社會價值與人口變遷之經驗研究：以台灣女性為案例 - 政大學術集成

全文

(1)國立政治大學社會科學學院亞太研究英語碩士學位學程 International Master’s Program in Asia-Pacific Studies College of Social Sciences National Chengchi University. 治政碩士論文大. 立 Master’s Thesis ‧. ‧ 國. 學. 社會價值與人口變遷之經驗研究：以台灣女性為案例. Nat. sit. A Study of Social Values and Demographic Change:. n. al. e. io. Empirical Evidence and the Implications for Taiwanese Women. Ch. engchi. i n U. v. Student：Torie Jeannette Gervais（紀小薇） Advisor: Dr. Mei-chuan Wei (魏玫娟). y. 中華民國 106 年 07 月 July, 2017.

(2) 社會價值與人口變遷之經驗研究：以台灣女性為案例 A Study of Social Values and Demographic Change: Empirical Evidence and the Implications for Taiwanese Women 研究生：紀小薇. Student: Torie Jeannette Gervais. 指導教授: 魏玫娟. Advisor: Dr. Mei-chuan Wei 國立政治大學. 政治大. 立亞太研究英語碩士學位學程. ‧ 國. 學碩士論文. ‧. n. er. io. al. sit. y. Nat. A Thesis. v. Submitted to International Master’s Program in Asia Pacific ni C. hengchi U Studies. National Chengchi University. 中華民國 106 年 07 月 July, 2017.

(3) ABSTRACT Confucianism has long influenced gender roles and family values in Taiwan. Yet despite Confucianism’s continued influence, fertility rates have dropped below replacement level, marriage is being postponed or rejected, percentages of men and women attending university are roughly equal, and women’s participation in the workforce and the. 政治大. government has increased. One proposed explanation is that Taiwan is following in the footsteps of Europe and undergoing the Second Demographic Transition. If Taiwan’s. 立. demographic changes were connected to similar value shifts as in Europe, women in. ‧ 國. 學. Taiwan would benefit from greater freedom to choose their own life course and greater gender equality. This research sought to determine if there is evidence to support that Taiwan is undergoing the Second Demographic Transition and what the implications are. ‧. for women. Cross-sectional data from 1985 and 2015 general questionnaires and the. y. Nat. 2011 family questionnaire of the Taiwan Social Change Survey was analyzed for a. sit. potential link between value orientation and household type using multinomial logistic. er. io. regression. In regression analyses, evidence did not support that Taiwanese society as a. al. n. iv n C Survey data was supplemented with interviews h ein-depth i U with Taiwanese women to h n c g determine the reasons behind and impact of their union formation and life course choices. whole is experiencing the SDT, though evidence did support value shift over time.. Interview data suggested that some sectors of society are undergoing the SDT while others are not. Furthermore, women reported gender equality and difficulty balancing their work and family life as barriers to following their desired life courses. Finally, this research concludes that further efforts are required to attain women’s freedom and equality in Taiwan.. Keywords: Taiwan, Second Demographic Transition, Social Values, Life Course Choices, Gender equality i.

(4) 摘要. 儒家思想長期影響了臺灣的性別角色以及家庭價值觀。然而儘管儒家思想持續影響，生育率仍下降至低於人口替代率，晚婚或不婚，男性與女性就讀大學的百分比幾乎相等而職場上與公家機關的女性比例亦有提升。其中一個解釋的角度便是台. 治政大體系的轉型有所關連，那麼台灣女性勢必可以從中得到更大的自由去選擇他們的人立生道路以及達到性別平等。本研究的目的便是企圖找到與此論點相關的證據以及探灣也正跟隨歐洲的腳步，歷經第二次人口轉型。如果台灣的人口轉型是與歐陸價值. ‧ 國. 學. 討這種人口轉型對於女性有什麼意義。我將運用「多元邏輯式回歸方法」 (multinomial logistic regression)來分析 1985 到 2015 年之間問卷的交叉數據(cross-. ‧. sectional data)以及 2011 年由中研院主導的台灣社會變遷調查計畫中所做的家庭調. sit. y. Nat. 查，試圖在價值導向與家庭型態當中找出潛在關聯性。在回歸分析法當中，證據不. io. er. 足以顯示台灣正在歷經第二次人口轉型，然則，這些證據卻足以證明台灣社會當中的價值轉變。除了數據分析，本論文也試圖融合對台灣女性的深入訪談進行研究，. n. al. Ch. i n U. v. 探討伴侶型態與不同生命歷程選擇的成因和影響。訪問數據顯示一部份的社群確實. engchi. 正在經歷人口轉型，其餘則不然。再者，這些女性描述了在追求自我的路途上，職場與家庭之間做出平衡的困難。最後，本研究總結台灣的性別平等之路仍須更多努力。. 關鍵字：台灣、第二次人口轉型、社會價值、生命歷程選擇、性別平等. ii.

(5) ACKNOWLEDGEMENTS. Thank you to Academia Sinica for providing data used in this thesis: Data analyzed in this research were collected in the First Round, First Year; Sixth Round, Second Year; and Seventh Round, First Year surveys. 政治大. of the research project "Taiwan Social Change Survey". The project was conducted by the Institute of Sociology, Academia Sinica (data gathered. 立. before the First Year of the Third Round were conducted by the Institute. ‧ 國. 學. of Ethnology, Academia Sinica), and sponsored by the Ministry of Science and Technology (formerly known as National Science Council), Republic of China.. ‧. The author would also like to thank all of the women who volunteered to participate in. y. Nat. sit. this study for being willing to share their stories and perspectives on being a woman in. n. al. er. io. Taiwanese society.. Ch. i n U. v. Additionally, thank you to all those who offered valuable feedback, including my. engchi. advisor, Dr. Mei-chuan Wei, and committee members, Dr. Ping-yin Kuan and Dr. Chiung-chu Lin. Special thanks to Dr. Kuan for providing advice about coding data and conducting a regression analysis. Finally, thank you to friends and family members who have encouraged me and supported my pursuit of higher education. Special thanks to classmate Dinah Gardner for being my comrade-in-arms throughout the thesis process.. iii.

