ECONOMIC DEVELOPMENT AND RELIGIOSITY: A RELIGIOUS CAPITAL APPROACH TO SECULARIZATION THEORIES REBUTTAL Sergio Figueroa Sanz Instituto Tecnológico Autónomo de México ITAM Río Hondo No. 1 Col. Progreso Tizapán México, D.F. México 01080 April 2009 Working Paper for the Association for the Study of Religion, Economics, and Culture (ASREC) Annual Meeting to be held on April 2009. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit, including notice, is given to the source.
- 2 - Economic Development and Religiosity: A Religious Capital Approach to Secularization Theories Rebuttal Sergio Figueroa Sanz April 2009 Abstract Secularization theories have been popular for decades. Many scholars seem to agree that religion s role in modern society has dissolved to the point of being irrelevant or simply incompatible with economic development. Recent research, on the other hand, has found evidence that suggests the contrary. This paper attempts to provide for an explanation to such disparity, arguing that both perspectives are partially accurate since religiosity behaves as a quadratic function of economic development defined in terms of investment in religious capital and its opportunity cost. Although data availability makes it difficult for the hypothesis to be categorically proven or disregarded, preliminary evidence based on the United Nations Human Development Index database and the World Values Survey is reviewed to support the idea that societies in early development stages, just as societies in very high development stages, show high levels of religiosity, while societies in mid to high development stages show depressed levels of religiosity. If the hypothesis is correct, secularization theories ability to predict religiosity in the long run is faint and their explanatory value is limited to a very specific stage in societies development process. Sergio Figueroa Sanz Instituto Tecnológico Autónomo de México ITAM Río Hondo No. 1 Col. Progreso Tizapán México, D.F. México 01080 sergiofigueroasanz@yahoo.com
- 3 - INTRODUCTION A detailed discussion of the literature regarding the perspectives under which religion has been studied would make one quite interesting dissertation and, accordingly, would require work with a much more ambitious scope than this brief paper s 1. Nonetheless, if we narrow the possibilities to the arena of Economics, the task becomes less daunting. Indeed, it s fair to say that one of the dominant currents of thought over the past decades relative to the study of the economic relevance of religions is embodied by secularization theories. These theories, in general, take religions to be archaic figments, simply incompatible with modernity, and predict that, eventually, no one will profess a religion 2. Secularists, indeed, believe that as economies further develop, the opportunity cost of religiosity increases, depleting adherents of incentives to use resources in religious behavior. But, do religiosity trends truly account for religion s eventual removal from people s lives? Secularists arguments typically assume religious practice to be irrational behavior and argue that religiosity and education are mutually exclusive 3, considering that religious practices are founded in ignorance-driven superstition. As development permeates societies, so does science, dwarfing the need for religions as means to make up for lack of knowledge. In addition, the argument follows, as development advances, so does urbanization. Religions, being constructions based on superstition, help cope with the uncertainties ever present in rural economies when ignorance continues to prevail 4. In this sense, as societies migrate from rural economies to urban economies, the usefulness of religion is further eroded while access to science is widespread and the exposure to uncertainty reduces. As a consequence of the previously described process, secularists argue, salaries tend to be more attractive 1 A comprehensive review can be found in Iannaccone (1998). 2 Wilson (1982), Berger (1979) and Berger (1990) are good examples of secularization thought. 3 Consider the following reference from the very popular Bierce (1958): Religion is the daughter of Hope and Fear, explaining to ignorance the nature of the Unknowable ; and the following attributed to Bertrand Russell: Religion is something left over from the infancy of our intelligence, it will fade away as we adopt reason and science as our guidelines. 4 In precarious conditions of scientific development, how to explain a drought, a hurricane, a plague, sickness, or death? Bergson (1946) defines a static religion as a religion of myths and fables that gives comfort to adherents when facing uncertainty and harsh conditions.
- 4 - in developed countries than in under-developed ones, accounting for more diversity in economic activity and higher productivity levels, in general. Developed, urbanized, educated economies will be less labor-intensive in their production processes and more capital-intensive 5. As economies continue to develop, the trade off between practicing a religion and working (i.e. perceiving a salary) biases adherents to drop religion and favor work. Finally, secularists observe that economic development brings along increases in longevity and health, in general, further reducing the need for religion as means to deal with uncertainties such as sickness or death. In this sense, given that individuals in developed economies are expected to live longer lives and to dispose of more income to use in prolonging healthy lives, the practice of religion will be intensified towards the end of adherents lives -yet almost ignored earlier, reducing overall levels of religious participation 6. As already suggested, the aforementioned mechanism towards secularization can be summarized as a simple increase in the opportunity costs of religiosity as development advances 7, driving adherents willingness to be religious down. Alas, although most data sets show deteriorating religious attendance figures, the World Values Survey shows that while some time was spent by 61% of in prayer and similar religious activities in the early 1980 s, this percentage increased to almost 64% by the early 2000 s most of which reported engaging in such activities on a daily basis. 5 Recall that capital needs not necessarily be physical. Current economic theory recognized the existence and relevance of non-physical forms of capital, such as human capital and social capital. Note that when a broader concept of capital is used, the fact that some developed countries continue to have an important dependence in labor in their production processes, such labor tends to be more qualified than that of under-developed countries. In this sense, qualified labor has an important component of capital embedded (human and social capital, that is). 6 Azzi and Ehrenberg (1975) incorporated to a utility function the concept of after life consumption and took it to be the objective of religious participation, within the context of adherents life cycle. Their results confirmed the described effect of longevity over religiosity. 7 Some research has been performed to confirm such statements but none have been entirely conclusive. For instance, Gruber and Hungerman (2006) show that secular activities, indeed, compete with religious service attendance. By analyzing the effect of the modification of blue laws - state laws that prohibit retail activity on Sunday, in the United States, the paper shows evidence that suggests that, under general conditions, allowing retail activity on Sundays lowers religious attendance levels and church donations and spending. However, other papers such as McBride (2005) conclude that economic growth an important factor in economic development, hasn t resulted in the decline of religion and that theories predicting the decline of religion due to rising opportunity costs of religious demand and supply ignore countervailing influences.
