Explanation of the Trends and Recent Changes in Spanish Society Regarding Belief in God: Atheism, Agnosticism, Deism, Skepticism, and Belief

: This article deals with the different religious counterpoints or beliefs that traditions of philosophy and thought have fundamentally deﬁned around ﬁve positions or options: atheist, agnosticism, deism, skepticism, and belief. The aim is to explore the sociological conditioning factors behind these proﬁles, offering a meaningful explanation to help us understand the trends and recent changes in Spanish society. To this end, we analyzed two surveys, for the years 2008 and 2018, of the International Social Survey Programme (ISSP) on Religion in the context of Spain. The results highlight the variables of gender, age, and ideology to describe speciﬁc proﬁles, explain religious choices, and understand trends. In addition to the trend towards secularization, the results manifest an increase in the number of atheists and a decrease of believers, show increasing radicalization and right-wing ideology among believers, and a greater grounding of the meaning of life in the existence of God alone.

In this classical and contemporary arcanum lies the realm of beliefs traditionally ascribed to four levels of analysis: deists, who believe in a god or an all-powerful mind; believers, who believe in God; agnostics or skeptics, who do not deny the existence of God, but either have doubts or point to its cognizable impossibility (Eller 2004); and atheists, who do not believe in God (Baggini 2021).Nuances are established between these rigid positions, but analytically they are difficult to discern (Smith 2016;Stark et al. 2005;Oppy 2009;Barnes 2003;Le Poidevin 2010).
If we frame current Spanish society within the model of modernization and secularization prevailing in the Western world, and essentially in Europe, we believe that the conditionality of these beliefs is essentially based on the socio-demographic and socioeconomic characteristics of Spaniards.That is, Spaniards' denial or doubting of the existence of God is conditioned, to a greater extent, by their lives and socio-historical trajectory.The former can be known empirically through their opinions, attitudes, beliefs, and behaviors, while the latter must be understood within the social and historical heritage of the different generations.Berger (2011) points out that socialization guarantees an uninterrupted consensus about the primary characteristics of the social world, and ensures the strength of the legitimation of the transmitted contents, particularly through religion.However, Demerath (2000) and Zuckerman (2009Zuckerman ( , 2020) ) rightly point out that millions of people are culturally religious through their identification with a religious tradition, but without believing in the religious content of the tradition, a trend that is observed explicitly in the Spanish case.
This type of research is not new at a global level (Joas and Wiegandt 2009), nor has it been sociologically explained through socio-demographic and socio-economic variables (Norris andInglehart 2012, 2015;Bruce 2003;Zuckerman 2009).However, it does represent a novelty in the Spanish case, where research has been scarce (Montero 1986;Llopis 2002;Bericat 2008;Pérez Agote 2010a, 2010b).Nevertheless, there are relevant databases held by the Centro de Investigaciones Sociológicas and the Observatorio del pluralismo religioso en España (Observatory of Religious Pluralism in Spain), initially created by public initiative in 2011 and, since 2021, part of the Fundación Pluralismo y Convivencia.
Pérez Agote (2010b) summarized the new socio-political context and recent religious changes in Spain represented by three trends: (a) the sweeping individual secularization seen among the Spanish population, whose non-believing Catholic positions (deism, agnosticism, and atheism) arrived later than in northern European countries; (b) the separation between Church and State, which is experiencing major difficulties due to the relevant power and influence of the Catholic Church; (c) the recent settlement in Spain of a large immigrant population with non-Catholic religious beliefs and experiences (Orthodox, Muslims, etc.).Our research question is set in the following context: given that Spain, by history and tradition, is essentially a Catholic country, what factors or variables intervene so that part of the population transgresses this tendency towards deist, agnostic, and atheist positions?For example, empirical research has shown that several social factors determine significant shifts from theistic, agnostic, and atheistic belief positions, including national trends (Zuckerman 2009(Zuckerman , 2020)).Other recent empirical research has found, for example, that intelligent, i.e., more educated, people may have less need for religious beliefs and practices (Zuckerman et al. 2013).
Apart from the well-recognized trend towards secularization in large parts of Western society, sociological attributes have much to say about this transfer of belief or spirituality (Stark et al. 2005).In this sense, we believe that this change is conditioned by Spanish social structure (Itçaina 2019).Furthermore, we expect socio-economic and socio-demographic conditions to influence and affect these attitudes.
To this end, in this study we present a descriptive analysis of the most significant theories on deism, agnosticism, atheism, and Christianity, from a philosophical perspective, and we take a sociological approach using data based on the most significant socio-demographic and socio-economic profiles of Spaniards.We subsequently conduct explanatory multiple regression analysis regarding the change in the trend towards more deistic, agnostic, or atheistic positions in the Spanish population in the decade between 2008 and 2018.