(6) TABLE OF CONTENTS ABSTRACT ......................................................................................................................... i 摘要..................................................................................................................................... ii ACKNOWLEDGEMENTS ............................................................................................... iii LIST OF TABLES ........................................................................................................... viii CHAPTER 1. Introduction ............................................................................................... 1. 政治大. 1.1 Research Background ................................................................................................ 4. 立. 1.2 Purpose of Research .................................................................................................. 8. ‧ 國. 學. 1.3 Chapter Outline ....................................................................................................... 10 CHAPTER 2. Literature Review .................................................................................... 12. ‧. 2.1 Demographic Theories and the Second Demographic Transition .......................... 12. Nat. sit. y. 2.2 Second Demographic Transition Empirical Studies ............................................... 17. er. io. 2.3 Existing Research on Causes of Demographic Change in Taiwan ......................... 21. al. iv n C Research Methods 26 h ................................................................................... engchi U n. 2.4 Theoretical Framework ........................................................................................... 25 CHAPTER 3. 3.1 Value Orientation and Family Formation: A Quantitative Analysis....................... 26 3.1.1 Data ................................................................................................................... 27 3.1.2 Independent variable......................................................................................... 28 3.1.3 Dependent variable ........................................................................................... 28 3.1.4 Hypotheses........................................................................................................ 30 3.1.5 Data Coding ...................................................................................................... 34 3.1.6 Analytical Strategy ........................................................................................... 36. iv.

(7) 3.2 Taiwanese Women’s Perspectives on the Influence of Social Values on Life Course Decisions: A Qualitative Analysis ................................................................................ 43 3.2.1 Sampling ........................................................................................................... 43 3.2.2 Data Collection, Instruments, and Analytical Strategy .................................... 44 3.2.3 Protection of Human Subjects .......................................................................... 45 CHAPTER 4. Mixed Evidence of the SDT in Taiwan ................................................... 46. 4.1 Changes Over Time ................................................................................................. 47 4.1.1 Profile of the Survey Respondents ................................................................... 47. 政治大 4.1.3 Changes in Religious Attitude, Attitude toward Education, and Family Value 立 Orientations ............................................................................................................... 55 4.1.2 Changes in Overall Value Orientation.............................................................. 50. ‧ 國. 學. 4.2 Value Orientation and Household Type in 1985 and 2015 ..................................... 59. ‧. 4.2.1 Regression Models for Overall Value Orientation and Household Type ......... 59 4.2.2 Regression Models for Family and Nonfamily Value Orientations and. Nat. sit. y. Household Type ......................................................................................................... 65. er. io. 4.3 Family Value Orientation and Household Type in 2011 ........................................ 69. al. iv n C 4.3.2 Overall Family Value h Orientation in 2011 ....................................................... 70 engchi U n. 4.3.1 Profile of the Survey Respondents ................................................................... 69. 4.3.3 Family Values by Gender Roles, Family Role, and Attitude Toward Marriage in 2011 ....................................................................................................................... 73 4.3.4 Regression Models for Family Value Orientation and Household Type ......... 76 CHAPTER 5. Taiwanese Women’s Perspectives .......................................................... 83. 5.1 Profile of the Interview Sample .............................................................................. 84 5.2 Value Orientation of Interviewees .......................................................................... 86 5.3 Relationship and Career Life Course Choices ........................................................ 90 5.3.1 Relationship Histories and Future Plans ........................................................... 91 v.

(8) 5.3.2 Work Histories and Future Plans ...................................................................... 93 5.4 Reasons for, Influences behind, and Satisfaction with Life Course Choices .......... 94 5.4.1 Social Expectations and Individual Desires ..................................................... 94 5.4.2 Influences on Decisions .................................................................................... 96 5.4.3 Workplace Equality and Balancing Career and Family ................................... 98 5.4.4 Nontraditional Choices and Societal Tolerance ............................................. 101 5.4.5 Satisfaction ..................................................................................................... 104 5.4.6 Other Important Issues.................................................................................... 107. 政治大 6.1 Factors behind Demographic 立 Behavior and Life Course Decisions ..................... 110. CHAPTER 6. Discussion and Conclusion ................................................................... 109. ‧ 國. 學. 6.1.1 Evaluating the Relationship between Value Orientation and Household Type based on Survey Data .............................................................................................. 110. ‧. 6.1.2 Evaluating the Relationship between Value Orientation and Life Course Decisions based on Interviews with Taiwanese Women......................................... 114. y. Nat. sit. 6.1.3 Second Demographic Transition or Rigid Family Values? ............................ 117. er. io. 6.2 Implications for Taiwanese Women ..................................................................... 121. al. n. iv n C ................................................................................................................................. 122 hengchi U. 6.2.1 Evaluating the Impact of External Pressure to Follow a Traditional Life Course. 6.2.2 Evaluating Avoidance of Marriage and Children ........................................... 123 6.2.3 Evaluating Other Factors that Negatively Impact Women’s Freedom to Determine their Life Courses .................................................................................. 125 6.2.4 Women’s Freedom to Control their Life Courses? ........................................ 126 6.3 Limitations ............................................................................................................ 128 6.3.1 Quantitative Analysis of Value Orientation and Family Formation............... 128 6.3.2 Qualitative Analysis of Impact of Family Formation Life Course Choices ... 128 6.4 Suggestions for Future Research ........................................................................... 129 vi.

(9) 6.5 Conclusion............................................................................................................. 130 REFERENCES ............................................................................................................... 133 APPENDIX A. Taiwan Social Change Survey Modules, Topics, and Sample Size .. 136. APPENDIX B. General Surveys: Taiwan Social Change Survey Questions ............. 137. APPENDIX C. Family Module: Taiwan Social Change Survey Questions ............... 139. APPENDIX D. Coding of Value Questions for General Surveys .............................. 140. APPENDIX E. Coding of Value Questions for Family Survey ................................. 144. APPENDIX F. Stage Two Interview Questions ......................................................... 147. APPENDIX G. Debriefing and Informed Consent Form ........................................... 150. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. vii. i n U. v.

