Religion and long-term investment

Table of Contents

Table of Abbreviations

Table of Tables

Table of Figures

Abstract

1. Introduction

2. Research questions

3. Comparison Turkey and Germany

4. Literature Review of Present-bias
4.1 Factors influencing Present-bias

5. Literature Review on Religiosity Measures
5.1 Problem of Measuring Religion
5.2 Overview of Previous Researchers on Religiosity Measures
5.3 Religious Attitude and Orientation
5.4 Religious Commitment and Involvement
5.5 The Measurement of Religious Dimensions in European Surveys
5.6 Relevant Dimensions

6. Religion and Present-Bias
6.1 Religion influences Self-Control
6.2 Religion influences Present-Bias

7. Hypotheses Development

8. Research Design and Methodology
8.1 Design
8.2 Sample
8.3 Questionnaire Development

9. Empirical Results
9.1 Sample Summary
9.2 Impatience under exponential and quasi-hyperbolic model

10. Hypothesis Testing

11. Discussion

12. Limitations and Future Research
12.1 Future Research

Appendices

Table of Abbreviations

illustration not visible in this excerpt

Table of Tables

Table 1. Overview of relevant measurements for religiousness

Table 2. Average discount values of delta (exponential)

Table 3. Average discount values of delta (hyperbolic) and beta

Table 4. Regression of Impatience and Present-bias - Germany (*Significance at the 5% level)

Table 5. Output of regression for independent variables and dependent variable - Germany

Table 6. Output of regression for independent variables and dependentt variable - Turkey

Table 7. Output of regression for religion and beta - Turkey

Table 8. Output of ANOVA for denomination and beta - Germany

Table 9. Output of correlation analysis for denomination and beta - Turkey

Table 10. Output of correlation analaysis for religion and beta - Germany (* Correlation is significant at the 0.05 level (2-tailed))

Table 11. Output of correlation analysis for religion and beta - Turkey (* Correlation is significant at the 0.05 level (2-tailed))

Table 12. Ouput of independent t-test for reminder and beta - Germany

Table 13. Output of independent t-test for reminder and beta - Turkey

Table 14. Output of independent t-test for the beta comparison between Germany and Turkey

Table 15. Output of independent t-test for the religion comparison between Germany and Turkey

Table of Figures

Figure 1. Impact of religion on present-bias and saving/borrowing behaviour

Figure 2. Measure of Religiousness

Figure 3. Relationship amnong the independent variables religion, age, gender, education, faculty department, denomination and the dependent variable present-bias, which simultaneously influences saving or borrowing behaviour

Abstract

By running a survey about the religiousness of students in Germany and Turkey, the impact on the present-bias parameter is analysed. Highly religious people suffer less present-bias problems than non-highly religious people do. For the independent variables age, gender, denomination, faculty department, and education, no significance was found. Increasing the consciousness of religion in participants’ mind showed a positive impact on present-bias, at least for Turkey. Using this parameter, which is achieved by surveys, saving/borrowing behaviour of a specific target group can be detected and directly addressed in order to stimulate national savings.

Keywords: Present-bias, Time-Inconsistency, Quasi-Hyperbolic Discounting, Exponential Discounting, Religion, Self-control, Impatience, Time Preferences

Acknowledgement

I gratefully acknowledge my supervisor Burak Can, who offered me the chance to be a part of a bigger research project guided by the sole exchange entity of Turkey – Borsa Istanbul. His patience, excellent work performance and friendly, open nature gave me the opportunity to improve my work with passion. Furthermore I would like to express gratitude to Orhan Erdem. He made the cooperation with Borsa Istanbul possible and supervised me during my stay in Turkey. His different viewpoints helped me to improve the quality of my work. I would also like to thank Filiz Erturk for her support to gather the data from Turkey. Without her help, I wouldn’t be able to compare two countries. Finally I also want to thank my boyfriend Jesko Wolff, who supported me during my whole studies. Giving me motivation and criticising me in a charming way, helped me to improve my work and myself.

Introduction

People mainly try to make rational decisions. Logical reasoning, while considering several influencing factors, is a complex process. At the same time, making a choice depends on the goals’ focus, which is set individually. Some people prefer having higher present outcomes, whereas, others deem future results as more importance. Therefore, short- and long-term goals also influence this decision making process. As Shefrin and Thaler (1988) prove, majority of people prefer sooner-smaller offers to later-larger ones. Even though they know exactly which offer would yield the highest outcome, affective^[1] systems induce a deviating selection (Meier and Sprengler, 2010).