Deism, Agnosticism, Atheism and Christianity
The deistic approach first appeared in pre-Socratic Greece, in the philosophical triad initiated by Cicero and later followed by Hume: "Anaxagoras, the disciple of and successor to Anaximenes, was the first thinker who asserted that the orderly arrangement of the universe was designed and realized by the rational power of an infinite mind" (Cicero 2008, p. 21).During the Enlightenment, Hume, following the argumentative triad model, opted for deism as the most likely choice.Asserting that the whole system of nature that we observe speaks for itself in favor of an intelligent author; the order of the universe is proof of the existence of an all-powerful mind (Hume 2021).Along the same lines, Voltaire (2021) reaffirmed that a pure and eternal religion could not produce evil.Kaufman (2009, p. 23) spoke of a natural god as the very "creativity" of the universe, his position considers God as that awesome mystery of serendipitous creativity manifesting itself throughout the universe from the Big Bang onward, first in cosmic and biological evolutionary developments, and then ultimately producing the human world of history, culture, highly complex levels of symbolization, and elaborate technologies.
Cicero introduced skeptical or agnostic positions, arguing that among the divergent opinions of uneducated people, none of them may be true (Cicero 2008).Huxley, a freethinker and exponent of agnosticism, concretized the argument by establishing a methodology following that of Descartes.In matters of the intellect, one should follow reason as far as it leads without regard to any consideration.Similarly, conclusions reached should not be claimed to be true if they are not proven or demonstrable.This would become the essence of principled agnostic faith, to which we should add Clifford's (Clifford 1999) contribution that believing something based on insufficient evidence is always wrong, everywhere and for everyone.
Atheism is based on knowledge of nature, which in turn is based on reason and experience and criticism of superstition, religious beliefs, and ignorance (Holbach 2014).It is also based on socializing influences (Mill 2019(Mill , 2020) ) or the supplanting of God by human reality (Sartre 2017).The idea would later develop into different versions (Rowe 2010) and interpretations, from friendly atheism (Martin 2006) to militant atheism (Sabater 1989).More contemporary interpretations include new atheism (Dawkins et al. 2019), which followed in the same wake as nineteenth-century atheism: religion must be tolerated; however, it must be countered, criticized, and exposed to rational argumentation in all its manifestations.In essence, the shift between theism, agnosticism, and atheism and Christianity is based on the decline of intolerance from the 19th century onwards and is essentially the result of the struggle against religious beliefs developed from the common sense of logic, and in the current era, from science (Mill 2019).
Believers base their belief in God on the tradition of belief, based on general assent and intergenerational feedback and the thrust of science and its exceptionality.Concerning scientism, Pascal (2019) affirmed that faith tells us what the senses ignore, being above them but not against them.This is also the position of Enlightenment thinkers such as Von Haller and Leonhard Euler, who found no contradiction between religion and science, but rather between reason and the eudemonism practiced in atheism (Smith 2010).Von Haller (2009) claimed that this new wisdom was the ruin of social life because it offered only particular and sensual happiness.His arguments were further developed by Caputo (2019Caputo ( , 2021)).James (2014) ascribed belief to the legitimacy of faith, voluntarily accepted, and Ratzinger (2006) adds religion as an experience, far from the space of 'ratio'.