(10) LIST OF TABLES. Table 1.1: Total Fertility Rates (TFR) for Select Countries ................................................5 Table 1.2: Subreplacement Fertility for Select Countries ....................................................5 Table 1.3: Singulate Mean Age at Marriage in Select Countries ........................................6 Table 1.4: Proportions Single by Gender and Age Cohort, Taiwan 1985-2016 ..................7 Table 3.1: Hypotheses for the Relationship between Value Orientation and Household. 政治大. Type for General Survey Round 1 (1985) and Round 7 (2015) and Family Module Survey Round 6 (2011) ......................................................................33. 立. Table 3.2: Analytical Framework for Multinomial Logistic Regression Models for. ‧ 國. 學. General Survey Data for Round 1 (1985) and Round 7 (2015) .......................39 Table 3.3: Analytical Framework for Multinomial Logistic Regression Models for Family. ‧. Module Survey Data for Round 6 (2011) ........................................................42. sit. y. Nat. Table 4.1: General Survey Sample Profiles by Age, Gender, Marital Status, Education. io. er. Level, and Income, Rounds 1 and 7 .................................................................48 Table 4.2: Overall Value Orientation Percentages and Value Orientation Percentages by. n. al. Ch. i n U. v. Age, Gender, Education, and Income for General Surveys, Rounds 1 (1985). engchi. and Round 7 (2015) .........................................................................................51 Table 4.3: Value Percentages by Religious Attitude, Attitude Toward Education, and Family Values for General Surveys, Round 1 (1985) and Round 7 (2015) ....56 Table 4.4: Multinomial Logistic Regression Results for Overall Value Orientation for General Surveys in 1985 and 2015 ..................................................................61 Table 4.5: Multinomial Logistic Regression Results for Family and Nonfamily Value Orientations for General Surveys in 1985 and 2015 ........................................67 Table 4.6: Family Survey Sample Profile Age, Gender, Marital Status, Education Level, and Income, Round 6 (2011) ...........................................................................71 viii.

(11) Table 4.7: Overall Family Value Orientation Percentages and Value Orientation Percentages by Age, Gender, Education, and Income for Family Module Survey, Round 6 (2011) ...................................................................................72 Table 4.8: Family Value Percentages by Gender Role, Family Role, and Attitude Toward Marriage for Family Module Survey, Round 6 (2011) ....................................75 Table 4.9: Multinomial Logistic Regression Results for Overall Family Value Orientation for Family Module Survey in 2011 ..................................................................78 Table 4.10: Multinomial Logistic Regression Results for Gender Roles, Family Role, and Attitude toward Marriage Value Orientations for Family Module Survey in. 政治大. 2011..................................................................................................................81. 立. Table 5.1: Profile of the Interview Sample, Sorted by Age ...............................................85. ‧ 國. 學. Table 5.2: Relationship Histories of the Interview Sample ...............................................92. ‧. n. er. io. sit. y. Nat. al. Ch. engchi. ix. i n U. v.

(12) CHAPTER 1. Introduction. Women’s life choices have long been constrained by women’s secondary status and the role society expects them to play. In Taiwan and other culturally Chinese societies, men’s and women’s roles have been influenced historically by Confucianism. After the Nationalists (Kuomintang) took control of Taiwan in 1945, their policies toward. 治政大curriculum, imbuing in children culture. The party controlled the education system and 立 the ideas of patriarchy. Both national education and state media glorified “virtuous wives women placed particular emphasis on Confucian values to preserve traditional Chinese. ‧ 國. 學. and good mothers,” advocating that women play a supporting role to the males in their family (Chang, 2009, p. 62-66). Hsiu-lien Annette Lu, a forerunner of feminism in. ‧. Taiwan, made it her goal in the 1970s to “challenge Kuomintang-sponsored Confucianism,” including double standards in the treatment of men and women, such as. Nat. sit. y. female chastity, preference for male children, respecting men but demeaning women, and. io. al. n. p. 76, 89-98).. er. the belief that women’s place is in the home while men’s place is outside (Chang, 2009,. Ch. i n U. v. Since the end of martial law in 1987, women’s rights have seen great. engchi. improvements. Laws have become much more equal since activists began campaigning for women’s rights in the late 1980s and since the mid-1990s women have played a greater role in politics. Access to education has also greatly increased. In 1971 women were 36.9 percent of students that passed the college entrance exam and by 2002 were 49.1 percent of all college students (Chang, 2009). In 2013 for the first time the proportion of women with higher educations surpassed the proportion for men. In 2015 42 percent of women had a university degree, compared to 38 percent of men (DGBAS, 2017a). Women’s workforce participation has increased, even after marriage and childbirth. In 1960 only 25 percent of women over the age of 15 worked (Ferris, 2004). 1.

