Confucian Identification, Ancestral Beliefs, and Ancestral Rituals in Korea

Abstract

Since Koreans do not consider Confucianism to be part of religion, conventional religious identification questions cannot accurately capture the number of Confucians in Korea. Using the Korean General Social Survey and other data sources, we aim to describe the identification, beliefs, and practices related to Confucianism, especially ancestral rituals, and to examine whether these beliefs and practices differ across religious groups. Contrasted with 0.2% of the adult population identifying their religion as Confucianism in the 2015 Korean Census, 51% considered themselves as Confucians when asked, “(Regardless of your religious affiliation) do you consider yourself a Confucian?” If we consider those who think that rites for deceased family members are Confucian, the proportion was 44%. Considering those who conduct ancestral rites at a gravesite as Confucians, the proportion was 86%, but was only 70% when we count those who perform ancestral rites at home as Confucians. We also found substantial differences among religious groups. In general, Buddhists were most likely and Protestants were least likely to identify with Confucianism, believe in the power of ancestors, and perform ancestral rites. Perhaps most telling is the result of religious none falling in the middle between Buddhists and Protestants in terms of identification, beliefs, and rituals of Confucianism. The differences of religious groups appear to reflect religious syncretism and the exclusivity of religion. It is overstating to declare a revival of Confucianism, but it is reasonable to say that Confucianism is not a dying tradition in Korean society.

Confucianism is the backbone of Korean culture. According to Tu (1998, p. 30), “Among all the dynasties, Chinese and foreign, the long-lived Choson(Yi) in Korea (1392–1910) was undoubtedly the most thoroughly Confucianized”. Confucianism as a state ideology has permeated the everyday life of Koreans for more than 500 years. Despite this, before 1950, 96% of Koreans did not identify with any religious group (Baker 2013b). During Japanese colonization (1910–1945) and the modernization period after the Korean War (1950–1953), Confucianism was treated as a backward ideology and was severely criticized. Recently, in the 2015 Korean Census, the Korean population aged 18 years and over comprised of 19% Protestants, 17% Buddhists, 8% Catholics, 0.8% other religious group, including 0.2% Confucians, and 54% religious none. The small number of people who identified Confucianism as their religion support that the outlook for Confucianism appears bleak (Ro 2016, p. 263).

Having noted that surprisingly very few people in China, Japan, Korea, and Taiwan reported Confucianism as their religious identity, Sun (2013, pp. 110–19) raised important questions about how to identify Confucians. She argued that conventional religious identification questions in surveys originating from a monotheistic religion, such as that of the Judeo-Christian tradition, cannot appropriately capture Confucians. She addressed this oversight by suggesting to examine ancestral ritual, a representative ritual of Confucianism in East Asia, as an alternative indicator for identifying Confucians.

Both conventional Confucian identification questions in surveys and performance measures of religious rituals can fail to represent people who conceive themselves as Confucians in Korea. There are two main types of conventional religious identification questions. The Korean Census, performed every 10 years since 1985, asks, “do you have a religion? and if you have one, what is it?” The response categories are Buddhism, Protestantism, Catholicism, Won-Buddhism, Confucianism, Chondogyo, Daesoonjinrihoe, Daejonggyo, and Other, please specify (see Kim et al. 2009 of the slight variation of Census questions from 1985 to 2005). Confucians identified in these Censuses accounted for 0.2% of the population in 2015, 0.3% in 2005, 0.6% in 1995, and 1.6% in 1985. Meanwhile, most other surveys do not include Confucianism as a response category, and those who want to identify as such need to answer “Other” or “Other (please specify).”The main problem with these religious identification questions is that Koreans, whether they conceive of themselves as Confucian or not, do not consider Confucianism a religion (Baker 2010, p. 68).

Besides these problems of identification questions, until recently, there has been a limitation in using ritual practices as an indicator for religious identity, as Baker (2013a, p. 190) noted.

“Traditional Korean religiosity was centred on ad hoc ritual-based communities. Participation in rituals was not normally taken as a sign of a long-term commitment to a particular religion. A person could participate in a Buddhist ritual one day, a shaman ritual the next and a Confucian ritual the day after without feeling any contradiction and without being seen as a Buddhist, a Shamanist or a Confucian exclusively.”