- 5 - Similarly, this survey shows that back in the 1980 s, roughly 74% of deemed themselves as religious, while over 81% did by the early 2000 s. May this suggest that secularists are missing a point?
- 6 - OBSERVATIONS ON SECULARIST ARGUMENTS AND RATIONALITY Secularization theorists arguments rely on assumptions that may not be entirely accurate 8, especially if it s noted that the key assumption in these regards is to deem adherents decision to be religious as irrational. This is outright mistaken. First and foremost, religiosity is a consequence of a decision by an individual -the adherent, to believe in what the particular religion he professes presents to be truth and to adopt certain rules of behavior that reflect such principles. To put it in other words: a religious individual is religious because he chooses to be. This means that adherents must draw some utility from religion; otherwise, they would choose to either convert to another religious denomination or simply be non-religious. The reasons behind such decisions are irrelevant for our work, and most Economic inquiries for that matter. To ponder about why someone derives utility from being religious is an analogous question to that of why someone would enjoy eating an apple for a snack 9. This means that whenever an adherent behaves in coherence with the principles and rules that religion implies, he maximizes utility. Let s label this as the intrinsic value of religion. Empirical evidence supports the relevance of religion's intrinsic value. Findings by Ellison (1991) and Hout and Greeley (2003) find a strong correlation between religiosity and self-reported measures of well-being. Gruber and Mullainathan (2002) show that the effect on self-reported happiness of attending weekly to religious services instead of never at all is comparable to the effect over well-being related to moving to the top income quartile from the bottom. Furthermore, religious practice can be deemed as rational also because of what could be labeled as religion's instrumental value: the value associated to certain consequences resulting from abiding by the normative constraints imposed by some religions which, attending to theoretical frameworks and empirical evidence, may be argued to be conducive to efficiency and economic growth. On the 8 Hadden (1987), for instance, argued that secularization theories foundations are nothing more than ideologies lacking of empiric and theoretical support. 9 Economic theory would identify these as a typical case of consumer analysis, where consumer's preferences are assumed to be given.
- 7 - theoretical frameworks side, social capital applications have become popular, viewing religion as clubs that allow for networking and social interactions that may enhance productivity and reduce costs in several markets, including the work of Glaeser, Laibson and Sacerdote (2000) and Putnam (2000). In this sense, religion as means to accumulating productive capital represents for quite an appealing incentive for adherents to be, effectively, religious. Another interesting approach is found on Schultz (2001), that focuses on showing that morals are not only a desirable condition for efficiency but a necessary one. By showing that moral systems constitute a rational normative constraints system, Schultz argues that only through such a system will market failures become mitigated and Pareto efficiency be achieved. Extending the argument to religious normative systems is only a matter of degree, especially if we consider that religion is both source and beneficiary of morals. Empirical evidence supports the significance of religion's instrumental value. Gruber (2005) makes a revision of several studies that find a positive relation among select measures of religiosity 10 and some variables typically associated with desirable economic outcomes. These studies also make important findings in referring to a clear reduction on deviant behaviors such as suicide, crime, and substance abuse. Gruber also references some studies that find a significant improvement in allocation of time, school attendance and work activity, as well as reductions in the frequency of socially deviant activity in young, churchgoing, African-American males. Similarly, Gruber references several studies that find a strong positive correlation between religiosity -usually measured by religious participation, and improved health. Gruber concludes through his own data analysis and framework, that a higher density of a given religion in an area leads to significantly more religious participation, and to better outcomes along a variety of dimensions, such as education, income, and marital status, confirming our views over the instrumental value of religion. Barro and McCleary (2003) provide for additional evidence on the instrumental value of religion, showing that...economic growth responds positively to the extent of religious beliefs, notably 10 Including participation in religious services.
- 8 - those in hell and heaven, but negatively to church attendance. That is, growth depends on the extent of believing relative to belonging. The result is based on the assumption that beliefs are religion's main output and church attendance one of the inputs. This is particularly true if we stress the fact that attendance is one of the inputs, not the only one. Similar conclusions were reached by Guiso, Sapienza, and Zingales (2003), as related to positive economic attitudes driven by religion. The existence of, both, the intrinsic value and the instrumental value of religion give more than sufficient support for the claim that professing a religion and religious participation are rational behaviors, refuting secularization s foundational assumption. Still, in referring to the quite nouvelle concept of religious capital -comprehensively overviewed in Iannaccone (2003), the argument against secularization can be made more robust.
- 9 - RELIGIOSITY AND RELIGIOUS CAPITAL INVESTMENT AS SYNONYMS The concept of religious capital narrows down a series of shared characteristics with human capital and social capital while attributing to it additional, exclusive traits. It recognizes the unique ability of religious congregations to foster the creation an accumulation of elements that improve productivity and efficiency, and that are obtained in the context of religious practice. Religious capital accumulation may be conceptualized as a by-product of religious activity, assuming that adherents chose their levels of religious involvement rationally. According to Iannaccone (2003), religious capital has several key features: it enhances the utility derived from current religious activity, it grows as a consequence of past religious involvement, it is context-attuned and hence of greatest value in religious settings similar to those in which it was obtained, and it is complementary with the religious activities and capital of other family members. This simply implies that adherents value their investments in religious capital because in doing so they maximize their utility. Perhaps this explains the fact that roughly 81.26% of the world s population practices religion 11. People, indeed, care for investing in religious capital, either because of religion s intrinsic value or its instrumental value or both. Measuring religious capital would require metrics relative to years of religious instruction, past religious involvement, and exposure to religious traditions, and knowledge of rituals, sacred texts, and familiarity with the members and activities of specific congregations, to name a few. Following these ideas, religiosity may be defined in terms of individuals willingness to invest in religious capital and, of course, their actual investments. It would be hard to disagree with the idea of labeling someone that heavily invests (time, money, labor, etc.) on learning and practicing his religion as a religious person. This definition may seem simplistic, especially considering other authors previous attempts to establish a punctual definition, often times in terms of attendance to religious 11 Based on the World Values Survey s question of if the respondent practices or not a religion - www.worldvaluessurvey.org.