(13) In 1970 that number had increased to about 30 percent and in 1990 to 45 percent (Chang, 2009). In 2015 women’s proportion of the labor force was 44.3 percent (DGBAS, 2017a). The percentage of women in skilled jobs has increased, from 34.8 percent of professionals and technical specialists and 8.9 percent of managers and administrators in 1971 to 43.6 percent and 14.6 percent respectively in 2002 (Chang, 2009). In 2015, women’s share of professional and technical positions reached 50.8 percent, passing the 50 percent mark for the first time (DGBAS, 2017a). Other significant social changes that influence women’s status are marriage and fertility rates. Women have been marrying later and having fewer children. Despite. 政治大. Confucian stress on procreation, the fertility rate has been falling since the 1960s (Leete, 1994). In 1970 the average number of children was four but by 2013 it had dropped to. 立. 1.07 (Cheung, 2015). Also, more middle class women are simply not marrying at all,. ‧ 國. 學. leading to the new term “single nobility” (dansheng guizu), referring to their high standard of living and freedom from family responsibilities (Chang, 2009, p. 163). A 1990 Census showed that in Taipei 19 percent of 30-34-year-old women and 11 percent. ‧. of 35-39-year-old women were unmarried (Leete, 1994). In 2010 the proportions. sit. y. Nat. remaining unmarried had increased to thirty-seven percent of women ages 30 to 31 and 21 percent of women ages 35 to 39 (Osteria, 2015). This suggests greater freedom to. er. io. make one’s own marriage choices. Combined with higher levels of education and more. al. n. iv n C and enjoy higher status than Taiwanese in the past. h e nwomen gchi U. women working in higher level positions, many women are now less dependent on others. Though women’s status has improved, it has not necessarily done so equally at all levels of society and women are still disadvantaged overall in comparison to men. In the working class, many still regard more than nine years of compulsory education unnecessary for women and rely on women’s manual labor in family businesses (Moskowitz, 2007). Rising wages make it difficult for families to keep small-scale manufacturing businesses running and some families force their daughters to work for far lower wages than they would need to pay a hired worker. Family obligations can also force women to stay in unhappy marriages, force daughters into prostitution, and hide. 2.

(14) domestic violence and rape. While not the norm, Moskowitz (2007) and Lee (2004) have conducted interviews and recorded such cases. In less extreme situations, a disproportionate amount of domestic tasks continues to fall on women. About 50 percent of women not in the workforce in 2016 gave “housekeeping” as the reason for not working (DGBAS, 2017a). Even for working women, expectations to be a “virtuous wife and good mother” are still common (Ferris, 2004, p. 358-359). Due to a combination of lower fertility rates causing an aging population and a lack of nursing homes, elderly care is also the responsibility of women unless they can afford to hire a domestic worker. Before marriage, parents still have a. 政治大. significant amount of power over a woman’s choice of spouse even though marriages are no longer arranged. After marriage many women lose their connection with their birth. 立. family because women are still seen as “marrying out,” though this has been slowly. ‧ 國. 學. changing. Higher levels of education can also make it difficult for women to find a partner. Traditionally women’s education was not seen as important so it was acceptable for a man to marry someone less educated but for a woman to do the same is considered. ‧. downward social mobility (Leete, 1994).. Nat. sit. y. In the workforce, despite laws requiring equal pay for equal work, views that. io. er. women should not receive the same levels of pay have lingered. Marsh (2004) compared survey answers from 1963 and 1991 measuring men’s attitudes toward equal pay and. n. al. Ch. i n U. v. found shockingly little change. In 1963, 67.6 percent said there should be no difference. engchi. while 28.6 percent said women should get less pay. In 1991 the percentage saying there should be no difference only increased to 73.1 and the percentage saying women should get less only decreased to 24.2. The percentage of men agreeing to equal pay in 1963 may have been progressively high, but the percentage in 1991 is incredibly low considering how many other social changes occurred over that time period. When asked for the reasons behind their response, men who believed women should not receive equal pay gave answers like women being innately less qualified, women only working for spending money rather than to support their family, and that women cannot devote themselves to work in the same way as men because caring for their family would interfere. When comparing the reasons behind not supporting equal pay among subjects. 3.

(15) in 1963 and 1991, Marsh found that the percentages had not greatly changed, but some of the reasons had. In 1963, answers were linked to men’s elitism, interaction with kin, and living with a married son. The latter two were seen as indicators of a more traditional lifestyle living in an extended family. In 1991, answers were linked to large workplaces, owning the workplace, living with a married son, and having a male heir. The last two are again linked to a more traditional lifestyle. This shows a continuing influence of Confucian values on men’s attitudes toward women. Evidence of continued Confucian influence is supported by Fetzer & Soper’s (2007) analysis of survey data determining the impact of Confucian values on Taiwanese. 政治大. society. They looked at the 1995 World Values Survey and 2001 East Asian Barometer and found questions that aligned with the Confucian values of family loyalty, social. 立. hierarchies, and social harmony. Then they measured support for those values and how. ‧ 國. 學. they impacted democracy and human rights. They found strong evidence of influence: One important finding from our analysis is that Confucian values do not. ‧. consistently undermine liberal democracy…Only on women’s rights do we find any possible pattern of conflict between human rights and. Nat. sit. y. Confucian values. Social harmony, in particular, seems always to decrease. io. er. adherence to the rights of women. Family loyalty, on the other hand, showed a similar pattern in 1995, but by 2001 was boosting support for. n. al. women’s rights (p. 53).. Ch. engchi. i n U. v. The mixed influence of traditional Confucian ideals on women’s rights shown in Fetzer & Sopers’s findings suggests that further research is needed to determine how values and value change affects women’s status in Taiwanese society.. 1.1 Research Background As mentioned in the previous section, marriage and fertility tends have been changing in Taiwan. Many scholars have noted the significant changes in marriage and fertility trends in East Asia beginning in the latter half of the twentieth century. The first of these changes to occur was dropping fertility rates. Japan led the way, with fertility 4.