(Baker 2013a, p.190)

Few studies have examined Confucian identification or ancestral rituals empirically. As Sun (2011) noted, previous works that dealt with changes of ancestral rituals were disproportionally based on ethnographic studies. In addition, previous studies on ritual change are based on a small number of interviews (Kim 2014a). In addition, Kaplan (2018) examined the vitality of Confucianism in terms of Confucian schools or programs. Due to these limits of data collection, reliable knowledge on the prevalence and correlations of Confucian identification or ancestral rites is lacking.

Using several different data sources, we aim to describe Confucian identification, beliefs, and practices. We focus on the relationship of these beliefs and practices with religious groups of Buddhists, Protestants, Catholics, and religious none.

Our paper fills a gap in the literature of Confucianism. First, using five recently updated data sources in Korea, we describe the various beliefs and practices related to ancestors or ancestral rituals. As far as we know, this is the first Korean study to provide a comprehensive empirical portrait about ancestral rituals as an indicator of Confucian identity. Second, the original religion in Korea was folk religion (Shamanism and Animism), and, in chronological order, Buddhism, Confucianism, Catholicism, and Protestantism entered Korea (Grayson 2002). Investigating Confucian identification and ancestral rites across religious groups sheds light on religious syncretism, which is a not rare but neglected aspect of religious identification (Hackett 2014; Kim 2005). This study also provides insight into how national religious context influences the effect of individual religiosity on attitudes and behaviors (Gebauer and Sedikides 2021; Finke and Adamczyk 2008; Stark 1996). Finally, our paper contributes to religious studies in East Asia, where a conventional religious identification question is not a suitable tool for measuring religion (Woodberry et al. 2020).

1. Confucian Rituals

There are four rites of passage, coming of age, marriage, mourning, and ancestral rituals, that Koreans perform in a Confucian style. According to Lee (1989), these four rites changed in the 20th century. The coming-of-age ceremony was abandoned in the early 20th century. The wedding retains only parts of the traditional ceremony, being performed in wedding halls instead of the bride’s home. Lee noted that funeral rituals changed the least compared with other rites of passage up to the 1980s. Ancestral rites consist of communal rites, death day rites, seasonal rites, and holiday rites (Lunar New Year or Korean Thanksgiving). Of the four family rites, ancestral rites are considered the most important because they are the manifestation of filial piety, which is the core value of Confucianism (Park 2010; Kim 2007, p. 164). Ancestral rituals are repayment for parental love, and descendants believe that “ancestors became family gods who assured the prosperity and continuity of the family” (Lee 1998, p. 253) and “neglecting one’s ancestors brings misfortune to the family” (Park and Muller 2014).

The basis for ancestral rites at home was that primogeniture descendants performed rites for ancestors, usually up to four generations. Women were not allowed to participate in the rituals, except to prepare food. The Choseon dynasty ended with Japanese colonization. The Japanese colonial government intended to change Korean culture to make it more like Japanese culture and established simplified family ritual standards. The Korean government in 1969 also implemented the new standards as a part of the modernization process (Lee et al. 2018). In addition, some further changes in ancestral rites and other ritual procedures included limiting the number of honoring generations from four to two generations for ancestral rites in the 1990s (Kim 2014a). According to a family survey (Ministry of Gender Equality and Family 2010) of Koreans who performed ancestral rites, 11% memorialized up to four generations of ancestors, 25% up to three generations of ancestors, 42% honored two generations, and 16% honored one generation of ancestors.

2. Influence of Confucianism on Buddhists, Catholics, Protestants, and Religious None

The relationships between religious groups and Confucianism can be two-fold if we consider the relationship between religion and paranormal (or nonofficial) beliefs (Rice 2003). On the one hand, people with a specific religion are less likely to adhere to Confucian beliefs or practices because specific religious doctrines or prescriptions prohibit adherence to another religion. On the other hand, religious people are more likely to observe Confucian rituals because of the commonality of spiritual orientations. These two plausible arguments need to represent the context of religious syncretism in Korea. Kim (2005) showed that 73% of Buddhists, 26% of Protestants, 41% of Catholics, and 53% of religious none agree that auspicious gravesites usher in good fortunes for descendants. After 10 years, excluding Protestants, there is little change in this belief, with 71% Buddhists, 35% Protestants, 45% Catholics, and 51% religious none (Gallup Korea 2015, p. 86).