The influence of experiences, feelings and emotions can change the idea of rational decision-making. Factors that could constantly have an impact on the decision making process are self-control and impatience. By considering different periods while making choices over time, such as savings, self-control and impatience, affect this process. According to these factors’ occurrence, they can lead to different time preferences. Due to these preferences, a constantly identical discounting^[2] process will not take place. Thus, people do not behave in a time consistent manner, which is called time inconsistency. Therefore people could show different preferences in two different points in time, which is represented in a present-bias parameter.

This knowledge helps to analyse and interpret different behavioural patterns of people. Evidence is already given for the relation between the present-bias problem and credit card borrowing behaviour of people (Meier and Sprengler, 2010). Thinking about a specific factor, which is evolved over time and embedded in the emotional state of a person and which could constrain the excessive borrowing behaviour of people, suggests that religion could have an impact in this decision making process (Shariff and Norenzayan, 2007). By focusing on such a factor, which is also embossed by the society and cultural aspects, impacts on the present-bias problem will be analysed. Therefore, this study analyses the relationship between an individual’s religion and present-bias, which indirectly affects people’s saving/borrowing behaviour.

First, a detailed discussion of the research questions will follow. Additional information and illustrations give more insights about the relationship between the main research question and the underlying sub-questions. Second, literature reviews, which build a foundation for comprehension and answers, will be represented. By making use of the extensive body of literature as a guideline, the main hypotheses of the study are filtered. Third, using primary data, statistical tests are undertaken in order to find results for the studies’ assumptions. Finally, limitations and suggestions for future researches are discussed to show the need for further analysis of the present study.

2. Research questions

After the breakdown of the worlds’ financial system in 2008, there was increased interest in people’s savings, borrowing, payment makings and risk managing. Financial systems play a crucial role in reducing poverty and inequality. By giving the poor the chance to save and borrow money, their lives can improve. While the increased access to credit is linked to benefits, disadvantages such as high level of debt should not be neglected. What has not yet to be considered, in detail, is how individuals decide whether they are going to repay their credits. In 2012, 22% of adults saved at formal financial institution, whereas 9% got a new loan from a formal financial institution in one year (Demirguc-Kunt and Klapper, 2012). From 2011 to 2012, the Eurozone showed decreased net savings of 1% and increasing net borrowings of 22,64% (OECD.Stat, 2014).

What are the exact reasons behind one’s low saving and high borrowing behaviour? The main problem is that, personal financial decision-making is not always based on receiving the highest outcome. Imagining a person with a long-term incentive to save some money for the future and short-term actions that completely differentiate from the actual desire, gives some insights into the variables that influence this decision making process. This action can lead to irrational behaviour (Harackiewicz et al. 2000). Thus, people tend to satisfy their immediate interest without respecting their long-term goal.

Considering the psychological impact on economic decisions provides a new perspective in explaining the behaviours that do not maximize outcomes. Risk aversions and time preferences influence an individual’s decisions, which can change accordingly (Sitkin and Weingart, 1995). Thus, people with different time preferences also do not decide constantly in the same way. This is known as time-inconsistency. Additionally, sight and magnitude effects^[3] provide evidence that the traditional discounting model of Samuelson (1937) is violated and that people do not discount different choices exponentially^[4]. Considering these preferences as a change in taste, leads to the question ‘by what factor this switch is caused?’ The impact of different time preferences leads to dynamic inconsistent behaviour, which is influenced by the presence. According to this, factors that could influence time preferences can highly vary. However this factor can be summed up as present-bias, which can simultaneously be seen as a self-control problem.

Factors that could abate this present-bias (i.e. self-control problem) can have multiple derivations. Loewenstein (2000) refers to them as visceral factors^[5] that alter desires. Looking for one factor that comprises internal as well as external impacts leads to ‘Religion’ (Fox and Calkins, 2003). Religious beliefs have an effect on individual and social behaviour (Greeley and Hout, 1999); they also form an individual’s moral standards (Wiebe and Fleck, 1980). Furthermore, it represents a long-term phenomenon that has developed over time. Thus, it is a valuable factor, which can determine behaviour (Kim et al., 2009). To analyse saving behaviours of people, it is crucial to find factors that have a direct influence on their behaviour. Because religion influences the daily life of people and impacts behaviour, the following will be analysed:

To what extent does religiousness influence the present-bias and thus the saving/borrowing behaviour of people living in Germany and Turkey?

