Chinese culture became more individualistic: Evidence from family structure, 1953-2017

Previous research has indicated that some aspects of Chinese culture became more individualistic. However, prior studies have suggested a decrease in individualism in other aspects of China. Thus, it was unclear whether China became more individualistic. Therefore, the current research investigated whether Chinese culture became more individualistic by examining historical changes in family structure. Specifically, I analyzed temporal shifts in the divorce rate and household size, which have been confirmed as valid representative indicators of individualism. Results showed that the divorce rate increased between 1978 and 2017 and household size decreased between 1953 and 2017, indicating a rise in individualism. Moreover, analyses suggested that the one-child policy was unlikely the sole and major factor in the decrease in household size. Additionally, the aggregated score of divorce rate and household size demonstrated a clear increase in individualism. Therefore, the present research provided further evidence of the rise in individualism in China.


Increase in individualism in China
Previous research has indicated that Chinese culture has become more individualistic over time. Taras, Steel, and Kirkman (2012) conducted a meta-analysis of studies that used Hofstede (1980)'s cultural framework (four dimensions; individualism, power distance, masculinity, uncertainty avoidance). In this research, they found that individuals in China came to hold more individualistic values from the 1980s to the 2000s, which shows that Chinese culture became more individualistic at the individual level. Not only at the individual level, but also at the cultural (group) level, Chinese culture has changed toward greater individualism. A previous study examined historical changes in individualism-collectivism in China between 1950and 2008(Hamamura and Xu, 2015. It analyzed changes in the frequency of first-person singular/plural pronouns used in Chinese published books over the period by utilizing a database of millions of books published in various languages (Google Books Ngram). It found that the rate of first-person singular pronouns increased and the rate of first-person plural pronouns decreased. These results suggest an increase in individualism in China (also see, Yu et al., 2016). Moreover, Zeng and Greenfield (2015) have indicated that the prevalence of words that were considered to reflect individualistic values/behaviors (e.g., "autonomy", "choose") increased in Chinese books published between 1970 and 2008, reflecting a shift toward greater individualism in Chinese culture.
Not only values but also behaviors may have become more individualistic. It has been claimed that unique names increased in China between 1950 and 2009, suggesting a rise in the need for uniqueness and individualism (Cai, Zou, Feng, Liu, & Jing, 2018;Bao, Cai, Jing, & Wang, 2021; but also see, Ogihara, 2020b).

Decrease in individualism in China
In contrast, regarding values, some studies have indicated that Chinese culture has become less individualistic. It has been reported that individuals in China came to hold less individualistic values between 1990 and 2007 (Santos, Varnum, & Grossmann, 2017). This study demonstrates that at the individual level, China has shifted toward a less individualistic culture. 1 Further, Zeng and Greenfield (2015) found that, contrary to their hypothesis, the prevalence of the two words ("obliged" and "give"; out of eight words examined) that were considered to reflect collectivistic values/behaviors increased in usage in Chinese books published between 1970 and 2008. This result may show that China became less individualistic at the cultural level. 2

Current research
Thus, it is unclear whether Chinese culture has become more individualistic. Particularly, there have been only two studies examining changes in the behavioral aspect in China (naming; Cai et al., 2018). Thus, it is important to examine cultural changes in China by using other behavioral measurements. Therefore, the current research investigated cultural changes in China using two other behavioral indicators of individualism that have already been validated but have not been used to examine cultural changes in China: divorce rates and household size. 1 Although this study described global trends of temporal changes in individualism, it has some limitations. For example, the authors aggregated three items from the World Values Survey to examine temporal shifts in individualistic values, but the inter-item correlations were not high (.11 < τs < .24; ordinal-level Kendall's τ), possibly implying that the validity of these items may be low. Moreover, as the authors already noted in their article, the data points for China were only four (1990, 1995, 2001, and 2007), and they covered a relatively short period of time (17 years). This limited number of data points might have led to a failure to detect actual historical changes in individualistic values in China. These limitations may be related to the inconsistency of the findings. Indeed, their study showed that most (39 out of 53) of the countries they examined indicated a substantial increase in individualistic values, whereas only five countries (China, Armenia, Croatia, Ukraine, and Uruguay) exhibited a nonnegligible decrease in individualistic values. 2 However, these words increased in frequency only slightly. As the authors emphasized in their article, the contrasting individualistic words ("choose" and "get") that were analyzed increased more remarkably (for the details of their interpretation, see Zeng and Greenfield, 2015).