(16) rates declining initially from 1925 to 1940, again in 1949, and reaching replacement levels by 1957 (Atoh, Kandiah & Ivanov, 2004). Beginning about the same time Japan reached replacement fertility, other countries in Asia began their declines. Singapore and Taiwan began their fertility decline in the 1950s, with a total fertility rate (TFR) of 6.0, and reached replacement in 1975 and 1983 respectively. Hong Kong’s TFR declined from 5.2 in 1960 to replacement in 1979 and Korea’s TFR declined from 6.0 in 1960 to replacement in 1984. Family planning policies and effective contraception likely contributed to these declines. Once reaching replacement levels, fertility rates continued to drop and have remained at subreplacement levels for the past 10-25 years in Japan, Singapore, Hong Kong, Korea, and Taiwan (Atoh, Kandiah & Ivanov, 2004). This trend. 政治大 Taiwan had the lowest TFR at 1.07 in 2013 (Cheung, 2015). Since hitting a low of 0.90 立 has been dubbed the “East Asian childbearing pattern” and among these countries,. 學. ‧ 國. in 2010, Taiwan’s TFR increased to 1.18 in 2015 (DGBAS, 2016). See Tables 1.1 and 1.2 for detailed fertility information in the region.. Table 1.1: Total Fertility Rates (TFR) for Select Countries, 1970-2013 South Korea. 3.42. 2.14. 3.09. ‧. Taiwan. 4.53. 4.00. 2.05. 1.75. 1.74. 1.27. 1.54. 1.87. 2000. io. 1.36. 1.60. 2010. 1.13. 1.39. 1.15. 2013. 1.13. 1980 1990. al. n. Source: Cheung, 2015. 1.04. C h 1.40 1.19 engchi U. 2.52. 1.59. 1.81. 1.47. 1.68. 1.23. 0.90. 1.19. 1.07. 2.83. er. 1970. y. Singapore. sit. Japan. Nat. Hong Kong. Year. v ni. Table 1.2: Subreplacement Fertility for Select Countries Year Reaching Subreplacement TFR 1981. No. of Years of Subreplacement TFR 32. Japan. 1975. 38. 2005 (1.260). Singapore. 1977. 36. 2010 (1.150). South Korea. 1984. 29. 2005 (1.076). Taiwan. 1985. 28. 2010 (0.895). Hong Kong. Source: Cheung, 2015. 5. Year of Lowest TFR 2003 (0.901).

(17) There have also been major changes in marriage, especially in ages at first marriage and increased divorce rates. Mid twentieth century, marriage in the Asia Pacific was nearly universal and being unmarried was considered an “aberration” (Jones, 2004). Over the last 40 years that fact has been changing. Looking at the singulate mean age at marriage (SMAM) for women from 1970 to 2010, Singapore, Japan, South Korea, and Taiwan have all seen increases in age of 3.7, 5.0, 6.8, and 7.8 years respectively (Jones, 2014). See Table 1.3 for details of SMAM trends in the region. During the period of subreplacement fertility in Taiwan from 1985 to 2016, the median age at first marriage has increased from 27.6 to 32.0 for men and 24.4 to 29.7 for women (Ministry of the Interior, 2017b). Delayed marriage is also a major contributing factor to subreplacement. 政治大. fertility (Cheung, 2015; Jones, 2007; Lesthaeghe 2010).. 立. 30.8. 31.2. Increase (years) 3.7. 26.9. 28.6. 29.7. 5.0. 28.4. 29.9. 30.0. 2.5. 24.2. 26.2. 27.0. ‧. 30.3. 26.5. 27.9. 3.7. 30.3. y. 32.9. 7.7. 27.1. 30.1. 6.8. 32.7. 8.1. 30.3. 7.8. Singapore South. Men. 27.2. 27.3. 28.5. Korea. Women. 23.3. 24.1. 25.5. 24.6. 25.3. 28.8. 23.9. 26.0. Japan. 1970. 1980. 1990. Men. 27.5. 28.7. 30.4. Women. 24.7. 25.1. Men. 27.8. Women. Nat. io. Men. Women. Source: Jones, 2014. a l22.6 Ch. n. Taiwan. engchi. sit. 學. 2010. ‧ 國. 2000. er. Table 1.3: Singulate Mean Age at Marriage in Select Countries, 1970-2010. i n U. 30.5. v 27.6. Proportions of the population marrying have also been dropping as rates of those remaining single have increased since the 1970s (Osteria, 2015). Singapore, Japan, Korea, Taiwan, and Hong Kong have all seen increases in the percentages of singles at ages thirty and over. In Japan, South Korea, and Taiwan, this trend has affected the entire population, not just those highly educated. In Taiwan there is evidence of a trend toward lifelong singlehood (Osteria, 2015). By the end of the year in 2016 in Taiwan, 50.9 percent of the population aged 30 to 34, 31.5 percent of the population aged 35 to. 6.

(18) 39, and 21.6 percent of the population aged 40 to 44 remained unmarried (Ministry of the Interior, 2017a). See Table 1.4 for more details of singlehood trends in Taiwan. Table 1.4: Proportions Single by Gender and Age Cohort, Taiwan, 1985-2016 Age 30-34. 35-39. 40-44. 1985. 1990. 1995. 2005. 2010. 2016. Men. 16.2. 22.6. 35.4. 41.2. 54.1. 59.4. Women. 9.4. 12.3. 16.0. 26.9. 37.2. 42.5. Men. 7.9. 10.6. 18.3. 21.0. 29.0. 37.3. Women. 5.7. 7.5. 9.0. 14.8. 20.9. 25.8. Men. 5.7. 6.8. 10.8. 12.7. 17.8. 24.2. Women. 3.2. 5.5. 6.7. 9.5. 14.4. 19.1. 政治大. Sources: Osteria, 2015; Ministry of the Interior, 2017a. 立. In addition to delaying marriage or remaining single, divorce rates in East Asia. ‧ 國. 學. have been rising. Rates in Japan, South Korea, and Taiwan are now similar to those seen in Europe (Jones, 2015). The crude divorce rate in Taiwan has increased from 0.83 in. ‧. 1981 to 2.83 at its peak in 2006, and then decreased to 2.28 in 2015 (DGBAS, 2016). Jones (2015) attributes this change to women’s increasing economic independence,. sit. y. Nat. pressures on marriage from living in urban environments, and “perhaps a retreat from. io. al. er. Confucianist and other familial norms toward greater individualism” (p. 333).. v. n. There is also some evidence of cohabitation in Asian countries, though it is. Ch. i n U. difficult to find surveys that ask detailed questions about it. According to data from the. engchi. World Value Survey Wave 6 (2010-2014), Japan, South Korea, Singapore, and Taiwan all have low levels of current cohabitation reported (less than two percent). Among those who did cohabitate, the greatest proportion in Taiwan was the 20 to 29 age cohort, in Korea were the 20 to 29 and 30 to 39 cohorts, and in Singapore and Japan were those age 40 and over. National Japanese surveys tell a different story: 20 percent of Japanese women aged 25 to 34 had previously cohabitated, double the percentage from ten years’ prior (Kobayashi & Kampen, 2015). According to KAP (knowledge, attitude, and practice) Surveys in 1998 and 2004 in Taiwan, women ages 20 to 49 who had ever cohabitated rose from 11.4 percent to 19.6 percent. Better information on cohabitation. 7.