The goal of this paper is to find out, whether religion indirectly influences the saving/borrowing behaviour of people living in Turkey and Germany; this is done by analysing the impact on present-bias. The investigation of further sub-questions is thereby supportive:

(1) Why is the comparison between Turkey and Germany value adding?
(2) What is exactly the present-bias and how can it influence people?
(3) How can religiousness and present-bias be measured?
(4) How can religion affect people’s saving behaviour?

The sub questions clarify the understanding of the present-bias and the measurements that are necessary for the underlying study. The following figure gives an illustration of the research problem.

illustration not visible in this excerpt

Figure 1: Impact of religion on present-bias and saving/borrowing behaviour

Hence, the impact religion has on the present-bias, which simultaneously influences the saving/borrowing behaviour of people, is the main underlying research study. Using the findings of existing studies show that time inconsistency is present and thus, rational behaviour cannot be constantly assumed. Analysing this relationship, gives indications of the invalidity of the traditional economic model, something that has proved by several studies (Thaler and Shefrin 1981, Loewenstein et al. 2002, Fehr 2002, Laibson et al. 2007, Meier and Sprenger 2010, Can and Erdem 2013). Additionally, the relationship between borrowing and saving behaviour of people and their present-bias also exists. A number of studies show that people with present-bias, have higher active borrowing levels and higher credit card debt (Laibson, 1997; Fehr, 2002; Meier and Sprenger, 2010: Heidhues and Köszegi, 2008). Therefore, this study presumes the relationship between present-bias and saving/borrowing behaviour and consequently, does not prove of this correlation.

Knowing to what extent religion influences the saving/borrowing behaviour of people helps financial institutions to approach religious people in an appropriate way in order to stimulate their saving behaviour. Furthermore, a deeper understanding of how people manage their personal finances can develop. Thus, people could improve their personal financial outcomes by focusing more on their long-term interests; this would also lead to increased national savings. Additionally, the importance of self-control for personal financial outcomes increases.

3. Comparison Turkey and Germany

The biggest religious groups in the world are Christians (31.5%) and Muslims (23.4%). Taking a Muslim country into consideration that does not represent a totalitarian ideology way of living his religion increases the chance to ask participants questions about religion, without their refusal to participate. In general, religion is a sensible topic in most Islamic countries and should therefore be handled with caution. Because most Islamic countries implemented the religion in their constitution, it is even harder to ask questions that measure their religiousness without questioning the participant’s whole approach to life (Nasr, 2005). Turkey is the only country with a large Muslim community (99% Muslims) in which Islam is not considered as the state’s religion and laicism is simultaneously anchored in the charter of a borough (Nasr, 2005). Considering the overall economic development of Turkey in recent years, shows that it is a rapidly developing country and is the largest national economy in Central and Eastern Europe (Trading Economist, 2014). With a GDP of $820-827 billion (2013) and a growth of 4% (2013), in comparison with other European member states, Turkey exhibits an above-average performance (Data Worldbank, 2014). The unemployment rate also decreased from 10.5 % (2004) to 9.8% (2013). The public debt of 36% of the GDP clearly underlies the Maastricht-threshold. The GDP based on purchasing-power-parity (PPP)^[6] per capita GDP of Turkey is also above the average compared to countries, having the Islam as the largest faith group. The average of these states according to the IMF is between $6,000.00 and $8,000.00 (2012 to 2013), whereas Turkey shows a higher wealth with $15,578.38 (2013). Even though Turkey’s economic growth slightly declined in recent years, looking at the whole picture is more important while selecting a Muslim country that shows a good size of economy and an above-average living standard.

Analysing saving/borrowing behaviour in an Islamic country where the population has the financial opportunity to afford savings is essential. The data about Turkey’s economic status represent the nation’s wealth and gives incentives to assume that people have the chance to save money. Otherwise, findings of saving behaviour could be strongly influenced by poverty, which would make it impossible to find a correlation between the interaction of religion and present-bias influencing the savings / borrowing behaviour of people.

Regarding the biggest denomination, 588 million Christians live in Europe (including Russia). Thus, Europe has the most Christians. A European country that is economically- and population-wise comparable with Turkey is Germany. Germany is the largest economy in the Eurozone and shows economic stability (Trading Economist, 2014). Being the second largest exporter in the world was the main driver of Germany’s growth in recent years. The record-low unemployment rate of 5.2% is next to Austria (4.8%), one of the lowest in Europe (Eurostat, 2014). With a population of 80.43 million (2013), it is almost comparable to Turkey (74 million in 2013). GDP (PPP) per capita GDP is $39,993.34 (2013), also above the Eurozone average of $34,510.94. Thus, comparing Turkey and Germany seems most appropriate according to the data of religious population, economic wealth and general development.