REVISED Amendments from Version 2
Following the reviewer's suggestion, I have added some explanations of the crude divorce rate and possible reasons why the household size decreased. Additionally, the notes of Table 1 and Table 2 have been included on the webpage.
Any further responses from the reviewers can be found at the end of the article Divorce rates and household size are behavioral measurements reflecting individualistic tendencies. In individualistic cultures, family structure tends to be freer and looser compared to that in collectivistic cultures (e.g., Georgas et al., 2001;Triandis, 1995). This leads people to live separately and independently of other family members, contributing to higher divorce rates and smaller households.

Indicators
As the measure of divorce rates, the divorce-to-marriage ratio was used. Data covered the period between 1978 and 2017 and came from the National Bureau of Statistics of China (2018). Data on household sizes covered the period between 1953 and 2017 and were drawn from the National Bureau of Statistics of China (2018).

Aggregated score
Although divorce rates and household size have been confirmed as valid indicators of individualism, each indicator may be simultaneously influenced by factors other than individualism. For instance, celebrity divorces may impact divorce rates. By aggregating the two indicators, the random errors of each indicator cancel each other out, successfully reducing the influence of these random errors. Thus, an aggregated score was calculated by averaging the divorce rate (z-transformed) and household size (z-transformed and reversed). This strategy of analysis has been frequently used in prior research (e.g., Diener et al., 1995;Grossmann & Varnum, 2015;Hamamura & Xu, 2015;Ogihara, 2018bOgihara, , 2020aSantos et al., 2017;Yu et al., 2016;Vandello & Cohen, 1999).

Results
Simple Pearson's and Kendall's correlation coefficients among each indicator are shown in Table 1. Figure 1 shows historical shifts in the divorce rate between 1978 and 2017 in China. The divorce rate significantly rose over the last 40 years. In 1978, 4.8 out of 100 couples divorced, but in 2017, 41.1 out of 100 couples experienced a divorce. The correlation between the year and the divorce rate was strongly positive (Table 1), suggesting an increase in individualism. Household size Figure 2 indicates temporal shifts in household size between 1953 and 2017 in China. Household size declined over the past 60 years. In 1953, the average household consisted of 4.3 people, but in 2017 the average was 3.1 people. The correlation between year and household size was highly negative (Table 1), which indicates a rise in individualism.

Divorce rate
The one-child policy and household size. One might expect that the decrease in the average household size was caused solely by the one-child policy and may thus not reflect an increase in individualism. China's one-child policy, which was introduced in 1980 and ended in 2015, penalized parents for having more than one child. This policy may have decreased the birth rate, in turn decreasing the average household size.
However, it is difficult to assert that the decrease in the average household size was caused solely by the one-child policy for three reasons. First, it is not necessarily correct that the one-child policy continued to decrease the birth rate (Whyte, Feng, & Cai, 2015). Indeed, after the one-child policy was introduced in 1980, the fertility rate did not continue to decrease ( Figure 3A). Contrary to a common myth, even after the implementation of the one-child policy, the fertility rate increased between 1983 and 1986, and between 2000 and 2017. For more than half period of time (22 years; 1983-1986, 2000-2017) when the one-child policy was introduced (36 years; 1980-2015), the fertility rate did not decrease.
Third, even after the one-child policy ended in 2015, the average household size continued to decrease. If the one-child policy had a strong influence on household size, after the policy ended, household size should have increased (but it decreased).
Thus, although the one-child policy may have contributed to the decrease in household size to some extent, its effect was not large. Therefore, the one-child policy was unlikely to be the sole and major factor in this decrease in the average household size.