- 10 - services 12 ; but when agreeing to such definitions, secularization theories become alluring as ever since most data sets do agree on religious attendance decreasing over time and as attendance s opportunity costs increase 13. My definition of religiosity is broader than religious attendance and may prove to be more useful to describe what s actually going on. Investment in religious capital may include a series of activities, many of which are indeed part of public religious services typically lead in temples; but several others are private behaviors. The latter may include prayer and meditation, religious reading, or religious goods consumption (images, magazines, books, movies, etc.). Bear in mind that practicing a religion truly entails a lifestyle on its own 14. I ll borrow from Figueroa (2007) the concepts of collective religious services and individual religious services, and note that religious attendance is typically measured as attendance to very specific collective religious services such as periodic gatherings for indoctrination and the celebration of holidays (e.g. the Christian mass) leaving out individual religious services from the picture. It could be true that, although collective religious services continue to loose popularity among adherents as their opportunity costs increases thanks to economic development, adherent s appetite for some forms of religious service remains. Most definitely, individual religious services are good candidates to fulfill this appetite if religious congregations offer fitting services. A simple equilibrium model for religious services markets as repeatedly explored by other authors, would reveal that as long as religious demand is positive and that it s profitable for such demand to be satisfied, incentives will be in place for supply to be readily available. In other words, if religious service providers identify that adherents prefer individual religious services to collective services, resources will be shifted to the provision of 12 Some authors, however, have been more insightful. Glock (1962) for instance, identified five dimensions of religiosity, reflecting its complexity. 13 Recall Gruber and Hungerman (2006), that included and analysis of the effects on religious attendance of blue laws modification in the U.S. Similarly, Barro and McCleary (2002) show a negative correlation between urbanization and religious attendance due to the very same effect of religious service attendance opportunity costs rising. Also recall that Barro and McCleary (2003) found a negative correlation between economic growth and attendance. 14 Zinnbauer et al (1997), for instance, associates the concept of religiosity to the frequency with which adherents attend church, pray and in a relatively conservative fashion.
- 11 - individual services those very same services left out by most, if not all, measurements of religious attendance. Perhaps this is actually happening and a broader definition of religiosity has the ability to capture it.
- 12 - RELIGIOSITY AS A QUADRATIC UNCTION OF DEVELOPMENT As already averred, secularization theories have been among the most popular currents of thought for decades regarding the role of religion in modern societies, predicting the eventual disappearance of religion from quotidian life. A handful of authors, however, have observed that over the past years, certain indicators of religiosity have increased, contradicting secularists predictions. Who s got it right anyway? This paper focuses on giving credit to both perspectives in formulating a hypothesis I ll label as Religiosity s U, establishing that religiosity behaves as a quadratic function of economic development defined in terms of investment in religious capital and its opportunity cost. Simply put, the hypothesis establishes that societies in early development stages, just as societies in very high development stages, show high levels of religiosity, while societies in mid to high development stages show depressed levels of religiosity. The mechanism is relatively simple and is based on the following assumption: Adherents will invest in religious capital as long as the benefits derived from such investments 15 outweigh their costs. Consider a variable composed of a vector of indicators of economic development that we ll call D Index, namely: (i) education attainment, (ii) health and longevity, (iii) purchasing power, and (iv) religious markets diversity levels. A high D Index value denotes high indicators in these components. On the other hand, consider another compound variable comprised of several indicators of religiosity as it was defined earlier and simply call it a Religiosity Index. High values of the Religiosity Index denote intensive investment in religious capital. The Religiosity s U behavior relatively simple behavior is illustrated below. 15 Such benefits can be directly related to the already discussed concepts of religion s intrinsic and instrumental values.
- 13 - Where Rb < Rc < Ra and Da < Db < Dc The hypothesis establishes that a society with low D Index values, such as that which corresponds to a society like A in the previous graph, will show high Religiosity levels. These Religiosity levels will be, however, consistent with low quality religious capital stock. That religious capital stock is of low quality implies that it s associated to superstition and ignorance, as reflected by low D Index values. Religious beliefs in a society with these characteristics will most likely serve as a mechanism to help adherents cope with uncertainty and lack of access to science. For instance, Mexican Prehispanic cultures attributed rain to a deity, instead of to the Bergeron-Findeisen process 16. Probably, A will be a highly rural economy, where individuals will be mostly focused in satisfying the most basic subsistence needs and where, given the very limited access to science and technology, productive activity will be highly labor-intensive. In addition, as denoted by this economy s low Da D Index value, health and longevity indicators will be low, and thus, religious practice will help to deal with sickness and death. Finally, this economy can be expected to be relatively poor and with limited 16 The scientific process that explains rain.