4. Literature Review of Present-bias

The rational^[7] behaviour of humans has been frequently question in recent years. H. Simon (1972) also appraised the rationality axiom; his exploration of the bounded rationality of people implemented a different perspective of explaining human behaviour. Thus, people use heuristics^[8] during their decision making process in order to shorten this procedure. Kahneman and Tversky (1974) provide several examples of the different use of these so-called heuristics. In doing so, they mainly use methods that refer to certainty, probability and possibility pattern of people’s behaviour. For instance, a phenomenon known as certainty effect, illustrates how people overweight outcomes that are considered certain, relative to outcomes that are merely probable. Thus, certainty about an outcome is preferred, whereas risky outcomes are neglected. Various examples of different heuristics are illustrated by Kahneman and Tversky (1974), but also by Pearl (1984), Gigerenzer (1999), Riel (1996), Gilovich et al.(2002), Schultze and Pfeiffer (2002), Epley and Gilovich (2006), Song and Schwarz (2009) and Nestler and Egloff (2009). Thus, a person’s feeling about a specific situation effects the decision making process, which can lead to irrational behaviour. Following Rubinstein’s’ (2003) suggestion and combining the psychological and economical levels of perspectives, gives a better understanding of the right use of economic models. Therewith irrational behaviour could for instance also appear when it comes to savings or spending of money. Examining savings as a choice on a specific point in time could also be influenced by a heuristic in which parts of information are neglected and the personal preference would be shifted into the focus (Carrillo and Mariotti, 2000).

Ainslie (1992) indicates that humans same as animals would prefer a smaller reward that can be received soon to a larger reward, which can be accepted later. A possible explanation for this behaviour would be that future rewards have been uncertain, which is simultaneously linked to higher risk (Fehr, 2002). The uncertainty of an event and thereby the risk that a person faces, can change over time and therewith leads to an inconsistent behaviour. Thus, people’s preferences change over time, which can be seen as time inconsistency (Rohde, 2010).

Other researchers such as Koopmans (1960) have already discussed the general idea of this paper. In his work on impatience he analysed the interest in time preference, which implies the privilege for receiving a desired outcome sooner, on an infinite time horizon. Böhm-Bawerk (1912) and Fisher (1930) gave inspirations for Koopmans (1960) work, by considering impatience as a characteristic of human economic preference in decisions, which include a finite time horizon. This impatience refers to a time preference, which can be understood as a present-bias that changes people’s behaviour. (Frederick, Loewenstein and O’Donoghue, 2002).

According to Strotz (1995), such behaviour entails a changing taste. People, who prefer $110.00 in 31 days instead of taking $100.00 in 30 days, also take $100.00 now instead of receiving $110.00 tomorrow. Even though people face a present-bias for a situation that compares the present with the future, they behave differently again when comparing the future with the future. At this point, people start behaving rational again. Therefore, a dynamic inconsistency exists, which makes it hard for economists to measure a model that combines all these different aspects of behaviour.

Myerson and Green (1995) supported these general findings with the aid of a study that considered the individual behaviour and showed that the temporal discounting was better described by hyperbolic function than by exponential function. This shows that people do not discount their options at different times in the same way and that preferences influence them over choices for now and for the future, this can be an explanation for different savings behaviour. This perception works against the traditional utility model of Samuelson (1937), which defined the utility discounting as exponential. According to this, the subjective discounting rate of a person is the same for different points in time and therewith, shows a time-consistency.

However, numerous researchers found evidence against this traditional discounting model of Samuelson (1937). Harris and Laibson (2001) for instance, figured out that the discount rates are greater for the short-run than they are for the long-run, which can be best expressed by the hyperbolic discount function. By using a more quantitative model, which includes buffer-stock literature, equilibrium concept and the hyperbolic Euler Relation, Harris and Laibson provide evidence that is more mathematical. Instead, Fishbein and Rubinstein (1982) represented evidence for the hyperbolic function by extending the risk aversion work by Pratt (1964) and Arrow (1965) with a field study that included the choice of different payment situations. A criterion for a non-stationary time preference and thus evidence for hyperbolic discounting was illustrated by Prelec (2004). Similar findings are also represented by the studies of Loewenstein and Prelec (1992), Kirby and Herrnstein (1995), Harvey (1995), Cairns and van der Pol (2000), and Thaler (1981).