A Sociological Approach to Religious Counterpoints
From a dichotomous approach, Collins (1992) states that there are only two positions on religion, according to whether one is a believer or not: those who assume that there is a supreme reality that transcends the social and those who understand it as an irrational superstition about things that do not exist.Collins points out that most social thinkers hold the second position."Utilitarians and rational reformers, in general, tended to look on religion as an archaic, irrational force. . . .Legal reformers saw religion as an institution of Inquisitors and heresy-hunters. . . .Radicals saw religion as the upholder of the status quo" (Collins 1992, p. 30).
Indeed, in classical sociology, counterpoints take a back seat due to different approaches: comparative socio-historical analysis (Weber 1992(Weber , 2018)), customs and ritualism (Durkheim 2008), materialism (Marx 2015a(Marx , 2015b)), conciliatory religion (Simmel 2012), symbol system theories (Geertz 2017;Bellah 2011).It could be argued that if God represents society, and therefore different types of society have different types of gods as has happened in evolutionary terms, in advanced societies the idea of the gods may condition the idea of God.As happened in evolutionary terms, the idea of God in advanced societies may be conditioned by the social structure, since logically the prevailing individualistic effect shapes a more personal and abstract idea of deity (Goffman 2017(Goffman , 2022)).Nonetheless, in our Western context, non-rational feelings of group solidarity, which have been violated intergenerationally during socialization and expressed in supernatural religious beliefs, continue to persist in a large part of the population (Berger 2011).This is even apparent in negative attitudes, both institutionally and personally, toward non-believers (Weber 1992;Zuckerman 2009).
More specifically, Berger (2011) insists on his concept of plausibility structure, or real social basis, taken for granted for the continuation of societies.Applied to any faith and religious institution, it is internalized by successive generations through the process of socialization, in the construction of their own realities.Over time, the discontinuity of plausibility creates uncertainties or gaps in that truth, evident in terms of faith and religious institutions.Berger (2011) explains the concept through the example of the religious world of pre-Columbian religious legitimations, confirmed in the collective activity taking place within the framework of that world: "Subjectively, they were real to the individual whose life was embedded in the same collective activity. . . .when the conquering Spaniards destroyed this plausibility structure, the reality of the world based on it began to disintegrate with terrifying rapidity" (p.58).

Methodology
In line with our research question concerning the factors or variables involved in part of the Spanish population adopting deist, agnostic, or atheist positions, being a historically Catholic country, our general objective was to analyze which socio-demographic and socio-economic variables determine this trend.
Firstly, the specific objectives were to establish the socio-demographic and socioeconomic profile of religious counterpoints in Spain: atheists, agnostics, deists, skeptical believers, and believers.Secondly, from an explanatory point of view and through logistic regression models, we aimed to estimate which structural or independent variables best explain the changes in the trend towards positions of atheist, deist, agnostic, skeptic, or believer in the Spanish population in the decade from 2008 to 2018.
The established hypotheses are that Spaniards' specific socio-demographic and socioeconomic characteristics have a greater incidence in some religious positions than others.Likewise, it is considered that not all the explanatory or independent variables analyzed (gender, age, marital status, ideology, and status) behave in the same way over time and for each religious position (atheist, deist, agnostic, skeptic, or believer).
We used the database produced by the Centro de Investigaciones Sociológicas (CIS) from the last two waves of the International Social Survey Programme (ISSP) study on Religion.This transnational collaborative programme conducts annual surveys on various topics related to the social sciences (Fernández-Prados et al. 2019).Our study used data from the surveys Religion II (ISSP) Study number 2776, October-November 2008, and Religion III (ISSP), Study number 3194, October 2017/January 2018 (Smith and Schapiro 2021).Although the second survey sample analyzed was somewhat smaller and therefore had a greater level of sampling error than the first, the methodological guidelines and technical specifications of the studies were similar (see Table 1).The survey question used in the research on religious difference as a dependent variable was the same in both questionnaires.In the ISSP Religion III questionnaire of 2008 (ISSP Research Group 2018), it was Question 16 of the survey, and in the ISSP Religion IV questionnaire of 2018 (ISSP Research Group 2020), it was Question 42 (Q13 in the English version).In both cases, the question was as follows: "Please indicate which statement below comes closest to expressing what you believe about God [please tick one box only] 1.I don't believe in God; 2. I don't know whether there is a God and I don't believe there is any way to find out; 3. I don't believe in a personal God, but I do believe in a Higher Power of some kind; 4. I find myself believing in God some of the time, but not at others; 5.While I have doubts, I feel that I do believe in God; 6.I know God really exists and I have no doubts about it; 7. Doesn't know; and 8.No answer".
This question was recoded to clarify concepts better, which generated some methodological and conceptual doubts, as indicated in the introduction.Category 1 responses were ascribed to atheists, Category 2 to agnostics, Category 3 to deists, Categories 4 and 5 to skeptic believers, and Category 6 to believers.