- 14 - competition among activities, in general 17. As a consequence, the opportunity cost of religiosity will be relatively low while the benefit of religiosity will be high since adherents will be particularly drawn to religion for it to help cope with ignorance, uncertainty, and poverty, as already argued. Religious congregations will also have very limited incentives to behave strategically or efficiently for that mater. Clearly, the state of religious practice under this context seems to mirror secularists observations about religion. In a mildly developed economy such as B, science and education will be more readily available, people will live longer healthier lives, and the average individual will be relatively wealthy. In this economy, technologies for production will allow for a departure from labor-intensive production schemes leaving time for leisure, and markets will be more competitive and vast, providing for intense secular competition to religious activities. Productivity will be higher and, thus, wages and ultimately wealth will increase. Access to health services will be more generalized and uncertainty will be less of a protagonist of quotidian life. This economy will be more urbanized as well. This implies that religion s appeal as means to mitigate ignorance, uncertainty and poverty will be dramatically reduced, depleting adherents from incentives to be religious, especially considering religiosity s increased opportunity cost. For instance, after individuals find out that they can replicate fire at will, the usefulness of investing resources in worshipping the god of fire is dwarfed. In this sense, a society like B will show D Index values that lie at the middle of the scale such as Db, while Religiosity will be critically low at Rb, just as secularists would predict. Indeed, as this society has continued to develop over time, moving to a value of Db from a value similar to Da, the opportunity cost of religiosity has increased while the usefulness of being religious has reduced. Simultaneously, religious congregations at this point will have done little, if anything at all, to behave competitively, resulting in them loosing adherents interest to secular activities. However, as this society s D Index increases from Db towards Dc, the Religiosity s trend will start to shift. 17 Indeed, underdeveloped economies tend to suffer from incomplete markets.
- 15 - As development advances from a level as Db, religious congregations should be expected to behave more strategically and react to the very low levels of Religiosity. The fierce competition of secular activities will drive competitive behavior in religious congregations for them to offer improved religious services in an attempt to increase adherent s interest in investing in religious capital. This attempt will be successful to the extent that development brings along further departure from laborintensive production and increases capital-intensiveness. This will imply that capital stock s value will also increase, placing incentives for investing in it this should include physical, human, social, and religious capital altogether. Religious congregations will start alluring adherents back as they succeed in providing for productive religious capital and adherent s willingness and capacity to invest in it continue to increase over time. This is a process that a society such as C will have had undergone. A society like C will show very high levels of development and high levels of Religiosity. At this point, this economy will be composed of extremely educated, healthy and wealthy individuals that, as Sacerdote and Glaeser (2001) show, have a larger propensity to participate of group activities, which could include religious services. Here, religious congregations will offer an array of services, some collective and some individual, that satisfy adherents demand for religious services while allowing them to access and accumulate valuable religious capital. Indeed, religious markets will be highly competitive. Some aspects of religious capital will be complementary to education given religious capital s shared traits with human capital 18, others will foster healthy lives since most religions encourage behaviors with economically desirable behaviors as already discussed, while others will enhance production processes since religious capital may increase productivity given its shared traits with human and social capital 19. In this society, investment in religious capital will be deemed as something desirable for production efficiency and adherents will assign some utility to being religious 18 Religious congregations offer courses, workshops, continued education, etc. Some even sponsor, directly, higher education institutions all over the world, promoting technical change. 19 Religious congregations may also be extremely efficient channels for networking, reducing transactional and factor costs. They may also provide for direct funding through schemes as the increasingly popular faith-based credit unions.
- 16 - again, only that for completely different reasons as the ones taken into consideration by society A. In this sense, levels of Religiosity will not only be quite high at Rc, but Religiosity s quality will be also high. Note, however, that religiosity s opportunity cost will be higher than ever, but its benefits will outweigh the costs, tilting the scale in favor of Religiosity in secularization s stead. Constructing the Religiosity s U entails assuming that economic development has also some impact over religious markets competition and that the level of market competition in religious markets is related to overall levels of religiosity. In general, it would seem fair to assume that when comparing two societies with identical characteristics except for the degree of competition in a given market, the society with more competitive markets should be deemed as more developed 20. The relation held by religious market competition and religiosity is, however, less trivial. This relationship is based on the observation that when religious monopoly prevails in underdeveloped economies, religiosity levels will be extremely high. If this religious monopoly becomes threatened by other religious congregations, the effect over the monopoly s adherence s Religiosity should be expected to be similar to the effect of increased education, health and longevity, and wealth indicators over Religiosity: induce a decrease. To the extent that religious competition is effective and motivates strategic behavior across congregations, religious services supply will fit adherent s demand better attracting back investment in religious capital after Religiosity had reached its critical low point. These trends play along with other indicators of the D Index. For instance, more income availability gives adherents more flexibility to invest in religious capital in a similar fashion as it allows them to invest more intensively in human and social capital. In turn, these investments grant religious congregations funding to allow them to provide for religious services of higher quality and that appeal to adherents. This will effectively happen if competition prevails in this economy s religious markets. If the trends described by Religiosity s U are correct, then we have at hand a tool for explaining 20 In terms of economic efficiency, market competition is preferable to lack of competition, and a more efficient society should always be deemed as more developed.
- 17 - why secularization theories appeared in the first place and for explaining why is it that some recent studies and surveys seem to contradict such theories. If empirical analysis confirms the Religiosity s U, it could be concluded that secularization theories are accurate in predicting religiosity s trends in certain cases, but are mistaken in assessing religions ultimate destination.
- 18 - SUPPORTING DATA Proofing Religiosity s U is no simple task. Variables associated to religiosity are extremely elusive and difficult to measure. Religious congregations rarely keep reliable records on objective metrics for collective religious service attendance and giving, much less regarding individual religious services. The United States has a relatively wide availability of data; this is far from true elsewhere. Notwithstanding, the World Values Survey (WVS) represents one of the most commendable efforts to solve for this deficiency and, for this reason, was used to provide for empirical support for the hypothesis. This survey includes information on religious affiliation, religious beliefs and religious behaviors for 77 countries. Some of this information was used to serve as proxy for the compound variable labeled as the Religiosity Index. The very popular and generally accepted Human Development Index () from the United Nations (UN), and its components, were used as a proxy for the D Index. The is built as a composite variable of a weighted average of each country s life expectancy, an index of educational attainment composed of variables such as adult literacy rates and enrollment at various educational levels, and per-capita gross domestic product adjusted to purchasing power parity in U.S. dollars. A higher value denotes higher development indicators. We used the as calculated by the UN and its separate components for the analysis. However, it must be noted that the doesn t include information relative to religious markets competitiveness. The WVS was used again to solve for this deficiency in creating a simple measurement of religious market concentration as proxy. The analysis was performed using the 77 countries for which the WVS has data.
