differences smaller than before.

Attitudes towards the market

Atheists have become more open to competitive, private markets, or non-atheist non-religious people are becoming less open to them, or both. Act is that the attitudes between atheists and the default is becoming smaller. The same trend can be observed between religious people of all kinds and the default, but there are relatively little correlations significant.

(24)

24

When extending the dataset, significance levels rise for values about the market and thrift. Comparing them with non-significant values is difficult, if not useless. The rare times a value is significant in both datasets, it is again observed that the difference to the default is smaller than before.

To conclude, it seems that in most cases, significance levels rise and the difference to the default diminishes. It is possible to conclude from these observations that behavior of religious people is becoming more similar to that of non-religious and atheist people. Unfortunately, this does not mean that everybody has developed attitudes that on average, correspond more with better economic outcomes.

(25)

25

5 Conclusion

The correlation between religion, in multiple forms, and economic attitudes is undeniable. What Guiso, Sapienza, Zingales (2002) found is still applicable, but some adjustments should be added. Since they treated missing values as zero and misidentified other variables, not all correlations they found were significant.

Although Guiso et al. found a correlation between religion and intolerance, it is no longer visible in the corrected version or in new data. Religious people who attend services are still more trusting of the government and more trusting of other people, but in lesser extent than in the smaller dataset. The approach towards women on the labor market is a predictable one; the less religious someone is, the more they believe women should work and that they are equal to men. This has been changing for the better lately, but the average conviction is still indisputable. On the other hand, on average, being religious means being more ethical towards legal rules. Atheists are more willing to cheat on taxes or accept bribes. However, all these correlations are evolving towards the default, a non-religious and non-atheist person who does not attend religious services. When looking at economic attitudes towards the market, no significant correlation becomes visible for religious people. Atheists seem to be less open to a competitive, private market. Finally, a correlation between hard work and religion is present but people who attend religious services weekly believe that thrift should be discouraged. To conclude, the correlation between both factors is ambiguous but it seems that on most subjects, atheists have an attitude that corresponds with more economic growth. On other subjects religious people show, on average, more of this economic attitude. What can be concluded beyond any doubt is that the difference in behavior within people of different levels of religiosity is smaller now the dataset is enlarged with more and recent data. The hypothesis introduced in the beginning of this thesis is thus rejected. However, it is possible that fundamentalism is not (yet) visible in the dataset, or too small to be detected. It should be kept in mind that more literature is and will become available on these economic attitudes, which could alter the conclusion of the findings. As suggested by Keely (2002), more attention should go to the institutional aspect of religion, which might have an impact on economic attitudes and economic growth. Guiso et al. also looked more into detail and found differences in attitude between religions and country demographics, something that is outside the scope of this paper. Further research could exist of this more detailed approach per

(26)

26

religion but hopefully without coding errors, misidentified variables and with more data, as has been the outset of this study. Another possibility is to try to find or form a variable for fundamentalism, in order to test the hypothesis if the difference in economic behavior between fundamentalists and non-fundamentalists is growing.

For policy implications, it should be noted that the relationship of religion and economic attitude is receptive to change. But it is much more important to know that not one type of religious characteristics - whether it is atheism, no religion, moderate practice or highly religious people – holds the key to ‘perfect’ economic behavior.

(27)

27

6 Sources

Acemoglu, & Johnson. (2005). Unbundling Institutions. Journal of Political Economy.

acemoglu, Johnson, & Robinson. (2001). The colonial origins of comparative development.

Alesina, & Perotti. (1996). Income Distribution, Political instability, and Investment . European

Economic Review, 1203-1228.

Alesina, A., & Giuliano, P. (2011). Family Ties and Participation. Journal of the European Economic

Association, 817-839.

Alon, I., & Chase, G. (2005). Religious Freedom and Economic Prosperity. The Cato Journal, 399-406.

Anderson, G. M. (1988, October). Mr. Smith and the Preachers. Journal of Political Economy, pp. 1066-1088.

Barro. (1996). Determinants of Economic Growth: a cross-country empirical study.

Barro, R. J., & McCleary, R. M. (2003). Religion and Economic Growth across countries. American

Sociological Review, 760-781.

Becker, S., & Woessmann, L. (2009). Was Weber Wrong? A Human Capital Theory of Protestant Economic History. Quarterly Journal of Economics, 531-596.

Beugelsdijk, Groot, d., & Schaik. (2004). trust and Economimc Growth: A robustness analysis. Oxford

Economic papers, 118-134.