Korea’s current religious landscape comprises of folk religion, Buddhism, Confucianism, Catholicism, and Protestantism, as a product of a long history of mutual interaction and transformation of each religion in organizational or practice aspects (Cho 1999; Grayson 1992, 2001). Buddhism was introduced into Korea in 372 AD and was the dominant religion until Confucianism surpassed it in the Choseon dynasty. Although Buddhism was oppressed during the Choseon dynasty, the relationship between Buddhism and Confucianism is symbiotic in terms of ritual practice. Buddhism holds the basic tenets of filial piety and ancestor worship and can adopt Confucian rituals, with Buddhist monks playing a significant role in such rituals (Kim 2014b).

Catholicism and Protestantism interacted with Confucianism in a context of social and political unrest. Catholicism is the first monotheistic religion introduced to Korea in 1784 when Catholic churches prohibited ancestral rites. This ban on ancestral rites caused severe conflict with the Choseon dynasty, which promoted ancestral rites. In 1791, an official persecution of Korea’s emerging Catholic community began and lasted for almost a century, taking the lives of both Korean Catholics and, from 1839 on, their French missionary pastors as well. In the Franco-Korean treaty in 1886, Catholicism was finally officially acknowledged in Korea. In 1939, Pope Pius XII overturned the anti-rite decree of 1742, allowing Korean Catholics to perform ancestral rites. About 100 years after the introduction of Catholicism, Protestantism emerged in Korea when the Choseon dynasty was weakened. Although political and social conditions were favorable for the implantation of Protestantism, Korean Protestants had the same dilemma against ancestral rites, which they considered an act of idolatry. Shortly after the arrival of American Protestant missionaries, Korean Protestants created Chudoyebe, a Christian memorial service, as a substitute for Confucian ancestral rites. In this way, Korean Protestants avoided the persecution faced by earlier Korean Catholics (Grayson 2009a). According to Grayson (2009b), the successful implantation of Christianity in Korea in a short period was based on resolution of conflict with the core values of Confucianism. However, unlike the Korean Catholic church, the Korean Protestant church has not formally approved ancestral rites (Kim 2015), and it remains a problem for church ministers and adherents (Park 2010).

It is not easy to conclude whether the half of the Korean population that are not affiliated with Buddhism or Christianity can be considered Confucians. On the one hand, religious none might contain Confucians because Koreans have acquired and shared Confucian values and ritual practices under the influence of 500 years of Confucian rules. On the other hand, with rapid modernization and weakened Confucian traditions, people who are not Buddhists or Christians cannot be Confucian because they have doubts about ancestors and about God.

3. Data

Some data were collected from the 2016 and 2018 Korean General Social Survey (KGSS), a national representative survey of Korean adults aged 18 and over who do not live in institutions (Kim et al. 2019). Since 2003, the Survey Research Center at Sungkyunkwan University has performed the KGSS by face-to-face interviews annually (2003–2014) and biennially (2016, 2018). The sample size and response rate for the 2016 KGSS and 2018 KGSS are 1051 (46%) and 1031 (43%), respectively. The KGSS asked questions about beliefs and practices related to Confucian identification, beliefs about ancestors, and visiting graves in both years.

The Gallup Korea Omnibus survey was designed as a face-to-face survey of a nationally representative sample of adults conducted in March 2021. The response rate was 32%. We collected responses to the question, “Do you conduct ancestral rites on Lunar New Year or Korean Thanksgiving? If you do, do you perform Confucian rites or Christian prayers?” Responses to this same question in the Religion of Koreans survey in 1984, 1989, 1997, 2004, and 2014 were collected.

Data were also collected from the Values in Crisis (VIC), a cross-national survey that aims to study value changes across nations. This survey is a non-probability online panel of Korean adults. In the first wave (N = 4000) collected in May 2020, respondents were asked if they felt close to the Confucian way of life and a few items for the reasons for their feeling of closeness. In the third wave of this survey (n = 2460) collected in March 2021, we again asked if respondents felt close to the Confucian way of life, and then we asked what comes to mind when they think of Confucianism in an open-ended format.

4. Analysis

Following suggestions by Hackett (2014) about measuring religious identity, we divide questions related to Confucianism into three groups, religious identity, beliefs, and practice. We first show the association between religious groups and religious identity, followed by those of beliefs and then practice. Multivariate binary logistic regression analysis was performed for the relationship between religious groups and Confucian identification (Yes = 1, No = 0), belief in the supernatural power of deceased ancestors (1 = Yes, 0 = No), belief in the impact of the soul of a dead person on descendants (1: strongly agree and agree, 0 = else), conducting ancestral rites at a gravesite (1: at least once per year 0: None), conducting ancestral rites at home (1: at least once a year 0: Never), and having a shrine or tablet in their home (1: Yes 0: No). We controlled age, sex, marital status, education, income, and urbanicity (see Appendix B for variable descriptions). For interpretation of bivariate analysis, we did not include the ‘others’ religious group because several small religious groups represent this category. In addition to bivariate analysis, we also briefly present the results of multivariate analysis related to religious group differences.