The dynamic inconsistency simultaneously leads to a self-control problem, which is reflected in the impatience of a person and which is reflected in the present-bias problem. If the discount factor varies with time, then the short-term and long-term plan changes. Thus, a person could have saving plans for the future, but already violates them for the present (Strotz, 1956). Therefore, self-control has an impact on savings behaviour. By taking this knowledge of time discounting into consideration, the saving behaviour changes (Thaler, 1981). Previous researchers analysed credit- and consumption outcomes and found that a hyperbolic function gives more and better insights about the behaviour. By illustrating borrowing pattern they showed that hyperbolic functions are better predictors than the exponential model (Laibson, Repetto and Tobacman, 2005). The study of Meier and Sprenger (2010 gives strong evidence. Analysing the relationship between present-biased time preferences with credit card borrowing, shows that present-based individuals are indeed more likely to have higher credit card balances. The combination of a field study that included an incentive choice experiment and objective data on credit card borrowing from the major credit bureaus in the US showed that people with present-bias borrowed 25 % more than individuals with dynamical consistency (Meier and Sprenger, 2010).

Integrating time-preferences into a formulation, leads to the quasi-hyperbolic function of Laibson (1997) and O’Donoghue and Rabin (1999). It converges the hyperbolic discounting function into discrete times, including present (t) and future activities (t+1) to (T-t). Thus, the following holds:

Abbildung in dieser Leseprobe nicht enthaltenAbbildung in dieser Leseprobe nicht enthalten.

(Abbildung in dieser Leseprobe nicht enthalten) represents the subjective utility of a reward in which (t) illustrates the present and (T-t) and t+1 the future behaviour. Abbildung in dieser Leseprobe nicht enthaltenis in this sense the present-bias. Having a Abbildung in dieser Leseprobe nicht enthalten<1 constitutes the hyperbolic discounting function, whereas a Abbildung in dieser Leseprobe nicht enthaltenof 1 demonstrates no present-bias (i.e. self-control problem or time-inconsistency) and thus, relates back to the primary model of Samuelson (1937). Abbildung in dieser Leseprobe nicht enthaltenis the time in the future that is discounted to the present. The combination of Abbildung in dieser Leseprobe nicht enthaltenAbbildung in dieser Leseprobe nicht enthaltenpoints out the discount factor that compares a choice with the present and the future, whereas only Abbildung in dieser Leseprobe nicht enthaltenis considered when comprising two choices in the future.

4.1 Factors influencing Present-bias

Understanding an intertemporal choice^[9] is not easy because of the different influencing situations, which again are affected by various personal motives (Frederick, Loewenstein and O’Donoghue, 2002). These motives are in turn impacted by the present-bias. Therefore, planning for the future is easier than putting actions into practice in the present. Therewith short-term and long-term goals could change according to personal preferences. Main factor that plays a role is self-control; this can also be reflected in patience. It influences whether a person has the willpower to save their money.

In order to increase the national savings, it is important to emphasize people’s strong self-control. Whereas Muraven and Baumeister (2000) argue for a self-control muscle that needs to be trained, other psychologists argue for personal characteristics that influences self-control (Hester, 1995). This study considers the influences that are not only inherent. According to Loewenstein (1996) people’s self-control is mainly influenced by visceral factors, which compromise driver states (e.g. sexual desire, hunger and thirst, emotions and moods, pain) that impact the decision making process. Consequently, visceral factors show two main aspects: 1. a direct hedonic impact and 2. effect on the relative desirability of goods and actions (Loewenstein, 1996). This is partially supported by Fox and Calkins (2003), who focus on intrinsic and extrinsic factors that influence self-control. Thus, it is more about the combination of internal and external influences instead of just focusing on them separately. The intrinsic influence represents the individual’s temperament and cognitive processes, such as attention and inhibitory control, whereas the extrinsic factor indicates the person’s environment including various relationships to peers and family and the influence of cultural expectations. Summarizing Fox and Calkins’ (2003) study, the impact of emotions on self-control will be most fruitful if the interplay between the intrinsic and extrinsic factors is studied over time. In general, a person’s characteristics, environment, peers, and education could influence self-control.

One possible variable that comprises all these factors is religion. Cloud (2000) defines it as “(…) the habitual expression of an interpretation of life, which deals with ultimate concerns of values.” Additionally, the general meaning of religion can be disposed to each generation (Cloud, 2000). Religion comprises the intrinsic and extrinsic influence and it mirrors the people’s daily behaviour. Therefore, it not only influences social and consumer behaviour (Fam et al., 2002), but also impacts moral standards (Wiebe and Fleck, 1980). These aspects are researched by other authors, such as Greeley (1977), Uppal (1986), McDaniel and Burnett (1990), and Fam and Waller (2000). Religiousness is also a phenomenon that developed over time and hence, represents a long-term interest (Kim et al, 2004). By influencing important life feelings, such as marriage, nutrition, moral standards, and social association, religion plays a pivotal role in life. However, the norms and values vary according to the extent of religious belief and their involvement in life (Khraim, 2010). Thus, analysing the religiousness of people and the impact on the present-bias (i.e. self-control) could give further insights into national saving/borrowing behaviour.