Blum, U., & Dudley, L. (2001). Religion and Economic Growth; Was Weber Right? Journal of

Evolutionary Economics, 207-230.

Botero, Djankov, & Porta, L. (2004). The Regulation of Labor. The Quarterly Journal of Economics.

Cole, M., & Block, W. E. (2013). Adam Smith on Religion and Economics. Journal of Prices & Markets, pp. 24-35.

Durlauf, S. N., Kourtellos, A., & Tan, C. M. (2012). Is God in the Details? A Reexamination of the Role of Religion in Economic Growth. Journal of Applied Econometrics, 1059-1075.

Easterly, & Levine. (1997). Africa's Growth Tragedy. Quarterly Journal of Economics, 1203-1250.

Easton, & Walker. (1997). Income, Growth, and Economic Freedom.

Eum, W. (2011). Religion and Economic Development - A study on Religious variables influencinf

GDP growth over countries. Berkeley: University of California.

Fanfani. (1936). Catholicism, Protestantism and capitalism.

Granata, I. L. (1996). The effect of cultural values on economic development: theory, hypotheses, and some empirical tests. American Journal of Political Science, 607-631.

Grier, R. (1997). The Effect of Religion on Economic Development: A Cross National Study of 63 former colonies. KYKLOS, 47-62.

(28)

28

Guiso, L., Sapienza, P., & Zingales, L. (2003). People's Opium? Religion and Economic Attitudes.

Journal of Monetary Economics, 225-282.

Gwartney, & Lawson. (1996). The concept and measurement of Economic Freedom.

Keely, L. (2002). Comment on: People's Opium? Religion and economic attitudes. Journal of Monetary

Economics, vol 50(1), 283-287.

Knack, & Keefer. (1997). Does social capital have an economic pay-off? Quarterly Journal of

Economics, 1251-1288.

Knack, & Zack. (2001). Trust and Growth. The Economic Journal, 295-321.

Landes, D. (1998). The Wealth and Poverty of Nations: Why some are so rich and some are so poor.

Lechman, & Kaur. (2015). Economic Growth and the Female Labour participation.

Lian, B., & Oneal, J. R. (1997). Cultural Diversity and Economic Development: A Cross-National Study of 98 Countries. Economic Development and Cultural CHange, 61-77.

Lopez-de-Silanes, L. P., & Shleifer. (1999). The quality of government. Oxford University Press.

Mauro. (1995). Corruption and Growth. Quarterly Journal of Economics, 681-712.

McCleary, R. M. (2008, May). Religion and Economic Development. Policy Review, pp. 45-57.

McCleary, R. M., & Barro, R. J. (2006). Religion and Political Economy in an International panel.

Journal for the Scientific Study of Religion, 149-175.

Montalvo, J. G., & Reynal-Querol, M. (2003). Religious Polarization and Economic Development.

Economic Letters, 201-210.

Porta, L., Lopez-de-silanes, Shleifer, & Vishny. (1997). legal determinants of external finance.

Roland, G. a. (2011). Which dimensions matter for long-run growth? American Economic Review, 492-498.

Smith, A. (1759). Theory of Moral Sentiments.

Tabellini. (2010). Culture and Institutions: economic development in the regions of europe. The

Journal of the European Economic Association, 677-716.

Tsani. (2013). Female labour force participation and economic growth. Economic Letters, 323-328.

Wang, Q., & Lin, X. (2014). Does Religious beliefs affect economic growth? Evidence from provincial-level panel data. China Economic Review, 277-287.

(29)