5. Results

Instead of presenting the usual religious identification question from previous surveys or censuses, the 2016 KGSS asked respondents, “(Regardless of your religious affiliation) do you consider yourself a Confucian?”

Table 1. Confucian identification by religion, age, sex, marital status, education, family income, and urbanicity (%), 2016 Korean General Social Survey (KGSS).

	Confucian Identification (%)
Yes	51
Religion		***
Buddhists	67
Protestants	38
Catholics	60
None	47
Other	58
Age group		***
18–24	49
25–34	44
35–44	43
45–54	52
55–64	60
65+	61
Sex		NS
Female	48
Male	54
Marital status		NS
Married	52
Never married/Divorced/Widowed	50
Education		NS
Less than high school	56
High school graduate	55
More than high school graduate	48
Family income (monthly)		NS
1st quartile	52
2nd quartile	49
3rd quartile	51
4th quartile	55
DK	44
Urbanicity		NS
Urban	50
Rural	57

Notes: *** p < 0.001, NS = not significant.

showed that fifty-one percent of people considered themselves Confucians; by religious group, 67% of Buddhists, 60% of Catholics, 47% of the unaffiliated, and 38% of Protestants identified themselves as Confucian. Among socio-demographic variables of age, sex, education, household income, and urbanicity, only age is statistically significantly associated with Confucian identification. About 60% of Koreans aged 55 years and older considered themselves Confucians, compared to less than 50% of people aged 18 to 44 years. Using binary logistic regression, we observe that the odds of considering oneself Confucian varies by religious group (Appendix A:

Table A1. Results of binary logistic regression on Confucian identification, beliefs, and practices.

	Confucian Identification 1		Supernatural Powers of Ancestors 2		Influence of Ancestors on Descendants 2
	OR	95% CI	OR	95% CI	OR	95% CI
Religion
None (referent)
Buddhists	2.161 ***	1.53, 3.05	2.295 ***	1.59, 3.32	4.65 ***	2.93, 7.37
Protestants	0.66 **	0.47, 0.93	0.79	0.52, 1.19	1.128	0.64, 2
Catholics	1.723 **	1.08, 2.76	1.393	0.89, 2.17	2.678 ***	1.54, 4.65
Other	1.002	0.27, 3.66	1.254	0.25, 6.28	8.588**	1.9, 38.75
Age group
18 to 24 (referent)
25 to 34	0.792	0.48, 1.32	1.422	0.78, 2.6	1.063	0.5, 2.27
35 to 44	0.775	0.44, 1.38	1.644	0.89, 3.04	0.908	0.41, 2.01
45 to 54	1.074	0.61, 1.9	1.913 **	1.05, 3.49	1.328	0.63, 2.82
55 to 64	1.561	0.82, 2.98	2.219 **	1.15, 4.29	1.063	0.46, 2.47
65 and over	1.831	0.93, 3.6	2.12 **	1.04, 4.33	1.962	0.82, 4.67
Female	0.744 **	0.57, 0.97	1.005	0.75, 1.34	1.305	0.9, 1.9
Married	0.816	0.58, 1.15	0.924	0.65, 1.31	0.873	0.56, 1.36
Education
Less than high school (referent)
High school graduates	1.533	0.94, 2.51	0.71	0.43, 1.18	0.811	0.43, 1.52
More than high school graduate	1.363	0.81, 2.3	0.785	0.45, 1.36	0.904	0.45, 1.81
Family income (monthly)
1st quartile (referent)
2nd quartile	0.839	0.5, 1.41	1.159	0.63, 2.13	0.992	0.45, 2.17
3rd quartile	0.871	0.52, 1.46	1.105	0.61, 2.01	1.372	0.65, 2.9
4th quartile	1.106	0.67, 1.82	1.027	0.58, 1.83	1.446	0.7, 3
DK	0.741	0.4, 1.36	0.747	0.4, 1.4	1.832	0.86, 3.88
Urban	0.889	0.62, 1.28	1.195	0.82, 1.75	0.761	0.48, 1.21

Notes: Source: ¹ 2016 KGSS, ² 2018 KGSS; OR = odds ratio; CI = confidence interval. ** p < 0.01; *** p < 0.001.