(19) will not be available until more surveys include the “ever cohabitated” question (Lesthaeghe, 2010). Overall, recent demographic trends in Japan, South Korea, Singapore, and Taiwan can be summarized by sustained subreplacement fertility, delayed marriage, decrease in overall percentage married, increase in percentage of those never married, increase in divorce rates, and limited evidence of cohabitation. Currently out of wedlock births are still uncommon, less than two percent of births for Japan, South Korea, and Singapore, and three percent in Taiwan (Cheung, 2015). Premarital sex may still result in conception, and “shotgun” marriages for this reason are occurring in Japan (Lesthaghe, 2010).. 立. 政治大. ‧ 國. 學. 1.2 Purpose of Research. The gains in status that women in Taiwanese society have achieved despite the continued influence of Confucianism and the demographic changes occurring suggest. ‧. that Taiwan might be undergoing the Second Demographic Transition (SDT). Delayed. y. Nat. marriage, lower fertility rates, increased women’s workforce participation, and higher. sit. education levels for women are all features of the transition. According to Lesthaeghe. er. io. (2010), as a society undergoes the SDT, there is a change from focusing on meeting. al. n. iv n C preference drift.” (p. 213). SDThnon-material needsU e n g c h i are called “higher order needs” and include “individual autonomy, expressive work and socialization values, selfmaterial needs to seeking to fulfill non-material needs: this is referred to as “Maslowian. actualization, grass-roots democracy, and recognition” (Lesthaeghe, 2014, p. 18113). This change in societal values, combined with high levels of development, is the driving force behind the demographic changes of the transition. However, researchers have questioned whether the Second Demographic Transition is a universal global phenomenon or if it is a feature of certain Western societies. The purpose of this research was to analyze whether Taiwan is undergoing the Second Demographic Transition and to determine what the implications of current social values are for women in Taiwanese society. Analyzing this phenomenon from a feminist. 8.

(20) perspective, the shift in societal values associated with the SDT can benefit women by normalizing more equal gender roles and giving women more freedom to control their life choices as more lifestyles become socially acceptable. Traditionally Confucianism has portrayed men and women as being divided into the respective categories of yang and yin as the natural state of affairs. Not only is this a false dichotomy but these roles are social constructions that are harmful for individuals in society. Women are not monolithic and there cannot be only one life course appropriate for all women. The main research questions guiding this thesis were: . Is Taiwan undergoing the Second Demographic Transition?. . What are the implications of current social values for Taiwanese women?. 政治大. 立. To answer the question of whether Taiwan is undergoing the Second Demographic. ‧ 國. 學. Transition, this research analyzed data collected by Academia Sinica’s Taiwan Social Change Survey. The motivation behind investigating this question was that if Taiwan is in the midst of SDT, then it seems likely that Maslowian preference drift would conflict. ‧. with Confucian values. Maslow’s higher order needs are more focused on the individual. y. Nat. and individual satisfaction while Confucianism emphasizes social harmony. There were. io. sit. two sub-questions guiding the survey data analysis:. Are social values in Taiwan shifting over time?. . Is there a relationship between an individual’s value orientation and. n. al. er. . Ch. household type?. engchi. i n U. v. To evaluate these questions, survey questions aligning with Confucian and other traditional values and higher order values associated with the SDT were compared over a time span of thirty years to see if there is evidence of a social value shift and if there is a relationship between values and the type of household in which the respondent lives. While there was clear evidence of value shift, this research’s findings failed to find a connection between values and demographic behavior, represented by household type. Survey data was supplemented with interviews with local women. Interviews were designed to determine a woman’s value orientation, her life course choices, and the. 9.

(21) reasons for, influences behind, and satisfaction with her life course choices. There were four sub-questions guiding the interview data analysis: . Are women’s values related to their life course choices?. . Are women being pressured by society or their families to follow a traditional life course?. . Are unmarried women actively choosing to avoid marriage and children?. . What circumstances in Taiwanese society negatively affect women’s ability to follow their desired life course?. Interview data provided a possible explanation for the lack of a connection. 政治大 sample was following behavior predicted by the SDT. In regard to social values and their 立 impact on women, most women in the sample reported instances of gender inequality and between values and household type in regression analyses: only part of the interview. ‧ 國. 學. the difficulty of having both a career and family as constraints on their choices. The findings of this research have important implications for social policy in. ‧. Taiwan, especially regarding population and gender equality. The findings of this. y. Nat. research cannot conclude that ideational shift alone is the driver of demographic change,. sit. but rather suggest that gender inequality is influencing women’s behavior. Policy needs. er. io. to address the barriers preventing women from having both a family and career.. al. n. iv n C will provide the wrong incentives. to thisU h eAccording i research’s findings, many women h n c g feel like they cannot have a family without hurting their careers and some women are. Otherwise any attempts to encourage fertility and marriage will be in vain because they. avoiding the burdens of marriage and children. From these findings it is clear that further steps need to be taken to attain greater equality in Taiwanese society.. 1.3 Chapter Outline The following chapters explore the social reasons behind demographic changes occurring in Taiwan since the 1980s and the impact the changes in social climate have on Taiwanese women’s choices regarding their life courses. Chapter Two discusses Western theories developed to explain similar demographic changes in Europe, research 10.

(22) approaches used in both the West and East, existing research on Taiwan’s demographic changes and social values, and the theoretical framework used in this research. Chapter Three explains the twofold analytical approach of this thesis: a quantitative analysis of values and household type in survey data and a qualitative analysis of Taiwanese women’s perspectives on how social values influence their decisions regarding life course choices. Chapter Four presents the findings of the survey data analysis and Chapter Five presents the findings of the interviews. The final chapter, Chapter Six, discusses the meaning and implications of the findings, the limitations of the research, suggestions for future research, and the final conclusions of this thesis.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 11. i n U. v.