Descriptive and Comparative Analysis between 2008 and 2018
In Spain, religious attitudes have undergone significant changes in a decade, maintaining an ontological logic associated with secularization processes, i.e., positions of belief are decreasing while positions of non-belief are increasing (see Table 2).Despite this general trend, believers were in the majority in both surveys (38.8% in the former and 30.6% in the latter).The above results notwithstanding, one in five believers have ceased to be believers.Skeptic believers and deists remained at similar percentages during the decade under review.On the other hand, the rate of agnostics grew slightly with an increase of 8%, and that of atheists increased considerably by 73%.An outline of socio-demographic profiles and their evolution according to each of the religious options, row and column percentages have been calculated by crossing the variables.Tables A1 and A2 are presented in Appendix A and include detailed percentages.From those, we highlight the following.
According to the gender row percentages, the majority option among women was that of believer.In the case of men, the believer option predominated in the first survey, while ten years later, the skeptical position was most frequent.The youngest respondents were mostly skeptical, although the position held by more than a quarter of 18 to 35-year-olds in the second survey was atheist.Adults aged 36-65 shifted from mostly believers to skeptic believers, and those aged 65 and over remained believers, albeit with a significant decline.In terms of education, there was a dichotomy between the lowest educational levels, those of "no education" and primary education, where the most frequent option was believer, with a decreasing tendency, and the higher educational levels where the majority opted for the skeptic believer stance, tending to increase slightly in the latter survey.Among married and separated people, the believer option remained the majority; among widowed and divorced people, the believer option was outnumbered by the skeptic believer option; and among single people, the atheist option.In general, there was a tendency towards skeptical believing positions, although in some socio-demographic strata, such as young men with higher education, left-wing ideology, and from the upper classes, non-belief and primarily atheistic positions were already the most frequent.
The column percentages allow us to define each religious profile and their evolution more clearly.Considering the over-representation of percentages concerning the marginalized, the following profiles and trends can be observed among the believers (see Table 3): female, from adult to older adult, with no education or primary to secondary education, married or widowed, from the center or right-wing ideology, particularly the right, from the working classes, particularly from among the old middle classes.Women also predominate among the skeptic believers, but in this case, they grew from young people to adults, with secondary to higher education levels, married, from the center and the right that has moved to the center, and from the middle to upper classes.Likewise, young, highly educated women with left-wing ideologies from the upper and middle classes stand out among the deists, though their marital status had evolved from married to divorced, separated, and single in the latter survey.Finally, atheists and agnostics kept a similar profile, men, young people with secondary or higher education, single or separated, left-wing, and upper class.However, those identifying as atheists showed a more defined profile with higher percentages.

Workers and old middle classes
Source: Author or authors' elaboration.CIS.

Explanatory and Comparative Analysis of 2008 and 2018
Table 4 below presents a set of logistic regression analyses explaining and predicting significant associations between variables.Thus, the dependent variables were the five dichotomized religious positions (atheist-non-atheist; agnostic-non-agnostic, etc.) and the six socio-demographic items (gender, age, education, ideology, marital status, and social class) as independent variables.Qualitative socio-demographic variables have been recoded as categories with "female" gender and "married" marital status being the options selected.A total of ten logistic regression analyses were conducted, two for each of the five religious positions (one set from the 2008 survey, and one from the 2018 survey).
The results show that the model or the set of the six variables explains little, although the R 2 achieved in the regression analyses of atheists and believers stands out somewhat.In contrast, the values of the predictors are, in most cases, significant or highly significant.In the case of the atheist position analyses, gender, ideology, age, and marital status are particularly significant.Education was the only variable that ceased to be significant between the first and the second survey.In comparison, the agnostic option lost three of the four variables that were significantly associated with sex alone.Deists can be predicted according to their ideology, as can skeptic believers, although the sense of sign changes in the latter, i.e., they are more associated with right-wing ideology.Finally, the belief position is predicted by most of the socio-demographic variables applied, with ideology, age, and sex standing out as highly significant in a very similar way as for the atheist stance, but with the opposite signs and increasing the statistical determinants and significance in the independent variables.The significant predictor variable in virtually all religious positions in the surveys is ideology.Likewise, gender and age significantly explain the dependent variables in most models and religious positions, especially in the "extremes" where they are very significant: believer and atheist.