29 Country Raised religiously

at home scriptie: When missing = 0 Scriptie: missing = missing Goes to church at least once a year Scriptie: When missing = 0 Scriptie: Missing = Missing Argentina 54.8 54.8 81.38 59.88 56.42 71.59 Armenia 23.8 23.8 25.31 72.75 72.75 74.65 Australia 64.06 64.06 64.06 46.34 46.34 46.34 Austria 82.05 82.05 82.56 67.67 66.10 67.39 Azerbaijan 59.59 59.59 61.56 67.73 67.73 70.99 Bangladesh 93.05 93.05 93.05 90.03 95.41 95.41 Belarus 25.65 25.65 26.47 46.77 46.77 47.65 Belgium 60.12 60.12 85.30 51.21 51.21 51.31 Bosnia 64.83 71.63 71.63 77.75 75.75 75.75 Brazil 76.08 75.48 75.48 68.41 66.84 66.84 Britain 23.84 24.25 61.23 30.38 42.64 Bulgaria 39.55 39.55 39.80 44.73 44.73 45.07 Chile 79.92 79.92 79.92 63.96 63.96 63.96 Colombia 90.95 90.95 90.98 80.2 80.2 80.44 Croatia 73.75 73.75 75.26 72.16 72.16 72.52 Denmark 19.8 19.8 43.20 42.68 42.68 42.71 Dominican 82.73 82.73 83.94 71.94 71.94 73.35 East Germany 43.71 34.88 Estonia 16.07 16.07 16.21 25.33 25.33 50.39 Finland 29.44 29.44 49.80 53.41 45.38 49.08 France 32.38 32.38 71.16 37.1 37.1 37.24 Georgia 44.43 35.71 36.12 73.81 74.2 74.91 Germany 43.44 54.38 48.91 49.14 Hungary 27.97 27.97 69.04 43.32 39.02 55.90 Iceland 32.29 32.29 76.01 53.22 53.22 53.32 India 91.39 91.39 91.39 89.27 89.27 89.27 Ireland 42.44 42.44 94.10 93.73 93.73 93.86 Italy 56 56 93.55 76.86 76.86 77.36 Japan 14.26 14.26 22.68 71.06 61.03 64.21 Latvia 27.58 27.58 28.07 44.56 44.56 57.48 Lithuania 68.09 68.09 68.67 38.43 38.43 76.74 Macedonia 58.09 58.09 58.09 71.16 71.16 71.16 Malta 44.77 97.96 93.02 93.90 Mexico 50.57 55.02 82.76 82.06 80.39 82.63 Moldova 63.72 63.72 64.97 75.2 75.2 75.51 Montenegro 37.08 37.08 37.39 52.08 52.08 52.08 N. Ireland 41.4 41.4 83.88 81.33 81.33 81.60 Netherlands 32.57 32.57 71.75 53.22 53.22 53.34 Norway 28.27 29.88 43.67 47.98 48.61 48.91 Peru 78.2 78.2 79.58 78.78 78.78 80.85 Poland 96.84 53.18 96.53 92.35 49.78 91.08 Portugal 83.8 83.8 83.94 59.92 59.92 59.92 Puerto Rico 86.77 86.77 87.67 81.19 81.19 81.54 Romania 76.97 76.97 76.97 85.4 85.4 85.64 Russia 15.96 19.27 19.58 27.13 30.49 30.55 S. Africa 68.52 68.52 88.11 48.78 47.56 79.30 Serbia 53.98 53.98 54.15 63.05 63.05 63.05 Slovenia 72.82 72.82 73.04 63.61 63.61 63.83 Spain 75.33 63.63 91.53 57.9 61.21 61.54 Sweden 21.13 21.13 31.72 37.77 37.77 38.44 Switzerland 68.26 68.26 69.84 58.23 58.23 62.08 Turkey 60.64 54.55 68.14 54.68 54.68 65.60 Ukraine 30.74 30.74 32.20 54.46 54.46 56.12 Uruguay 67.3 67.3 67.3 34.8 34.8 34.80 USA 48.28 48.28 82.12 74.57 74.57 74.89 Venezuela 84.5 84.5 85.14 77.92 77.92 77.92 West Germany 43.3 56.34 All countries 53.62 52.10 66.25 59.25 59.13 63.99

7 Appendix

TABLE 1, PANEL A: The original statistics of Guiso et al. are the second and fifth column. The third and sixth column are the reproduction when missing values are treated as zero, and the fourth and seventh column is the reproduction when missing values are treated as missing instead of zero.

(30)