(23) CHAPTER 2. Literature Review. The demographic changes that began occurring in Asia in the second half of the twentieth century were preceded by similar trends in Europe. For this reason, many of the main demographic theories were developed in the West to explain the particular changes in. 政治大. behavior in that part of the world. Later scholars noticed the similarity of demographic trends in Asia to those that occurred in the West and proposed that some of these theories. 立. may apply outside of the West as well. Other scholars agreed that the demographic. ‧ 國. 學. changes appeared similar but disagreed that the causes behind them were the same. This chapter gives an overview of theories and approaches for explaining and determining the cause of demographic trends. Section 2.1 summarizes the main relevant Western. ‧. demographic theories and section 2.2 outlines important research approaches of existing. y. Nat. studies analyzing demographic transition in the West and in the East. Section 2.3. io. sit. summarizes important recent studies that explain the causes of demographic change in. n. al. er. Asia that also include Taiwan in their analyses. Finally, section 2.4 explains the theoretical framework of this research.. Ch. engchi. i n U. v. 2.1 Demographic Theories and the Second Demographic Transition There are three main approaches to analyzing changes in fertility, marriage, and family formation trends: those that analyze economic causes, those that analyze sociocultural causes, and those that incorporate both types of causes into their analysis. In this section key theories for each of these approaches will be summarized. Primarily theories from 1980 on will be considered as the demographic changes of interest to this research first emerged in the 1970s.. 12.

(24) The neoclassical economic approach to fertility behavior was introduced by Becker (1960). His focus was to find a better explanation for the fertility trends that contradicted the Mathusian assumption that increased income would result in more children. Becker characterized children as a “consumption good” that is a source of utility for parents and the number of children would be determined by factors like consumer tastes, “quality” of children (economic quality, not moral), family income, cost of children, and supply (ability to have children and contraception allowing tradeoffs in quantity and quality of children). Becker’s framework, while adding an important facet to the consideration of fertility cannot completely predict fertility choices. In his own words, by adding “tastes,” to his framework it “permits, although it does not predict,. 政治大 argues that the tradeoff between quantity and quality of children is a key determining 立. fertility differences that are unrelated to "economic" factors.” (p. 211). Becker (1981). factor in fertility trends.. ‧ 國. 學. Easterlin (1980) used a macroeconomic model to explain the baby boom and subsequent bust of the 1960s and 1970s. He proposed a cyclical model of fertility in. ‧. which every other generation would vary in relative size, resulting in alternating. sit. y. Nat. conditions. Generations with relatively smaller population sizes would have better opportunities for employment while generations with larger population sizes would have. io. er. worse opportunities. Better opportunities would lead to earlier marriage and higher. al. n. iv n C and lower fertility. Lesthaeghe (2014) out that the cyclical theory is not supported h e npoints gchi U fertility rates, thus creating the next large generation which would have delayed marriage. by the sustained subreplacement fertility that spread through Europe beginning in the 1970s. Cleland and Wilson (1987) challenge the predominant economic approach to. explaining fertility. In examining data from multiple developing countries, they found that surveys only showed drops in desired number of children after fertility rates had already begun falling, proving that an economic model determining fertility preference through demand could not have been the cause. Their study concludes that ideational change is a far more likely to affect fertility behavior.. 13.

(25) Second Demographic Transition theory integrates many different features beyond fertility into one flexible model. The concept was first proposed by Lesthaeghe and van de Kaa in 1986 (Lesthaeghe, 2014). It does not disregard either economic or cultural explanations, but considers them both insufficient when considered alone and nonredundant when considered together. Though a certain level of development is required for the transition to begin, the main focus of analysis is ideational change. Unlike many other theories that separate demographic features and look at their causes in isolation (such as theories that analyze fertility without considering marriage), SDT theory incorporates fertility trends, marriage trends, and other life course choices as related to societal changes. The basic premise of the theory initially was that “change in. 政治大 including the dissolution of unions” (van de Kaa, 2002, p. 6). 立. attitude toward the family was seen to affect the whole process of family formation,. ‧ 國. 學. The SDT originated as an explanation for demographic phenomena that began in the 1970s, such as “the baby bust, the systematic postponement of marriage and parenthood, subreplacement fertility, the rise of alternative forms of partnerships, and. ‧. parenthood outside marriage” (Lesthaeghe, 2014, p. 18112). It is a theory explaining. sit. y. Nat. changes in fertility and union formation trends, using Maslow’s theory of shifting needs. io. er. as the underlying cause of ideational change driving the transition. Maslow’s theory (1954) predicted that increasing wealth and higher levels of. n. al. i n U. v. education, such as seen in advanced industrialized countries, would lead to “shifts away. Ch. engchi. from needs associated with survival, security, and solidarity. Instead greater weight is attached to individual self-realization, recognition, grassroots democracy, expressive work, and educational values” (Lesthaeghe, 2014, p. 18113). SDT theory combines Maslowian preference drift with the idea that neither neoclassical economic logic nor cultural features are sufficient explanations for demographic change. Both are integrated into the theory “by treating ideational changes as exogenous influences that add stability to trends over and beyond economic fluctuations” (Lesthaeghe, 2014, p. 18113). Second Demographic Transition theory rejects the conclusion that the First Demographic Transition (FDT) would result in stable populations characterized by zero growth, high life expectancy, and replacement fertility. It furthermore rejects cyclical 14.