5. Literature Review on Religiosity Measures

5.1 Problem of Measuring Religion

The general purpose of this paper is to find out if there is a correlation between the present-bias and religion. The factor religion is a broad and highly complex phenomenon, which covers a variety of meanings. Simultaneously, these meanings have a wide range of interpretations (Kirckpatrick & Hood, 1990). Psychologists working in the field of religion have adopted three approaches: unidimensional, bipolar and multidimensional. Religion is still very multidimensional in its construct and thus, can be considered from numerous perceptions, such as personal, organizational, behavioural, and cultural. Due to this complexity, classifications of the dimensions and categories of religion are inevitable. Numerous researchers deal with measures about religion. Therefore, constraints must be made while examining and measuring religion. The actual list of dimensions is multitudinous. Thereby, researchers can be classified into following scales while measuring the extent of a person’s religion: Religious beliefs and practices; Scales of religious attitudes; Scales of religious orientation; Scales of religious development; Scales of religious commitment and involvement; Scales of religious experience; Scales of religious/moral values or personal characteristics; Multidimensional scales of religiousness; Scales of religious coping and problem-solving; Scales of spirituality and mysticism; God concept scales; Scales of religious fundamentalism; Scales of death/afterlife; Scales of divine intervention/religious attribution; Scales of forgiveness; Scales of institutional religion; Scales of related construct (Hill & Hood, 1999).^[10]

Each of these dimensions includes a large number of different scale measurements, in which the researchers separately focus on different sub-aspects. Therefore, this heritage of scale measurements makes it challenging to find appropriate approaches. The actual challenge is to find out first, which dimension would be the best for the purpose of the study and afterwards, to find a scale that is useable for this. Hence, before starting the actual research, it should be clear what the purpose of the study is and which dimension would fit to it. By first analysing the basic concepts of researchers, who measure religiosity, a starting point for further analyses is given.

5.2 Overview of Previous Researchers on Religiosity Measures

Different conceptualizations of religion have been widely proposed by Thouless (1935), Allport (1950), Glock (1954, 1959, 1962), Golden (1960), Fukuyama (1961), Faulkner and DeJong (1966), and Gorlow and Schroede (1968). Traditionally religiosity was seen on a unidimensional construct. Thus, the person’s church attendance and denomination was the primary measure (Bergan, 2001). However, people who practice their religion extensively are not necessarily high in religiosity. The main reason for this could be the difference in their general incentive. Thus, practice could be seen more as a routine action than devotional (Khraim, 2010).

Glock (1962) essentially introduced the general term of different dimensions. Glock and Stark (1966) base their work on previous researches, such as G. Lenski (1961); they looked for a greater analytical precision to distinguish between the different religious expressions. Thereby, five dimensions evolved: 1. Personal religiosity (experiential); 2. Acceptance of the belief system (ideological); 3. Participation in religious activities (ritualistic); 4. Knowledge of the belief system (intellectual); 5. Ethical consequences of these dimensions and the descriptions derived from them (consequential). Whereas religious affiliation focuses on the distinction between different religious groups, religiosity rather considers the attitudes, behaviour and values of a person (McAndrew & Vaos, 2011)^[11]. Thus, the consequential dimension of Glock and Stark (1966) does not entirely show a clear understanding of the extent to which religious consequences are part of religious commitment. Nevertheless, Glock and Stark (1966) developed a comprehensive way of measuring religiosity. Thereby, they did not only use one or two dimensions of religiosity, unlike previous research, but rather included variety of dimensions.

The model of Glock and Stark (1966) also faced criticism. Weigert and Thomas (1969) indicated the overlap between two dimensions. According to their argumentation, belief and knowledge dimension are both part of the ideological domain. Apart from this aspect, the independence of the other categories is not questioned. Clayton (1974) even goes a step further by suggesting that all of Glock and Starks’ (1966) dimensions are aspects of religious belief. King and Hunt (1975) offer contradicting views by supporting strong evidence for the multidimensionality claimed by Glock and Stark (1966). Their extensive research program, using an inductive approach via the technique of factor analysis showed a similar scale to Glock’s belief, experiential and ritual dimensions. Additionally, Hilty (1988) also reported results that enforce a multidimensional approach.