30 Country Goes to church at

least once a week

Scriptie: Missing values = 0 Scriptie: Missing = missing Does not believe in God Scriptie: Missing values = 1 Scriptie: Missing = missing Number of respondents Scriptie Argentina 22.59 22.59 28.66 7.58 7.58 7.84 3086 3086 Armenia 7.30 7.30 7.49 13.2 13.2 14.36 2000 2000 Australia 16.5 16.5 16.50 18.65 18.65 19.90 2048 2048 Austria 25.41 25.41 25.91 11.92 11.92 13.31 1460 1460 Azerbaijan 5.64 5.64 5.92 2.15 2.15 2.21 2002 2002 Bangladesh 63.08 88.92 88.92 1.38 1.38 1.38 1525 1525 Belarus 4.83 4.83 4.92 29.03 29.03 33.52 3107 3107 Belgium 27.53 27.53 27.59 22.28 22.28 24.63 3937 3937 Bosnia 31.00 37.00 37.00 13.92 10.38 10.95 1200 800 Brazil 34.63 33.67 33.67 1.26 1.52 1.52 2931 1782 Britain 9.95 9.80 13.85 12.92 12.90 20.17 3808 3744 Bulgaria 6.32 6.32 6.36 41.12 41.12 46.09 2106 2106 Chile 26.64 26.64 26.64 3.76 3.76 3.80 2500 2500 Colombia 45.68 45.68 45.81 0.45 0.45 0.90 6025 6025 Croatia 22.32 22.32 22.44 18.14 18.14 19.15 1196 1196 Denmark 2.67 2.67 2.67 31.65 31.65 36.25 2212 2212 Dominican 43.41 43.41 44.25 7.19 7.19 7.26 417 417 East Germany 9.68 62.22 2345 Estonia 1.82 1.82 3.63 21.49 21.49 48.23 2029 2029 Finland 3.65 3.65 3.94 10.82 10.82 20.53 2578 2578 France 10.4 10.4 10.44 33.02 33.02 36.21 2202 2202 Georgia 9.53 9.21 9.30 6.56 6.67 6.84 2593 2008 Germany 14.45 14.52 34.21 38.86 6768 Hungary 12.38 12.38 17.74 33.17 33.17 40.11 2463 2463 Iceland 2.33 2.33 2.34 16.94 16.94 18.11 1629 1629 India 49.1 49.1 49.10 5.99 5.99 6.03 4540 4540 Ireland 81.46 81.46 81.57 2.75 2.75 2.81 2217 2217 Italy 35.44 35.44 35.68 10.07 10.07 10.90 3366 3366 Japan 3.06 3.06 3.22 25.45 25.45 39.08 3269 3269 Latvia 3.38 3.38 4.36 19.26 19.26 29.52 2103 2103 Lithuania 7.77 7.77 15.51 5.97 5.97 14.00 2009 2009 Macedonia 10.95 10.95 10.95 14.47 14.47 16.16 995 995 Malta 89.53 90.38 0.23 0.23 860 Mexico 48.22 47.92 49.26 4.47 4.95 5.03 4878 5732 Moldova 10.77 10.77 10.82 8.94 8.94 9.41 984 984 Montenegro 7.08 7.08 7.08 30.42 30.42 34.93 240 240 N. Ireland 50.65 50.65 50.81 3.25 3.25 3.36 616 616 Netherlands 22.79 22.79 22.84 29.54 29.54 32.04 2238 2238 Norway 5.07 5.21 5.24 28.74 28.15 30.47 3612 3417 Peru 41.78 41.78 42.88 1.98 1.98 2.01 1211 1211 Poland 59.92 30.66 56.08 1.48 / 2091 2091 Portugal 39.07 39.07 39.07 10.04 10.04 10.62 1185 1185 Puerto Rico 51.55 51.55 51.77 0.69 0.69 0.69 1164 1164 Romania 18.59 18.59 18.64 5.98 5.98 6.29 1103 1103 Russia 1.67 1.95 1.95 30.3 36.19 43.03 6775 4001 S. Africa 31.94 31.94 53.26 1.6 1.6 1.63 7267 7267 Serbia 5.78 5.78 5.78 27.34 27.34 30.81 1280 1280 Slovenia 22.43 22.43 22.51 33.3 33.3 36.46 2042 2042 Spain 29.64 31.72 31.89 11.59 11.08 11.81 13370 7661 Sweden 4.55 4.55 4.63 39.44 39.44 46.28 3010 3010 Switzerland 18.91 18.91 20.16 6.74 6.74 15.90 2612 2612 Turkey 32.41 32.41 38.89 1.23 1.23 2.45 2937 2937 Ukraine 9.82 9.82 10.12 19.85 19.85 23.47 2811 2811 Uruguay 13.2 13.2 13.20 13.9 13.9 14.17 1000 1000 USA 43.27 43.27 43.45 3.21 3.21 3.29 5706 5706 Venezuela 30.92 30.92 30.92 0.92 0.92 0.92 1200 1200 West Germany 16.98 19.35 4423 All countries 23.75 24.45 26.46 14.77 14.68 17.44 149653 140491

(31)

31

TABLE 1B: The Distribution of population by religious denomination and country (in percentages). One could compare to TABEL 1, Panel B from Guiso et al. and see for the largest parts the same numbers.

(32)

32

TABLE 1C:

Religiosity by religious denomination, easily comparable to Guiso et al.