). Buddhists (OR = 2.1) and Catholics (OR = 1.7) have higher odds of Confucian identification than religious none, while Protestants (OR = 0.66) have lower odds of Confucian identification than religious none.

As

Table 2. Feeling close to the Confucian way of life and reason (%), 2020 VIC.

	Buddhists	Protestants	Catholics	None		Total
Feeling close to Confucian Way of Life (Yes)	59.5	38.3	46.3	33.6		39.3
N	487	899	436	2135		3957
A reason you feel close to the Confucian way of life
You feel an obligation to your family	99.3	98.5	98.5	96.4		97.7
You see Confucian values as part of being Korean	91.4	87.8	85.1	85.5		87.1
You value Confucian rituals and practice in your life	73.4	57.0	57.7	57.2	***	60.3
You feel a duty toward your ancestors	85.5	55.5	74.8	60.4	***	65.9
Confucian values are a moral guide in your life	73.1	71.8	64.4	69.6		70.1
N	290	344	202	718		1554

Note: *** p < 0.001.

shows, in the VIC, about 40% of respondents noted the impact of the Confucian way of life. These impacts were surveyed. Almost all respondents (98%) chose “you feel an obligation to your family” as an impact, followed by at least 60% reporting the effects of “Confucian values as part of being Korean,” “Confucian values are a moral guide in your life,” “a duty toward your ancestors,” and “value Confucian rituals and practice in your life.” Two of these reasons are significantly associated with religious groups. Buddhists (73%) are more likely than other religious groups (about 57%) to value Confucian rituals and practices. Buddhists also more frequently report “a duty toward your ancestors” as an important item than do other religious groups, although Catholics are more likely than Protestants or religious none to report this as a reason.

Unlike the closed-ended questions in 2020 VIC, the 2021 VIC asked respondents to respond in their own words as to what comes to their mind when they think about Confucianism. As shown in

Table 3. The first thing comes to your mind about Confucianism (%), 2021 VIC.

	%
Living ethically/Following rules	19.4
Respect older people	16.8
Ancestral rites/Ancestor worship	14.5
Filial piety/Family	10.9
Politeness	7.9
Consideration/Respect for others	4.5
Custom/Tradition	4.4
Conservative/Ascetic	4.2
Gender inequality/Patriarchy	4.2
Others (Loyalty, Confucius, Choseon dynasty, etc.)	8.3
No answer/Don’t know	5.0
Total	1056

, about 19% mentioned living an ethical life or following rules. Respect for older people (17%), ancestral rites/ancestor worship (15%), filial piety/family (11%), and politeness (8%) were also listed. Consideration, custom, conservative, and gender inequality each was reported by less than 5% of the respondents.

Ancestor worship is a key point of Confucianism.

Table 4. Beliefs about ancestors (%), 2018 KGSS.

	Buddhist	Protestants	Catholics	None	Other	Total
Do you believe in…the supernatural power of deceased ancestors?
Yes, definitely	12.4	3.5	5.6	3.7	28.6	5.7
Yes, probably	33.2	18.6	26.4	20.1	-	22.9
No, probably not	36.3	31.2	48.8	40.2	14.3	38.6
No, definitely not	18.1	46.7	19.2	36.0	57.1	32.8
Total	193	199	125	492	7	1016
Do you think the soul of a dead person can affect descendants?
Strongly agree	9.0	2.1	5.7	1.3	14.3	3.5
Agree	24.3	8.2	16.4	7.8	28.6	12.2
Neither agree nor disagree	27.5	16.0	25.4	28.4	42.9	25.5
Disagree	28.0	45.9	41.0	44.1	-	40.7
Strongly disagree	11.1	27.8	11.5	18.5	14.3	18.0
Total	189	194	122	476	7	988

Notes: - indicates no cases.