Nevertheless, Glock and Stark (1966) were also not sure about the fifth dimension of their findings. Other researchers, such as Fichter (1969) and Fukuyama (1961) also declare this dimension. According to Fichter (1969), the first dimension, which inserts the personal religiosity (experiential), is more a psychological than sociological, due to its subjectivity, which is typical Protestant. This would mean that the dimensions are not really an abstract of religion but rather characteristics of its empirical manifestations in specific social contexts. However, it should be noted that the dimensions of a theoretical concept vary based on the researcher discipline (J. Billiet, 2002).

5.3 Religious Attitude and Orientation

By analysing the relationship between the present-bias and religion, it is important to find out, how a person’s religion influences behaviour. The individual’s attitude towards religion gives next to general religious beliefs, information about positive and negative feelings of these beliefs. It represents the evaluative reaction and possible insights of an individual’s orientation. Knowing how religion influences behaviour, gives more insights about the person’s way of thinking. Poppleton and Pilkington (1963) created the Religious Attitude Scale, which measures the general attitude toward Christianity. The agreement or disagreement with several statements about the necessity of religion for life, well-being and moral development of the respondent, is the composition of the scale. Due to similarity, but general approaches of previous scales, several researchers came up with a more specific way of measurement (Thurston and Chave (1929); Eysenck (1947); Brown and Lowe (1951)), Poppleton and Pilkington (1963).

A crucial aspect for this study is the intensity of the religion. The personal involvement and commitment plays a major role for the religious orientation. Thus, there is a distinction between how people use and live their religion (Allport, 1950). This differentiation between intrinsic and extrinsic involvement is mainly driven by the work of Allport (1950), which was revised in 1967, by Allport and Ross. People with the same religion could have different motivations. Therefore, it is important to differentiate the driving forces. Whereas one person could be driven by his own strong commitment to the religion, other people could be solely motivated by the religious behaviour of the peers.

Due to the social desirability bias, the answers of the participants can be influenced by the fact that they do not want to admit to unusual beliefs or practices (S. McAndrew & D. Vaos, 2011). Thus, in order to measure the extent of religion, it is crucial to consider questions that cover the individual intrinsic and extrinsic part. Another influencing factor is that people also exaggerate their attendance at religious services. Hadaway et al. (1993) compared the actual attendance of people in church with their self-reported attendance in a study. The result was a remarkable difference. One of the latest studies by Hadaway and Marler (2002) suggests that the actual number of people attending religious in the US service was 22% instead of the self-reported 40%. In Britain, a comparable outcome was found (Brierley, 2000). This data shows a general behavioural phenomenon, which can be seen in different countries. In general, it is a problem of overestimating one’s behaviour; this is studied by various researchers, such as J. Kruger and D. Dunning (1999), R. Baumeister (2002), and A. G. Düttmann (2007). Therefore, the inclusion of intrinsic and extrinsic items might overcome this bias (Hill & Hood, 1999).

What is named as intrinsic and extrinsic involvement by Allport and Ross (1967) is the implicit and explicit religion in the study of Nesti (1990). Thereby, implicit religion is a request for meaning that occurred by subjectivity and is expressed by means of symbols and practices, which is in turn known as explicit religion. Mueller (1980) instead contemplates religion as a cultural system in which extrinsic acts (e.g. involvement in social work, church attendance) cannot be seen as a dimension of religion. However, his theoretical reflection has no serious impact on survey research practice.

The work by Allport and Ross (1967) served as a foundation for varies scales. Gorsuch and Venable (1983) modified their work. The aim was the establishment of an age universal orientation scale that is useable for adults and children. It is evident that the general ideas are mostly adapted and modified according to the researcher’s concept. King and Hunts’ (1972) religious position scale entails two parts, the cognitive salience and the extrinsic religious orientation. While the first part is the deeper meaning of the religion for the personal life, the second part is responsible for the pragmatic role of religion. It is comparable to the intrinsic and extrinsic separation of religion by Allport and Ross (1967). Even though the scale items are mainly phrased for a variety of Christian and non-Christian groups, the questions essentially focus on church members and are hence, are not expressed in a universal way.