- 19 - TABLE 1 Countries analyzed 1 Iceland 27 Portugal 53 Dominican Republic 2 Norway 28 Czech Republic 54 China 3 Australia 29 Malta 55 Armenia 4 Canada 30 Hungary 56 Turkey 5 Ireland 31 Poland 57 Jordan 6 Sweden 32 Argentina 58 Peru 7 Switzerland 33 Chile 59 Philippines 8 Japan 34 Slovakia 60 Iran 9 Netherlands 35 Lithuania 61 Georgia 10 France 36 Estonia 62 El Salvador 11 Finland 37 Latvia 63 Algeria 12 United States 38 Uruguay 64 Viet Nam 13 Spain 39 Croatia 65 Indonesia 14 Denmark 40 Mexico 66 Moldova 15 Austria 41 Bulgaria 67 Egypt 16 United Kingdom 42 Romania 68 Kyrgyzstan 17 Belgium 43 Saudi Arabia 69 South Africa 18 Luxembourg 44 Belarus 70 Morocco 19 New Zealand 45 Bosnia and Herzegovina 71 India 20 Italy 46 Russian Federation 72 Pakistan 21 Hong Kong, China 47 Albania 73 Bangladesh 22 Germany 48 Macedonia 74 Zimbabwe 23 Greece 49 Brazil 75 Uganda 24 Singapore 50 Venezuela 76 Nigeria 25 Korea 51 Colombia 77 Tanzania 26 Slovenia 52 Ukraine The WVS selected indicators for each country were taken from the survey s Religion and Morale section, as evaluated by the following questions: (i) do you believe in God?, (ii) do you believe in life after death, (iii) do you take some moments of prayer, meditation or contemplation or something like that?, (iv) how often do you pray to God outside of religious services?, (v) do you belong to a religious denomination?, and (vi) if you belong to a religious denomination, to which one? The first two questions were used as proxies to adherents willingness to invest in religious capital, regardless of if they actually do is seems fait to assume that investing in religious capital in the first place requires for some belief in either God or a similar concept, or life after life in some shape
- 20 - or form. Questions (iii) to (v) are indicators of actual investment in religious capital 21. The used indicators are the percentages out of total answering that they attend religious services at least once a week and that dedicate some time to prayer outside of the context of what I ve called collective religious services these could be labeled as the highly religious (i.e. those that invest more heavily in religious capital). Including some indicator of religiosity outside the context of collective religious service is very important for this analysis, considering that most previous studies have narrowed their conclusions to religious attendance typically only to collective religious services. As already suggested, this paper intends to show that attendance to collective services indicators should be complemented by data on individual religious services to account for shifts in adherents demand relative to the type of services they value most in certain given circumstances, as accounted by their economic development stage. Finally, question (vi) in addition to including information on self-reported belonging, was used to create a proxy on religious markets concentration. This question refers to selfidentification as belonging to a given religious denomination regardless of how intensively they invest in religious capital. These figures were used to create an extremely simple indicator of religious markets concentration, defined as 1-p, where p is each country s highest reported percentage of belonging to one single denomination (i.e. a higher value of 1-p denotes more religious markets diversity ). In this sense, data from the WVS helped construct the variable labeled as Religiosity, including information on adherents willingness to invest in social capital (questions (i) and (ii)) and actual investment in religious capital in certain aspects (questions (iii) through (v)). The index was simply 21 Some U.S. surveys such as the Baylor Religion Survey 2005 and the Religion and Public Life Survey 2002 have i included questions relative to adherents expenditure in religious goods, their involvement in volunteer activities or educative activities within their religious congregations, and the frequency with which they consult religious texts. Obviously, including similar metric for this paper would be very helpful and would help the hypothesis analysis much more conclusive; however, these kinds of metrics are only available for the U.S. and are extremely limited for other countries. Having a complete data set for all the analyzed countries in this document is still well beyond the current availability of information. The WVS most definitely doesn t cover such metrics yet.
- 21 - calculated as the arithmetic mean of these indicators. Question (vi) was used to include the one indicator that the doesn t regarding the definition of the D Index. The UN s data base also includes a calculation on inequality which was also borrowed for the analysis: the Gini coefficient. This indicator is not weighed-in the but was included to test its relevance in gauging Religiosity against different popular indicators of economic development.