Aggregated score
The aggregated score of divorce rate and household size showed a clearer and more consistent pattern of increased individualism than each indicator ( Figure 4). The correlation between the year and the aggregated score was strongly positive (Table 1), suggesting an increase in individualism.
Crude divorce rate When the crude divorce rate (rate per 1,000 persons in general, not limited to married people) was used instead of divorce-to-marriage ratio, the results were unchanged (Table 2; Figure 5; Figure 6). The Pearson's and Kendall's correlation coefficients between crude divorce rate and divorce-to-marriage ratio were .95 and .92, respectively.  Note. The aggregated score was calculated by averaging the divorce rate (z-transformed) and household size (z-transformed and reversed).

Discussion
The current research showed that Chinese culture became more individualistic over the past 60 years by using behavioral indicators that had not been sufficiently examined. Past research showed the rise in individualism in China Cai et al., 2018;Hamamura & Xu, 2015;Taras et al., 2012;Yu et al., 2016;Zeng & Greenfield, 2015). However, some studies indicated the fall of individualism in China (Santos et al., 2017;Zeng & Greenfield, 2015). Thus, it remained unclear whether Chinese culture became more individualistic. Particularly, it was unclear whether China shifted toward greater individualism in the behavioral aspect. The present research investigated temporal changes in the two behavioral indicators (divorce rate and household size), which have been used to examine cultural changes in other countries (e.g., Hamamura, 2012;Grossmann & Varnum, 2015;Ogihara, 2018bOgihara, , 2020a. Results showed that the divorce rate increased dramatically between 1978 and 2017, and that household size steadily shrank between 1953 and 2017, suggesting an increase in individualism in China. Further, analyses indicated that the onechild policy was not the major factor of the decrease in household size. Many factors could explain this decrease in household size (e.g., an increase in people living alone, a decrease in households with multiple generations). Moreover, the aggregated score of divorce rate and household size demonstrated a clearer pattern of the increase in individualism than each indicator. Therefore, this research contributes to the accumulation of the literature demonstrating the rise in individualism in China by using different behavioral indicators of individualism.

Similar phenomena in Japan
It is more likely that these mixed changes are found in historically collectivistic cultures that are becoming more individualistic. Indeed, such changes were also found in Japan, another country in East Asia (for a review, see Ogihara, 2017Ogihara, , 2018a).
On the one hand, people in Japan came to live more independently from other family members (Hamamura, 2012;Ogihara, 2018b), increasingly gave more unique names to their babies (Ogihara, 2015(Ogihara, , 2021a(Ogihara, , 2021bOgihara et al., 2015), and came to hold more individualistic values (Hamamura, 2012;Taras et al., 2012), indicating an increase in individualism. On the other hand, a decrease in and non-change of individualistic values have been reported (Hamamura, 2012).
It is suggested that these mixed changes may be related to difficulties in adapting to a new environment (e.g., Ogihara, 2016;Ogihara & Uchida, 2014;Ogihara et al., 2014Ogihara et al., , 2016. Still, it is difficult to say that there is a sufficient amount of research on how cultures change and how people adapt to such cultural changes in East Asia. Therefore, it is important to investigate historical changes in these cultures in more detail.

Limitations and future directions
The current research used family structure (divorce rate and household size) as representative indicators of individualism. Nevertheless, it is still unclear whether other aspects of individualism show the same trend. To examine cultural changes in China, it is necessary to investigate various aspects of cultural changes by analyzing different indicators from a broad set of perspectives.
Although the present research demonstrated that Chinese culture became more individualistic, it is not clear why this change arose. There are many possible factors to cause this change (e.g., economic wealth, social mobility, subsistence styles, ecological/societal threats, and urbanization). Future studies should answer this question. Note. The aggregated score was calculated by averaging the crude divorce rate (z-transformed) and household size (z-transformed and reversed).

Open Peer Review
This research demonstrates the changes in these two indices that are known to be related with individualism of the society.
There are a couple of minor comments on this brief report: Table 1 is missing which parts (above and below) of the diagonal are about Pearson's correlation and Kendall's correlation coefficients.