Some Trends
In order to assess the trends, a graph (Figure 1) was developed to represent the different religious positions in the two surveys.The graph shows the data of the two most significant continuous independent variables (age and ideology) based on the two Cartesian axes according to their standardized values (z-scores).The result is a 10-point scatter plot with the five positions in the two surveys, and four quadrants.The upper right quadrant represents the older, right-wing ideology and the lower left quadrant represents the younger, left-wing ideology.Religious positions are placed in these two quadrants, with believers and skeptic believers in the first quadrant, and atheists, agnostics, and deists in the second (see Figure 1).
The trends in the positions of "non-believers" in a personal God, atheists, agnostics, and deists, experienced a movement between the two surveys that more closely adjusted to the average age of the population, which represents ageing, especially among agnostics, and to a particular ideological moderation leading them to move slightly to the "center".On the other hand, the tendencies of the "believing" religious position showed a clear tendency toward right-wing ideology, and although this population is also ageing, as the descriptions indicate, it is not as old as the overall population, hence its "downward" tendency along the axis of age.Likewise, the skeptical believers show a slight tendency towards the ideological right and evident ageing that places them in the quadrant of the oldest people.
Another graph was also developed (Figure 2) to represent the different religious positions, according to two axes or items that questioned the meaning of life: "only because God exists" and "it only has the meaning that each person gives to it".The Z-scores for the two Likertformat questions were calculated, and four quadrants were defined, with the five religious options being placed in only two quadrants.On the one hand, in the upper left quadrant, defined by greater importance of the meaning of life for oneself and less significance accorded to the existence of God, are atheists, agnostics, deists and skeptic believers; on the other hand, in the lower right quadrant, defined by the lesser importance of the meaning of life for oneself and greater importance for the existence of God, are believers alone.The trends in the positions of "non-believers" in a personal God, atheists, agnostics, and deists, experienced a movement between the two surveys that more closely adjusted to the average age of the population, which represents ageing, especially among agnostics, and to a particular ideological moderation leading them to move slightly to the "center".On the other hand, the tendencies of the "believing" religious position showed a clear tendency toward right-wing ideology, and although this population is also ageing, as the descriptions indicate, it is not as old as the overall population, hence its "downward" tendency along the axis of age.Likewise, the skeptical believers show a slight tendency towards the ideological right and evident ageing that places them in the quadrant of the oldest people.
Another graph was also developed (Figure 2) to represent the different religious positions, according to two axes or items that questioned the meaning of life: "only because God exists" and "it only has the meaning that each person gives to it".The Zscores for the two Likert-format questions were calculated, and four quadrants were defined, with the five religious options being placed in only two quadrants.On the one hand, in the upper left quadrant, defined by greater importance of the meaning of life for oneself and less significance accorded to the existence of God, are atheists, agnostics, deists and skeptic believers; on the other hand, in the lower right quadrant, defined by the lesser importance of the meaning of life for oneself and greater importance for the existence of God, are believers alone.Religious positions are indicated on this scatterplot on a continuum above the bases that represent where meaning to life was located.The atheist, agnostic, skeptic believer, and deist positions, at least in the dimension or axis of the greater meaning of life given by God, have tend to approach or adjust towards the axes of one survey then the other, i.e., to "moderate" their religious positions based on the meaning of life.In contrast, the Religious positions are indicated on this scatterplot on a continuum above the bases that represent where meaning to life was located.The atheist, agnostic, skeptic believer, and deist positions, at least in the dimension or axis of the greater meaning of life given by God, have tend to approach or adjust towards the axes of one survey then the other, i.e., to "moderate" their religious positions based on the meaning of life.In contrast, the believing position has tended to polarize by increasing its distance in the latter survey, reflecting the increased importance of God's existence and reduced importance of self in giving meaning to life.Socially, there is proximity on this Cartesian map between the atheist, agnostic, and deist positions, which are placed at one end of the continuum, while believing skeptics tend to be in the middle and believers at the opposite end of the range.