TABLE 1D:

Table 1D “reports summary statistics for the variables used in the regression analysis. Variable 1 is based on the following

question: “Generally speaking, would you say that most people can be trusted or that you can't be too careful in dealing with people? The variable is equal to 1 if participants report that most people can be trusted and zero otherwise. Variables 2 and 3 are based on the following question: “On this list are various groups of people. Could you please sort out any that you would not like to have as neighbors?” (Variable 2: People of a different race; Variable 3: The immigrants). Variable 4 is the combination of variables 2 and 3 and is equal to 1 if either variable 2 or 3 or both is equal to one. Variables 5, 6 and 7 are based on the general question: I am going to name a number of organizations. For each one, could you tell me how much confidence you have in them: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all? Answers are coded 1-4, we recoded them so that a higher number represents a higher degree of confidence. Organizations we considered are the government, the police and the armed forces. Variables 8 comes from the answers to the question Do you agree or disagree with the following statement: When jobs are scarce, men should have more right to a job than women? Answers are coded 1-4, we recoded them so that a higher number represents a higher degree of agreement. Variables 9 comes from the answers to the question: Do you think that women should have children in order to be fulfilled or is this not necessary? The answer needs children is coded as one, the answer not necessary is coded as a zero. Variables 10-12 come from the answer to the question For each of the following statements I read out, can you tell me how much you agree with each. Do you agree strongly, agree, disagree, or disagree strongly? The statements are: Being a housewife is just as fulfilling as working for pay (variable 10), Both the husband and wife should contribute to household income (variable 11), A university education is more important for a boy than for a girl (variable 12). We recoded them so that a higher number represents a higher degree of agreement. Variable 13 is based on a question similar to variables 5 to 7, except that the organization mentioned is the legal system. Answers are coded 1-4, we recoded them so that a higher number represents a higher degree of confidence. Variables 14 to 18 are based on the following question: “ Please tell me for each of the following statements whether you think it can always be justified, never be justified, or something in between, using this card.” Answers are in the range 1-10, with 1 = never justifiable and 10= always be justifiable. “Claiming government benefits to which you are not entitled” (var. 14). “Avoiding a fare on public transport” (var. 15). “Cheating on taxes if you have a chance” (var. 16). “Buying something you knew was stolen” (var. 17). “Accepting a bribe in the course of their duties” (var. 18). Variables 19-21 and 23 to 26 are based on the following question: Now I'd like you to tell me your views on various issues. How would you place your views on this scale? 1 means you agree completely with the statement on the left; 10 means you agree completely with the statement on the right; and if your views fall somewhere in between, you can choose any number in between. The statement on the left is normally the opposite of the statement on the right. Statements considered are (we report only statement on the right): We need larger income differences as incentives for individual effort” (var. 19); “Private ownership of business and industry should be increased“ (var. 20); “Competition is good. It stimulates to work hard and develop new ideas” (var.21); People should take more responsibility to provide for themselves (var. 23);“In the long run hard work usually brings a better life” (var. 24); “Wealth can grow so there is enough for everybody” (var.25). Variable 22 is based on the answer to the question Here is a list of qualities that children can be encouraged to learn at home. Which, if any, do you consider to be especially important? We code as 1 if the respondent list as important “Thrift, saving money and things”. Variable 26 is based on the question: Why, in your opinion, are there people in this country who live in need? Here are two opinions: Which comes closest to you view? We code as one the answer “They are poor becausee of laziness and lack of will power” and zero the answer “They are poor because society treats them unfairly”.” – Guiso et al.

Religious Denomination Raised religiously at home Scriptie: missing values included scriptie: missing values = missing Goes to church at least once a year Scriptie: Missing = 0 scriptie: missing values = missing Goes to church at least once a week Scriptie: missing = 0 scriptie: missing values = missing Catholic 70.3 68.00 89.53 75.3 75.57 78.45 38.8 39.43 40.93 Protestant 47.4 41.55 64.94 59.5 54.56 59.67 25.0 17.25 18.86 Jew 44.6 26.38 56.70 57.0 60.19 66.93 16.4 20.62 22.93 Muslim 67.9 73.62 76.84 67.8 70.93 76.59 34.5 38.28 41.33 Hindu 83.8 91.19 92.09 84.7 84.98 88.98 44.4 44.17 46.25 Buddhist 22.2 26.72 49.79 75.8 67.73 70.66 8.3 6.14 6.41 Others 51.0 48.48 52.62 68.2 70.29 72.82 17.0 19.79 20.50 No Religion 29.61 33.60 22.62 25.40 2.53 2.84 Missing 25.10 30.39 11.44 18.03

(33)

33

TABLE 1D

In table 4 it is again easy to compare between Guiso et al. and own research. The first line is always what Guiso et al. found while the second line called “scriptie” is own research.