illustrates how people conceive of the power or influence of ancestors. Slightly less than one in three Koreans (5.7% “Yes, Definitely” and 22.9% “Yes, Probably”) believe in the supernatural power of deceased ancestors, and less than one in five people (3.5% “Strongly Agree” and 12.2% “Agree”) think that the soul of a dead person can affect descendants. Of the religious groups, Buddhists (45.6%) are almost twice as likely than Protestants (22.1%) or religious none (23.8%) to believe in the supernatural power of ancestors. Buddhists (33.3%) also are three times more likely than Protestants (10.3%) and religious none (9.1%) to value the influence of ancestors on descendants. Catholics are less likely than Buddhists but more likely than Protestants or religious none to believe both of these ideas. In multivariate analysis (Appendix A: Table A1), only Buddhists (OR = 2.3) are more likely to believe in the supernatural power of ancestors than religious none. In terms of the belief of the influence of a dead person on descendants, both Buddhists (OR = 4.7) and Catholics (OR = 2.7) are more likely to believe it than religious none.

Sun (2013) acknowledged that ancestral rituals might not necessarily indicate the practice of Confucianism.

Table 5. Religion associated with rites for deceased family members by religious group (%), 2018 KGSS.

	Buddhists	Protestants	Catholics	None	Others	Total
Confucian	32.5	17.0	48.9	61.9	45.5	44.2
Confucian + Buddhist	14.2	-	1.1	7.5	-	6.8
Confucian + Christian	-	8.0	6.4	1.3	9.1	3.0
Confucian + Buddhist + Christian	0.4	0.4	-	-	9.1	0.3
Confucian + Other	-	-	2.1	-	-	0.2
Christian	2.5	69.2	21.3	5.2	9.1	19.7
Christian + Other	-	1.3	-	-	-	0.3
Buddhist	45.8	2.7	6.4	17.1	9.1	19.5
Buddhist + Christian	0.8	0.4	4.3	0.6	-	1.0
Buddhist + Shinto	0.8	-	-	-	-	0.2
DK	0.8	-	3.2	4.4	9.1	2.6
Others	0.8	0.4	5.3	2.1	9.1	1.8
Daoist	0.8	-	-	-	-	0.2
Shinto	0.4	-	1.1	-	-	0.2
Muslim	-	0.4	-	-	-	0.1
Total	240	224	94	480	10	1048

Note: Categories with + sign include respondents who offered multiple responses. Cells with – sign indicate zero cases belonging to the cell.

demonstrates the subjective perception of people affiliated with different religions about ancestral rites, “Do you consider the rites for your deceased family members to be… Confucian, Buddhist, Shinto, Daoist, Christian, Muslim, Other, or DK.” Even though respondents could choose more than one religious group, the majority chose only one. Although it is not the majority, 44% of people think that the rites for deceased family members are Confucian, followed by about 20% each associating them with Christianity and Buddhism, respectively. Among those who chose more than one category, about 7% chose Confucianism and Buddhism, and 3% chose Confucian and Christianity. These responses show that ancestral rites are not regarded as exclusive to Confucianism.

When we examine the distribution of religions associated with ancestral rites, there are significant differences. A total of 33% of Buddhists, 17% of Protestants, 49% of Catholics, and 62% of religious none consider ancestral rites to be Confucian. Interestingly, Catholics (49%) are much more likely to assign them as Confucian than Protestants (17%). Not surprisingly, Koreans with no religious affiliation are much more likely than any other group to attribute rites for deceased family members to Confucianism. However, 17% of those with no religious identification believe the rituals for deceased family members are Buddhist.

Table 6. Ways to practice ancestral rites during holidays (%), Gallup Korea.

	1992	2000	2002	2006	2009	2013	2014	2021
Confucian way	80	78	79	78	75	70	71	58
Christian way	12	14	11	15	16	15	14	10
No ancestral rites	8	7	9	7	9	14	12	32

Notes: Percentages do not add to 100% due to rounding or exclusion of the other category.

shows the 30-year trend toward conducting ancestral rites based on the question, “Do you conduct ancestral rites on Lunar New Year or Korean Thanksgiving? If you do, do you perform Confucian rites or Christian prayers?” Little change is shown in the small percentage of Koreans not performing ancestral rites at holidays between 1992 and 2009, and a slight decrease is noted for such practice from 2013 to 2014. In contrast, in 2021, 32% of Koreans did not perform ancestral rites. The near 20% increase from 2014 to 2021 might be related to COVID-19. Koreans were very concerned about family gatherings, where most ancestral rites are performed. Binary logistic regression showed that the odds of conducting ancestral rites were 3.36 times higher for Buddhists than for religious none, while the odds for Protestants were 0.66 times lower than for religious none (Appendix A:

Table A3. Results of binary logistic regression on Confucian identification, beliefs, and practices (cont.).