Conclusions
The statistical work carried out has allowed another approach to a long-standing and complex field of study rooted in philosophical debates as old as they are current.In any case, at least three conclusions can be drawn that open new challenges or future lines of research.
Firstly, the conceptualization of different religious positions encounters the methodological limitations of asking about them in standardized survey research questionnaires.The rich nuances of the history of thought, the history of Spain, and personal trajectories are lost when they are transformed into a sociological questionnaire or a question with five or six alternatives.Furthermore, the analysis of the results of the surveys made it necessary to focus interest on the evolution and explanation of the two extreme positions, atheists and believers, due to the growth of the former group and the reduction of the latter in a distinct process of secularization, and because they are the ones most significantly associated with the independent or explanatory variables (Stonawski et al. 2015).The results have also shown that religious positions considered theoretically or philosophically close to each other are further apart sociologically; thus, agnosticism and deism are closer to atheism than to skeptic belief, both in terms of the meaning of life given and the sociological profile analyzed.
Secondly, the results of the sociological explanation offered by the independent or socio-demographic variables of each religious position focus attention on three in particular: gender, age, and ideology (Froese and Bader 2008).The former requires a more detailed analysis than presented in this study.However, a differentiated secularization process is discernible among women who have moved from belief to skepticism and deism rather than towards atheism or agnosticism, positions which have a more masculine bias.Age and ideology correlate, so that they build a continuum from atheists (young and left-wing) to believers (older and right-wing).Stonawski et al. (2015) reported on the age structure of the Spanish population by religion and religiosity, revealing that older age groups are dominated by Christians of high religiosity, while those with no religion have a younger age profile, and minority religions are younger still; the youngest age groups are somewhat more religious, which follows from the greater fertility of the more religious coupled with intergenerational transmission from parents to children.On the other hand, recent study points to the emergence of a "culturalized religion" and its potential connection with the expansion of the radical right in Spain (Ruiz Andrés 2022).The relevance of these last two variables in explaining the different religious positions highlights the importance of generational evolution on the one hand, and on the other the historical roots of religious beliefs associated with conservative ideologies in the Spain of national Catholicism.
Finally, the most striking and obvious trend found was the sharp increase in atheism and the drastic reduction in the percentage of believers over the decade.Although the process of Western secularization can explain this, it perhaps requires further comparative analyses involving other countries in a similar cultural environment in order to gauge more precisely the pace of this process, which seems to be accelerating (Norris and Inglehart 2015).Another singular trend observed in the study of the two surveys was the ideological moderation of the non-believing religious positions and the radicalization of the more be-lieving positions.Ribberink et al. (2018, p. 220) conclude in their analysis that anti-religious attitudes are strongest among the non-religious in countries with a Catholic religious heritage, such as Spain; the same cultural climate that encourages nominal affiliation of the religious encourages opposition to religion among the non-religious.That is to say, the movement of believers in Spain to the right of the political spectrum, along with definitions of the meaning of life based more exclusively on the existence of God, point to a process of polarization that requires further analysis focused on this issue alone.

Figure 2 .
Figure 2. Religious options according to the importance of the meaning of life in God or each other (At = Atheist; Ag = Agnostic; De = Deist; Sk = Skeptic believer; Be = Believer).

Table 1 .
Fact sheets of the studies.

Table 2 .
Religious options in Spain between 2008 and 2018.

Table 3 .
Profiles of religious options and their evolution.

Table 4 .
Determinants according to logistic regression analysis for each belief type.

Table A2 .
Socio-demographic and socio-economic characteristics of Spaniards in 2008 and 2018 (column percentages).