The intrinsic religious motivation scale by Hoge (1972) measures different ways of being religious. The Christian tradition influenced the establishment of the scale, but it is applicable to a wide range of other religions (Hoge, 1972). The scale measures the motivation behind religious activity. In general, it adds conceptual clarity to the relationship between religiousness and other factors of interest. The items of the scale are designed in a way that steps into the dimension of Hunt and King (1971), which is in turn derived from the motivational aspect of Allport and Ross (1967). One important idea of Hoge (1972) is that unlike Mueller (1980), he assumes that intrinsic and extrinsic faith lie on the same dimension with different ends, rather than assuming that intrinsic and extrinsic faith illustrates two separate dimensions.

Koenig, Patterson and Meador (1997) established the Duke Religious Index (DUREL), which measures organizational, non-organizational and intrinsic religiosity. Due to Hoge’s high correlation of 0.72 for the intrinsic religiosity scale, the researchers made use of this part of the scale. A completely distinct measurement from what the intrinsic and extrinsic scales measure is what is Batson and Schoenrade’s (1991) Quest. Finding the truth about the religion and being on a quest, is the main intention of the scale. Self-criticism and perceptions of religious doubts and openness to change seems to be significant (Batson & Schoenrade, 1999). The main question that arises by opponents, such as Hood and Morris (1985), is if the Quest scale measures anything about religiousness at all. Due to additional questionings about the reliability and validity of the scale, Batson and Schoenrade (1991) analysed studies that used the scale to contrast subgroups with different subjective degrees of religion. Sequentially, the Quest scale measures a specific dimension of religion.

5.4 Religious Commitment and Involvement

In order to measure the commitment and involvement of religion, it is recommended not to simply ask respondents about their extent of commitment to religion, but rather focus on specific aspects.

The religious emphasis scale by Altemeyer (1988) focuses on the significance and extent to which religion was emphasized by parents. The scale is useable for adolescent from similar cultures. A more general religiousness measure, compared to the work of Altemeyer (1988), was established by Sethi and Seligman (1993). By correlating religiosity with attributional styles and optimism, Sethi and Seligman (1993) focused on three main aspects: 1. Religious influence in daily life; 2. Religious involvement; 3. Religious hope.

Roof and Perkins (1975) developed the salience in religious commitment scale. It is similar to the idea of the intrinsic religion, in which the individual importance of being religious and being attached is questioned. Thereby, the researchers looked at the extent to which adults consider their belief important. The scale contains three parts; two multiply choice formats and the last part entails a statement about religion with a Likert-scale for answer opportunities. The Chronbach’s alpha of 0.72 and the correlation coefficient of 0.81 between the salience measure and the doctrinal commitment show evidence that the salience scale is indeed measuring some characteristics of religiousness. However, the general usefulness of the scale is still questionable, due to his narrowed range of scores. Thus, if a respondent’s scores 10 or 11, on a scale of 11-points, he/she is considered to have a level of religious salience.

A different modification of the scale developed by Allport and Ross (1967) is measured by the religious commitment scale by Pfeifer and Waelty (1995). The researchers conducted a study with psychiatric patients in order to test the religious commitment. 51 statements that can be answered with yes or no serve as a measurement of religiousness. There is no specific test of reliability available. The correlation between the religious orientation scale by Allport and Ross (1967) and the religious commitment scale by Pfeifer and Waelty (1995) is 0.88. The following represents an overview of the most relevant measurements.

[...]

---

^[1] Affective system: Experience of feeling or emotion (Hogg, Abrams and Martin, 2010)

^[2] Discounting: Determining the present value of a payment, which is to be received in the future

^[3] Sight and magnitude effects: Represents a habit of people, who discount gains more than losses and small outcomes more than large one’s (Loewenstein et al., 2002)

^[4] Exponential discounting: Assumes a constant discount rate in time (Can and Erdem, 2013)

^[5] Visceral factor: Refer to a wide range of emotions, drive states and feeling states that grab people’s attention and motivate them to engage in specific behaviors (Loewenstein, 2000).

^[6] Purchasing Power-Parity (PPP): Compares differences in living standards between countries as a whole by including relative cost of livings and inflation rates of countries.

^[7] Rational behavior: Implies consistent maximization of a well-ordered function, such as a utility or profit function (see Becker, 1962)

^[8] Heuristics: Mental short cut that allows people to solve problems quickly and effectively. Thus, they do not have to think constantly about their next course of action (see Psychology about, 2014).

^[9] Intertemporal Choice: This kind of choice illustrates how the current decisions made by an individual can affect the options that become available to them at a future time. First introduced by Adam Smith (1776)

^[10] All possible dimensions are summarized in the measures of religiosity by Hill & Hood (1999). Hundreds of scales are documented.

^[11] McAndrew & Voas (2011) summarized the available major datasets for secondary analysis and explain the importance of religion