- 22 - QUANTITATIVE ANALYSIS AND RESULTS The Religiosity s U was defined as a quadratic function of the D Index. Being this a rough first approach to its validation, two linear equations were defined with opposing trends for two groups of countries defined as low development countries and high development countries, as follows. TABLE 2 Countries divided according to their development level High development Low development 1 Iceland 39 Croatia 2 Norway 40 Mexico 3 Australia 41 Bulgaria 4 Canada 42 Romania 5 Ireland 43 Saudi Arabia 6 Sweden 44 Belarus 7 Switzerland 45 Bosnia and Herzegovina 8 Japan 46 Russian Federation 9 Netherlands 47 Albania 10 France 48 Macedonia 11 Finland 49 Brazil 12 United States 50 Venezuela 13 Spain 51 Colombia 14 Denmark 52 Ukraine 15 Austria 53 Dominican Republic 16 United Kingdom 54 China 17 Belgium 55 Armenia 18 Luxembourg 56 Turkey 19 New Zealand 57 Jordan 20 Italy 58 Peru 21 Hong Kong, China 59 Philippines 22 Germany 60 Iran 23 Greece 61 Georgia 24 Singapore 62 El Salvador 25 Korea 63 Algeria 26 Slovenia 64 Viet Nam 27 Portugal 65 Indonesia 28 Czech Republic 66 Moldova 29 Malta 67 Egypt 30 Hungary 68 Kyrgyzstan 31 Poland 69 South Africa 32 Argentina 70 Morocco 33 Chile 71 India 34 Slovakia 72 Pakistan 35 Lithuania 73 Bangladesh 36 Estonia 74 Zimbabwe 37 Latvia 75 Uganda 38 Uruguay 76 Nigeria 77 Tanzania
- 23 - The classification was not arbitrary; it was made considering each country s, separating the sample according to its median 22. In addition, each segment was separated into two: one which considers all, and another that considers only with a high level of education as classified by the WVS. This, then, means that the analysis considered four samples: (i) The sample for all in low development countries, (ii) the sample for highly educated in low development countries, (iii) the sample for all in high development countries, and (iv) the sample for highly educated in high development countries. The following table shows such samples descriptive statistics. TABLE 3 Samples' descriptive statistics development indicators religious capital investment indicators (a) religious capital investment indicators (b) Low development countries Variable Std. Obs. # Mean Dev. Min. Max. * 39 0.725 0.106 0.467 0.850 * 39 0.713 0.141 0.265 0.853 * 39 0.808 0.125 0.466 0.956 Purchasing * 39 0.653 0.116 0.335 0.844 Gini * 39 0.398 0.091 0.260 0.586 Religious Diversity ** 39 0.297 0.219 0.000 0.727 Belief in life after death 37 0.673 0.263 0.025 1.000 Belief in God 37 0.904 0.162 0.145 1.000 Prayer at least once a week 24 0.533 0.245 0.045 0.919 Payer (yes/no) 21 0.725 0.239 0.254 1.000 Religious service attendance at least once a week 39 0.356 0.247 0.000 0.942 Belong to a religious denomination 38 0.811 0.216 0.093 1.000 Religiosity 39 0.671 0.204 0.046 0.952 Belief in life after death 37 0.683 0.247 0.025 1.000 Belief in God 37 0.925 0.149 0.184 1.000 Prayer at least once a week 24 0.573 0.225 0.089 0.866 Payer (yes/no) 21 0.751 0.221 0.294 0.972 Religious service attendance at least once a week 39 0.375 0.243 0.022 0.923 Belong to a religious denomination 38 0.842 0.195 0.061 1.000 Religiosity 39 0.691 0.197 0.042 0.947 22 The value was 0.85.
- 24 - development indicators religious capital investment indicators (a) religious capital investment indicators (b) High development countries Variable Std. Obs. # Mean Dev. Min. Max. * 38 0.922 0.038 0.852 0.968 * 38 0.886 0.046 0.770 0.954 * 38 0.959 0.031 0.856 0.993 Purchasing * 38 0.922 0.640 0.768 1.000 Gini * 35 0.341 0.072 0.247 0.549 Religious Diversity ** 38 0.361 0.198 0.024 0.739 Belief in life after death 37 0.572 0.156 0.237 0.925 Belief in God 37 0.742 0.161 0.396 1.000 Prayer at least once a week 30 0.355 0.164 0.116 0.744 Payer (yes/no) 32 0.651 0.155 0.318 0.901 Religious service attendance at least once a week 38 0.208 0.176 0.024 0.874 Belong to a religious denomination 38 0.712 0.190 0.246 1.000 Religiosity 38 0.541 0.135 0.287 0.901 Belief in life after death 37 0.570 0.162 0.272 0.862 Belief in God 37 0.786 0.150 0.390 0.995 Prayer at least once a week 30 0.389 0.177 0.145 0.805 Payer (yes/no) 32 0.636 0.153 0.373 0.912 Religious service attendance at least once a week 38 0.212 0.179 0.025 0.822 Belong to a religious denomination 38 0.753 0.183 0.248 1.000 Religiosity 38 0.559 0.135 0.279 0.897 * Reported by the UN's Human Development Report 2007 ** Calculated as 1-p, where p is the highest concentration in a single religious denomination as reported by the WVS's 1999-2004 wave Probabilities based on the WVS's 1999-2004 wave Arithmetic mean from