	Conducting Ancestral Rites
	OR	95% CI
Religion
None (referent)
Buddhists	3.359 ***	2.17, 5.2
Protestants	0.658 **	0.49, 0.88
Catholics	0.639	0.4, 1.01
Other	0.72	0.13, 4.13
Age group
19 to 24 (referent)
25 to 34	1.152	0.73, 1.82
35 to 44	1.277	0.76, 2.14
45 to 54	2.004 **	1.16, 3.47
55 to 64	2.277 **	1.26, 4.12
65 and over	2.189 **	1.13, 4.25
Female	0.717 **	0.57, 0.9
Married	0.944	0.67, 1.33
Education
Less than high school (referent)
High school graduates	0.632	0.39, 1.03
More than high school graduate	0.693	0.4, 1.21
Family income (monthly)
1st quartile (referent)
2nd quartile	1.401	1, 1.96
3rd quartile	1.125	0.79, 1.61
4th quartile	1.29	0.91, 1.83
DK	0.979	0.61, 1.57
Urban	1.401	1, 1.96

References Notes: Source: 2021 Gallup Korea; OR = odds ratio; CI = confidence interval. ** p < 0.01; *** p < 0.001.

).

Conducting family (ancestral) rites at a gravesite was asked about in the 2016 KGSS, “how many times do you conduct rites for the spirits of your deceased family members (ancestors) at their gravesite (cremation memorial site) (example: Performing ancestral rituals, offering prayers, burning incense, offering food, and/or wine)?

Table 7. Conducting ancestral rites at a gravesite (%), 2016 KGSS.

	Buddhists	Protestants	Catholics	None	Others	Total
Once per year	20.1	23.7	18.9	26.5	11.1	23.6
Twice per year	27.2	31.7	29.5	35.7	66.7	32.6
Three or four times per year	29.7	15.6	20.0	17.2	11.1	19.9
More than four times per year	15.1	8.5	13.7	8.2	0.0	10.3
Never	7.9	20.5	17.9	12.4	11.1	13.6
Total	239	224	95	476	9	1043

shows the distribution of answers among religious groups. About 14% never conducted family (ancestral) rites. About 33% of respondents visited a gravesite twice per year, which was the most common category. Buddhists differ substantially from Protestants among the religious groups, and Buddhists (92.1% at least once per year) are at least 12% point more likely to perform ancestral rites than Protestants (79.5% at least once per year). As shown in Appendix A (

Table A2. Results of binary logistic regression on Confucian identification, beliefs, and practices (cont.).

	Conducting Ancestral Rites at Gravesite		Conducting Ancestral Rites at Home		Having a Shrine or Tablet at Home
	OR	95% CI	OR	95% CI	OR	95% CI
Religion
None (referent)
Buddhists	1.756	0.99, 3.12	1.564 **	1.07, 2.29	3.359 ***	2.17, 5.2
Protestants	0.516 **	0.33, 0.81	0.773	0.54, 1.1	0.658 **	0.49, 0.88
Catholics	0.574	0.31, 1.07	1.29	0.77, 2.16	0.639	0.4, 1.01
Other	0.85	0.13, 5.58	2.588	0.43, 15.76	0.72	0.13, 4.13
Age group
18 to 24 (referent)
25 to 34	0.547	0.26, 1.14	0.482 **	0.28, 0.84	0.34	0.1, 1.16
35 to 44	0.615	0.27, 1.41	0.505 **	0.27, 0.95	0.097 **	0.02, 0.56
45 to 54	1.175	0.5, 2.78	0.685	0.36, 1.29	0.453	0.13, 1.61
55 to 64	1.378	0.51, 3.71	0.645	0.32, 1.3	0.365	0.09, 1.49
65 and over	1.025	0.39, 2.72	0.957	0.45, 2.04	0.773	0.19, 3.08
Female	0.519 **	0.35, 0.77	0.829	0.63, 1.1	0.687	0.37, 1.26
Married	1.224	0.78, 1.92	1.306	0.91, 1.87	1.128	0.5, 2.55
Education
Less than high school (referent)
High school graduates	1.134	0.57, 2.25	0.798	0.47, 1.35	0.985	0.38, 2.52
More than high school graduate	1.426	0.67, 3.03	1.217	0.69, 2.14	0.627	0.21, 1.88
Family income (monthly)
1st quartile (referent)
2nd quartile	1.768	0.93, 3.38	1.311	0.75, 2.28	1.567	0.53, 4.62
3rd quartile	2.064 **	1.06, 4	1.301	0.75, 2.26	1.361	0.44, 4.22
4th quartile	3.387 ***	1.75, 6.56	1.291	0.76, 2.19	2.533	0.88, 7.33
DK	2.756 **	1.25, 6.06	1.79	0.93, 3.45	1.089	0.27, 4.34
Urban	1.178	0.71, 1.94	0.755	0.51, 1.12	0.343 **	0.18, 0.67