○
As a more comprehensive test, the crude divorce rate is used in addition to the divorce-tomarriage ratio. Please clarify how crude divorce rate is calculated. Are they the ratio of individuals who got divorced amongst 1,000 married people or people in general?
○ As this study is looking at social structural change in China, which I agree is intricately related with the cultural values of the members of the society, it would be important not to overgeneralize the finding to changes in individualism. It is possible that there is a mismatch between individuals' cultural values and societal structure (household size, in particular), therefore, it should be clearly noted that the cultural values systems of individuals may not have caught up with the changes in the social system as indicated by the increased divorced rates & decreased household size.

○
The author argues that the reason why the one-child policy might not have driven the decrease in household size is due to the disconnect between the policy and birthrate. Then, additional explanation for why the household size decreased may render insight into the changes in culture in China. Is the creased household size due to decreased households living with grandparents or an increase in one-person household? Some interpretation of this may be helpful.

Are sufficient details of methods and analysis provided to allow replication by others? Partly
If applicable, is the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required.
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed. This research demonstrates the changes in these two indices that are known to be related with individualism of the society.
There are a couple of minor comments on this brief report: Table 1 is missing which parts (above and below) of the diagonal are about Pearson's correlation and Kendall's correlation coefficients.
Thank you for your helpful comment. The PDF and the submitted manuscript of this article have a note of Table 1 ("Note. Scores in the upper half indicate Pearson's correlation coefficients, while those in the lower half represent Kendall's correlation coefficients. Both results were consistent with each other. The aggregated score was calculated by averaging the divorce rate (z-transformed) and household size (z-transformed and reversed).") However, as you pointed out, this note is missing on the webpage of this article. I asked the F1000Research office to modify this editorial error.
As a more comprehensive test, the crude divorce rate is used in addition to the divorce-tomarriage ratio. Please clarify how crude divorce rate is calculated. Are they the ratio of individuals who got divorced amongst 1,000 married people or people in general?
I appreciate your suggestion. Following your suggestion, I have added an explanation of the crude divorce rate as below. It is the rate per 1,000 persons in general, not limited to married people.
"When the crude divorce rate (rate per 1,000 persons in general, not limited to married people) was used instead of divorce-to-marriage ratio, the results were unchanged (Table 2; Figure 5; Figure 6)." Thank you for your comment. In fact, I discuss these complex changes in individualism, which includes not only values but also other aspects such as behaviors and cultural products, in China (and Japan) in this paper. The indices I used in this paper (i.e., divorce rate and household size) have been confirmed to be valid indicators of individualism and show the changes in some aspects of individualism. Thus, I explain this point as a limitation in the Limitations and future directions section as below. If you think this is insufficient, I would appreciate it if you could specify which part is insufficient.
"The current research used family structure (divorce rate and household size) as representative indicators of individualism. Nevertheless, it is still unclear whether other aspects of individualism show the same trend. To examine cultural changes in China, it is necessary to investigate various aspects of cultural changes by analyzing different indicators from a broad set of perspectives." The author argues that the reason why the one-child policy might not have driven the decrease in household size is due to the disconnect between the policy and birthrate. Then, additional explanation for why the household size decreased may render insight into the changes in culture in China. Is the creased household size due to decreased households living with grandparents or an increase in one-person household? Some interpretation of this may be helpful.
I appreciate your constructive suggestion. Following your suggestion, I have added a possible explanation of why the household size decreased as below.
"Many factors could explain this decrease in household size (e.g., an increase in people living alone, a decrease in households with multiple generations)." Thank you for your further consideration of this manuscript.
I look forward to hearing from you at your earliest convenience. Sincerely,

Yes
Are the conclusions drawn adequately supported by the results? Yes check whether household size continued to decrease even when Chinese parents are encouraged to give birth to more than one child since 2015).
It would be better to disclose the limitations of previous findings, which would help readers better evaluate different pieces of evidence. For example, in Santos et al.'s (2017) study, they only had 4 data points for China (1990China ( , 1995China ( , 2001China ( , 2007 on three self-report items of cultural values 1 . This could be main limitations of this specific finding and would make the idea that "individualism is decreasing in China" less convincing. The author may provide further critical comments when summarizing the literature.

2.
Since  has provided more valid evidence for the increasing name uniqueness in China 2 and has addressed Ogihara's (2020) 3