WVS's selected indicators (a) Includes only with high levels of education, as classified by the WVS (b) Includes all Testing the hypothesis was based on multivariate linear regressions, expecting to find, for low development countries, a negative relationship between development indicators and religious capital investment indicators; whereas a positive relationship for high development countries. In addition, it was expected that these relationships are stronger for highly educated. The following tables summarize the analysis main results, showing each distinct development indicator relative to the Religiosity index and its individual components, for each reviewed sample. For each one, a separate regression analysis was performed, to separate each variable s effect from others. The tables show each model s coefficients for each exogenous variable i (i.e.,, and
- 25 - Purchasing ), the Gini coefficient, and Religious Diversity; as well as each model s R 2. Below each coefficient s value, the respective standard errors appear in parenthesis. TABLE 4 Regression analysis results Religiosity index exogenous variables all highly educated Purchasing low development countries Religious variable i Gini Diversity R 2-1.031 ** 0.460-0.223 0.346 (0.266) (0.305) (0.130) -0.729 ** 0.282-0.313 * (0.234) (0.330) (0.150) -0.788 ** 0.479-0.082 (0.223) (0.313) (0.222) -0.745 ** 0.645-0.171 (0.269) (0.336) (0.139) -1.073 ** 0.626 * -0.243 (0.267) (0.306) (0.130) -0.746 ** 0.444-0.333 * (0.236) (0.333) (0.151) -0.801 ** 0.647 * -0.096 (0.226) (0.317) (0.132) 0.267 0.310 0.232 0.386 0.299 0.337 Purchasing -0.817 ** 0.828 ** -0.192 (0.267) (0.344) (0.139) 0.288 high development countries exogenous variables Religious variable i Gini Diversity R 2 0.867 1.054 ** -0.170 0.309 all highly educated Purchasing (0.538) (0.292) (0.101) 0.513 0.912 ** -0.168 (0.415) (0.274) (0.103) 0.247 0.907 ** 0.145 (0.819) (0.307) (0.106) 0.594 1.106 ** -0.183 (0.328) (0.295) (0.101) 1.183 * 1.093 ** -0.140 (0.521) (0.282) (0.098) 0.592 0.890 ** -0.134 (0.413) (0.273) (0.103) 0.803 0.966 ** -0.096 (0.810) (0.303) (0.105) Purchasing 0.803 ** 1.160 ** -0.158 (0.314) (0.284) (0.097) 0.287 0.254 0.323 0.339 0.278 0.253 0.364 p<0.10 * p<0.05 ** p<0.01
- 26 - Belief in life after death exogenous variables all highly educated Purchasing variable i low development countries Gini -0.990 ** 0.701-0.372 (0.343) (0.399) (0.169) -0.618 * 0.554-0.435 * (0.302) (0.432) (0.194) -0.930 ** 0.722-0.230 (0.267) (0.381) (0.158) -0.588 0.873-0.312 (0.341) (0.433) (0.179) -1.120 ** 0.923 * -0.324 (0.356) (0.414) (0.176) -0.744 * 0.743-0.408 * (0.314) (0.449) (0.201) -0.962 ** 0.950 * -0.166 (0.285) (0.407) (0.168) Religious Diversity R 2 0.303 0.223 0.363 0.196 0.339 0.264 0.362 Purchasing -0.727 * 1.130 * -0.262 (0.354) (0.450) (0.186) 0.236 high development countries exogenous variables Religious variable i Gini Diversity R 2 0.774 1.103 ** -0.017 0.225 all Purchasing (0.698) (0.381) (0.136) 0.407 0.970 * -0.012 (0.535) (0.356) (0.137) 0.473 1.007 * 0.015 (1.049) (0.389) (0.140) 0.527 1.151 ** -0.031 (0.431) (0.390) (0.137) 0.209 0.199 0.232 highly educated 0.938 1.180 ** 0.005 (0.639) (0.348) (0.124) 0.469 1.017 ** 0.011 (0.494) (0.329) (0.127) 0.852 1.105 ** 0.052 (0.965) (0.358) (0.128) Purchasing 0.608 1.225 ** -0.010 (0.395) (0.357) (0.125) 0.290 0.261 0.258 0.294 p<0.10 * p<0.05 ** p<0.01
- 27 - Belief in God exogenous variables low development countries Religious variable i Gini Diversity R 2-0.453 0.517-0.289 ** 0.307 (0.210) (0.245) (0.104) all highly educated -0.404 * 0.410-0.353 ** (0.174) (0.248) (0.111) -0.306 0.531 * -0.228 * (0.175) (0.250) (0.103) Purchasing -0.274 0.597 * -0.262 * (0.202) (0.257) (0.106) -0.595 ** 0.665 * -0.320 ** (0.214) (0.249) (0.105) -0.498 * 0.535 * -0.394 ** (0.179) (0.255) (0.114) -0.405 * 0.684 * -0.240 * (0.180) (0.258) (0.106) 0.321 0.276 0.250 0.390 0.391 0.346 Purchasing -0.395 * 0.777 ** -0.288 * (0.209) (0.266) (0.110) 0.319 high development countries exogenous variables Religious variable i Gini Diversity R 2 0.543 1.420 ** -0.201 * 0.410 all Purchasing (0.593) (0.323) (0.115) 0.296 1.334 ** -0.198 (0.452) (0.301) (0.116) 0.414 1.371 ** -0.176 (0.884) (0.328) (0.118) 0.350 1.452 ** -0.210 (0.367) (0.332) (0.116) 0.402 0.398 0.411 1.105 1.478 ** -0.184 0.369 (0.651) (0.355) (0.126) highly educated 0.588 1.289 ** -0.178 (0.505) (0.336) (0.130) 0.809 1.361 ** -0.134 (0.995) (0.369) (0.133) 0.338 0.323 Purchasing 0.730 1.538 ** -0.204 (0.401) (0.363) (0.127) 0.377 p<0.10 * p<0.05 ** p<0.01