Notes: Source: 2016 KGSS; OR = odds ratio; CI = confidence interval.; ** p < 0.01; *** p < 0.001.

), multivariate analysis of predicting ancestral rites at a gravesite (1 = at least once a year vs. 0 = Never) shows that, compared with religious none, only Protestants (OR = 0.52) are less likely to do so.

Table 8. Conducting ancestral rites at home and having a shrine or tablets (%), 2016 KGSS.

	Buddhists	Protestants	Catholics	None	Others	Total
Almost every day	0.0	0.9	3.2	0.4	0.0	0.7
Once or twice a week	1.7	2.7	4.2	1.3	20.0	2.1
Once or twice a month	3.8	2.2	1.1	1.9	0.0	2.3
A few times per year	57.3	38.4	49.5	49.4	60.0	48.9
Once a year	15.5	20.1	17.9	15.7	10.0	16.7
Never	21.8	35.7	24.2	31.4	10.0	29.3
Having a shrine or tablets at the home	6.7	2.7	4.3	5.0	18.2	5.0

shows responses to the question “How many times do you conduct rites for the spirits of your deceased family members (ancestors) at home? (example: Performing ancestral rituals, offering prayers, burning incense, offering food and/or wine).” About 30% did not conduct ancestral rites at home, and among those who perform ancestral rites at least once a year, the most common response was a few times per year. Buddhists and Catholics are more likely to conduct ancestral rites at home for at least once a year than Protestants and religious none. Multivariate analysis (Appendix A: Table A2) revealed that, compared with religious none, Buddhists (OR = 1.6) are more likely perform ancestral rites at home. However, Protestants or Catholics are not statistically different from religious none. In contrast to ancestral rites at home, having a shrine or tablets at home is rare (5%), and there are no statistical differences among the religious groups.

6. Conclusions

The rich data on Confucian identity, ancestral beliefs, and ancestral rites from recent surveys revealed the number of Confucians and their relationships with religious groups. Contrasted with 0.2% of the adult population identifying as Confucian in the 2015 Korean Census, 51% considered themselves Confucian when the question was phrased, “(Regardless of your religious affiliation) do you consider yourself a Confucian?” If we consider those who think that rites for deceased family members are Confucian, the proportion would be 44%. Considering those who conduct ancestral rites at a gravesite as Confucian, it would be 86%, but it is only 70% when we count those who perform ancestral rites at home as Confucian. In sum, although we reported a wide range of Confucian identification, from 44% to 86% depending upon the measures, the results of our study are a good starting point to the empirical study of Confucians in Korea.

In addition to counting the number of Confucians, we found substantial differences across religious groups about Confucian identification, beliefs about ancestors, and conducting ancestral rites. In general, of all the groups, Buddhists were most likely and Protestants were least likely to identify with Confucianism, believe in the power of ancestors, and perform ancestral rites. This might be due to the long coexistence of Buddhism and Confucianism that promotes mutual interactions and inclusivity. Regarding the differences between Catholics and Protestants, we speculate that acceptance by a religious authority influences the views of Catholics and Protestants on ancestral rituals. Perhaps most telling is the result of religious none falling between Buddhists and Protestants in terms of identification, beliefs, and rituals of Confucianism. The differences between religious groups on ancestral rituals appear to reflect religious syncretism, and religious none should not be regarded as Confucian.

Although we cannot pinpoint the number of Confucians in Korea, several potential indicators of Confucian identity support the continuation of Confucian tradition. It is an exaggeration to announce a revival of Confucianism, but it is safe to conclude that Confucianism is not a dying tradition in Korean society. Measures used for identifying Confucians instead of the conventional religious identification question will better allow us to determine the stability or change of the Confucian tradition in East Asian societies. Future research on Confucian identity requires creative new measures and depends upon continuous data collection.