Variable Mean Median SD Interq. Range Min Max N 0.29 0 0.45 1 0 1 Scriptie 0.30 0 0.46 1 0 1 1327237 0.12 0 0.33 0 0 1 Scriptie 0.12 0 0.33 0 0 1 130155 0.14 0 0.35 0 0 1 Scriptie 0.14 0 0.35 0 0 1 129712 0.39 0.4 0.27 0.4 0 1 Scriptie 0.20 0 0.40 0 0 1 127153 2.3 2 0.93 1 1 4 Scriptie 2.36 2 0.91 1 1 4 63286 2.53 3 0.91 1 1 4 Scriptie 2.54 3 0.91 1 1 4 133895 2.61 3 0.94 1 1 4 Scriptie 2.63 3 0.93 1 1 4 132120 1.85 1 0.91 2 1 3 Scriptie 1.87 2 0.91 2 1 3 106797 0.53 1 0.5 1 0 1 Scriptie 0.57 1 0.50 1 0 1 126065 2.81 3 0.87 1 1 4 Scriptie 2.81 3 0.86 1 1 4 96293 1.76 2 0.71 1 1 4 Scriptie 3.16 3 0.75 1 1 4 97248 2.03 2 0.89 1 1 4 Scriptie 2.06 2 0.89 2 1 4 53717

8. When jobs are scarce, men should have more right to a job than women

Panel D: Summary statistics of economic and social attitudes

Attitudes toward others

1. Trust people (0,1)

2. Intolerant towards other races (0,1)

3. Intolerant towards the immigrants (0,1)

4. Average intolerance

Attitudes toward the government

5. Trust the government

6. Trust the police

7. Trust the armed forces

Attitudes toward women

9. Do you think that women should have children in order to be fulfilled

10. Being a housewife is just as fulfilling as working for pay

11. Both the husband and wife should contribute to household income

12. A university education is more important for a boy than for a girl

(34)

34 2.52 3 0.88 1 1 4 Scriptie 2.54 3 0.88 1 1 4 134321 2.27 1 2.25 2 1 10 Scriptie 2.27 1 2.25 2 1 10 132729 2.53 1 2.42 2 1 10 Scriptie 2.51 1 2.41 2 1 10 128198 2.6 1 2.47 3 1 10 Scriptie 2.58 1 2.46 2 1 10 132350 1.76 1 1.73 0 1 10 Scriptie 1.76 1 1.73 0 1 10 132091 1.73 1 1.7 0 1 10 Scriptie 1.73 1 1.68 0 1 10 134552 5.91 6 2.97 4 1 10 Scriptie 5.94 6 2.96 4 1 10 105637 6.17 6 2.82 4 1 10 Scriptie 6.15 6 2.84 4 1 10 102026 7.45 8 2.51 4 1 10 Scriptie 7.48 8 2.49 4 1 10 100305 0.32 0 0.47 1 0 1 Scriptie 0.33 0 0.47 1 0 1 138464 5.18 5 3.06 6 1 10 Scriptie 5.25 5 3.07 6 1 10 105253 6.54 7 2.91 5 1 10 Scriptie 6.56 7 2.9 5 1 10 102019 6.48 7 2.77 4 1 10 Scriptie 6.49 7 2.76 4 1 10 98481

Attitudes toward the market Attitudes toward legal rules

13. Trust the legal system

14. It is justified to claim government benefits you are not entitled to?

15. It is justified to avoid a fare on public transport?

16. It is justified to cheat on taxes?

17. It is justified to buy a stolen object?

18. It is justified to accept a bribe?

23. Individual responsibility versus government assistance

24. Hard work improves life versus success is more a matter of luck and connections.

25. Wealth can grow so there's enough for everyone versus one can get rich only at expense of others

19. Inequality of income gives incentives to individuals versus income should be made more equal

20. Private ownership should be increased versus government ownership should be increased

21. Competition is good versus competition is harmful

Attitudes toward thriftiness

22. Do you think to be especially important that children be encouraged to learn at home “thrift, saving money and things”

The invisible hand of God? : an update of the relation between religion and economic attitudes