- 28 - Prayer at least once a week exogenous variables all highly educated low development countries Religious variable i Gini Diversity R 2-1.049 ** 1.159 ** -0.419 ** 0.636 (0.279) (0.362) (0.146) -0.880 ** 0.867 * -0.620 ** (0.258) (0.407) (0.174) -0.695 * 1.267 ** -0.223 (0.272) (0.408) (0.167) Purchasing -0.832 ** 1.549 ** -0.384 * (0.257) (0.373) (0.154) -1.072 ** 1.372 ** -0.396 (0.309) (0.401) (0.162) -0.846 ** 1.110 * -0.583 ** (0.292) (0.461) (0.197) -0.777 * 1.462 ** -0.187 (0.288) (0.431) (0.177) 0.608 0.532 0.593 0.622 0.571 0.557 Purchasing -0.832 ** 1.769 ** -0.359 * (0.285) (0.414) (0.170) 0.576 high development countries exogenous variables Religious variable i Gini Diversity R 2 0.506 1.203 * -0.114 0.261 all highly educated (0.822) (0.432) (0.146) 0.428 1.140 * -0.113 (0.620) (0.408) (0.145) 0.527 1.048 * -0.112 (0.307) (0.438) (0.147) Purchasing 0.397 1.247 * -0.125 (0.527) (0.443) (0.147) 0.856 1.201 ** -0.047 (0.750) (0.394) (0.133) 0.527 1.087 ** -0.040 (0.573) (0.377) (0.134) 0.051 1.055 ** -0.029 (1.221) (0.409) (0.137) Purchasing 0.663 1.274 ** -0.065 (0.477) (0.400) (0.133) 0.264 0.254 0.267 0.287 0.274 0.247 0.305 p<0.10 * p<0.05 ** p<0.01
- 29 - Payer (yes/no) exogenous variables all highly educated Purchasing Purchasing exogenous variables all low development countries Religious variable i Gini Diversity R 2-0.903 * 0.640-0.440 0.393 (0.375) (0.497) (0.223) -0.750 * 0.365-0.594 * (0.345) (0.562) (0.256) -0.770 * 0.724-0.290 (0.347) (0.500) (0.222) -0.597 0.981-0.363 (0.335) (0.509) (0.232) -1.019 * 0.784-0.375 (0.398) (0.528) (0.237) -0.849 * 0.471-0.550 (0.367) (0.598) (0.272) -0.830 * 0.890-0.208 (0.375) (0.540) (0.240) -0.705 1.169 * -0.291 (0.355) (0.540) (0.246) 0.362 0.368 0.314 0.413 0.381 0.369 0.340 high development countries Religious variable i Gini Diversity R 2 0.332 1.169 ** -0.056 0.295 (0.717) (0.371) (0.128) -0.033 1.103 ** -0.049 (0.542) (0.351) (0.128) 0.416 1.159 ** -0.044 (1.147) (0.375) (0.129) Purchasing 0.361 1.230 ** -0.069 (0.459) (0.377) (0.129) 0.289 0.292 0.306 highly educated 0.599 1.023 * -0.055 (0.785) (0.406) (0.140) -0.028 0.908 * -0.043 (0.597) (0.387) (0.141) 1.228 1.065 ** -0.026 (1.244) (0.406) (0.139) Purchasing 0.561 1.102 * -0.073 (0.499) (0.410) (0.140) 0.203 0.185 0.215 0.224 p<0.10 * p<0.05 ** p<0.01
- 30 - Religious service attendance at least once a week low development countries exogenous variables Religious variable i Gini Diversity R 2-1.613 ** 0.834 ** -0.055 0.584 all Purchasing (0.262) (0.300) (0.128) -1.086 ** 0.571-0.180 (0.255) (0.360) (0.164) -1.253 ** 0.864 ** 0.166 (0.225) (0.316) (0.131) -1.208 ** 1.133 ** 0.022 (0.288) (0.360) (0.149) 0.427 0.539 0.421 highly educated -1.622 ** 0.813 * -0.103 (0.279) (0.320) (0.136) -1.005 ** 0.572-0.203 (0.277) (0.391) (0.178) -1.286 ** 0.843 * -0.120 (0.235) (0.329) (0.137) Purchasing -1.288 ** 1.129 ** -0.031 (0.293) (0.366) (0.152) 0.550 0.353 0.523 0.428 exogenous variables variable i high development countries Gini Religious Diversity R 2 all highly educated 0.105 0.737-0.160 (0.715) (0.387) (0.134) -0.026 0.712 * -0.157 (0.542) (0.358) (0.135) -0.186 0.685-0.162 (1.045) (0.392) (0.136) Purchasing 0.162 0.779 * -0.166 (0.439) (0.396) (0.135) 0.587 0.882 * -0.088 Purchasing (0.646) (0.350) (0.122) 0.182 0.771 * -0.081 (0.495) (0.328) (0.123) 0.357 0.812 * -0.067 (0.956) (0.358) (0.124) 0.475 0.945 * -0.101 (0.393) (0.355) (0.121) 0.132 0.130 0.132 0.135 0.170 0.152 0.153 0.186 p<0.10 * p<0.05 ** p<0.01
- 31 - Belong to a religious denomination low development countries exogenous variables Religious variable i Gini Diversity R 2-0.723 * 0.155-0.366 * 0.246 all (0.282) (0.324) (0.142) -0.549 * 0.019-0.439 ** (0.238) (0.336) (0.156) -0.437 0.167-0.274 (0.243) (0.339) (0.144) Purchasing -0.637 * 0.315-0.331 * (0.266) (0.333) (0.141) 0.222 0.177 0.230-0.611 * 0.367-0.457 ** (0.311) (0.358) (0.157) 0.251 highly educated -0.460 0.253-0.518 ** (0.261) (0.369) (0.171) -0.323 0.378-0.380 ** (0.265) (0.369) (0.157) 0.235 0.200 Purchasing -0.601 0.516-0.433 ** (0.289) (0.361) (0.153) 0.261 high development countries exogenous variables Religious variable i Gini Diversity R 2 2.170 ** 0.819 * -0.400 ** 0.315 all highly educated (0.765) (0.414) (0.144) 1.505 ** 0.480-0.403 ** (0.592) (0.392) (0.147) 0.389 0.415-0.343 * (1.254) (0.470) (0.163) Purchasing 1.395 ** 0.911 * -0.428 ** (0.465) (0.419) (0.143) 2.306 ** 0.958 * -0.399 ** (0.800) (0.433) (0.150) 1.294 * 0.573-0.391 * (0.642) (0.425) (0.160) 1.457 0.693-0.314 (1.291) (0.484) (0.168) Purchasing 1.499 ** 1.062 * -0.429 ** (0.484) (0.437) (0.149) 0.286 0.141 0.332 0.314 0.231 0.165 0.336 p<0.10 * p<0.05 ** p<0.01