(26) fertility theory proposed by Easterlin (1980), in which economic booms and busts would lead to corresponding fluctuations in fertility. Both of these theories suggest “equilibrium as the end point” of fertility trends (Lesthaeghe, 2014, p. 18112). SDT instead predicts “sustained subreplacement fertility” combined with “a multitude of living arrangements other than marriage, a disconnection between marriage and procreation, and no stationary population” (Lesthaeghe, 2014, p. 18112). Due to low fertility, migration is a key element in offsetting population decline. To understand the Second Demographic Transition, it is necessary to compare it to the First Demographic Transition and highlight the changes. There are three main. 政治大. aspects to compare: marriage trends, fertility trends, and social changes. The FDT spanned from the 1860s to the 1960s. The percentage of the population that was married. 立. rose and remarriage in cases of widowhood or divorce were the norm. Divorce was. ‧ 國. 學. uncommon, as was cohabitation. The age at first marriage also dropped, hitting the lowest point in the 1960s. Beginning in the 1970s with the onset of the SDT, these trends sharply changed. First marriage was delayed, overall percentages of married couples. ‧. decreased, the divorce rate increased, and remarriage became less common. More. y. Nat. couples lived together prior to marriage and cohabitation became an alternative to. io. sit. remarriage. For some couples, cohabitation or living separately (“living apart together,”. n. al. er. or LAT) replaced marriage (Lesthaeghe, 2014, p. 18112-18113).. Ch. i n U. v. For fertility trends, the First Demographic Transition was characterized by lower. engchi. ages at first birth and lower fertility rates at later ages due to contraception, though contraception was also prone to failure. The majority of childbearing occurred within marriage and it was unusual for a married couple not to have children. During the Second Demographic Transition fertility was affected by the three revolutions of the 1960s: contraceptive, sexual, and sex (gender). More effective contraception allowed first births to be postponed while the sexual revolution meant lower ages at becoming sexually active. Questioning of divided gender roles and more women working outside the home made delayed birth more desirable for women. The result of later ages for first birth was subreplacement fertility. Additional SDT fertility changes included more. 15.

(27) children being born outside of marriages and more cases of childfree1 couples (Lesthaeghe, 2014). Finally, the social conditions of the First Demographic Transition were the foundation on which the Second Demographic Transition was built. During the FDT period, society was primarily focused on “increasing household real incomes, improving working and housing conditions, raising standards of health, improving human capital through mass education, and providing a safety net for all via the gradual construction of a social security system” (Lesthaeghe, 2014, p. 118114). To meet these needs, religious and political institutions encouraged the nuclear family unit with a gendered division of. 政治大. labor: this was seen as the best way to meet material needs while protecting morality from the corruption of industrialization. Participation in community organizations was. 立. also a key characteristic of this time period, leading to a strong emphasis on solidarity.. ‧ 國. 學. Once the material concerns of the FDT were met, “higher order needs” emerged that were associated with the SDT and “centered on a triad: self-actualization in formulating goals, individual autonomy in choosing means, and a claim of recognition for their. ‧. realization” (Lesthaeghe, 2014, p. 118114). Individualization and nonconformity. sit. y. Nat. replaced solidarity, leading to a rejection of authority, a rise in secularization, withdrawal from community organizations, and alternative choices to the rigid family model.. io. n. al. er. Relating to gender, the SDT also resulted in more equal roles, higher levels of education. i n U. v. for women, and more financial independence for women (Lesthaeghe, 2014).. Ch. engchi. The Second Demographic Transition was first observed in Northern and Western Europe and some critics doubted the universality of the trend. However, after the collapse of communism in 1989, the features of the SDT immediately appeared in Central and Eastern Europe (Lesthaeghe, 2010). Southern Europe, initially thought to be an exception to the theory, experienced a delayed but more extreme version of the transition. Marriage and fertility occurred at even later ages and “scarcely exhibited signs of fertility recuperation after age 30” (Lesthaeghe, 2010, p. 222). Combined with twenty plus Current feminist discourse distinguishes between “childless” and “childfree.” The former is thought to be pejorative because it implies one is lacking something and the latter is considered to support women’s choice to not have children. This research uses the word “childfree” to refer to women who do not have children, whether due to deliberate choice or other factors. “Childless” is not used in order to avoid stigmatizing women who do not have children. 1. 16.

(28) percent of younger female generations never having children at all, SDT in Southern Europe has resulted in “lowest-low fertility.” Southern Europe also differs from Northern and Western Europe due to its “strong family system” in which children live with their parents until marriage, women may move in with their husband’s family, traditional gender roles are more persistent, and elderly parents rely on their grown children (Lesthaeghe, 2010, p. 223). The family system explains why there was a lag in cohabitation and having children outside of marriage in Southern Europe when compared to Northern and Western Europe: moral values in a culture can delay certain aspects of the SDT (Lesthaeghe, 2010).. 政治大. Lastly, Second Demographic Transition theory predicts that its demographic changes “are likely to emerge in non-Western societies, provided that they equally. 立. develop a greater accentuation of Maslowian “high order needs” in tandem with the. ‧ 國. 學. growth of solid democratic institutions protecting respect for diversity” ((Lesthaeghe, 2014, p. 118113). Allowances are made for the culturally specific context as “historical path dependency” can result in “heterogeneity in the pattern of the SDT” (Lesthaeghe,. ‧. 2010, p. 225). Therefore, it is important to test the model using empirical data in each. sit. y. Nat. location in which the SDT is suspected to be occurring. Lesthaeghe (2010; 2014) concludes that while some evidence exists to suggest the transition is underway in Asia,. io. n. al. er. more research is needed.. Ch. engchi. i n U. v. 2.2 Second Demographic Transition Empirical Studies This section summarizes key empirical studies that test if the SDT is occurring. First, it outlines two relevant approaches (Lesthaeghe & Surkyn, 1988; Surkyn & Lesthaeghe, 2004) that have found empirical evidence in Western countries to support the existence of the SDT. Second, this section outlines the main study that tests if the SDT is occurring in Asian countries. Specifically, focuses on studies that link value changes to demographic features using value surveys because that is the source of data this research utilized as well.. 17.

(29) Lesthaeghe and Surkyn (1988) is the first empirical attempt at proving the SDT’s existence in Europe. The study operationalizes a cohort based and education-driven model of ideational change and uses data from the 1981 European Values Studies (EVS) to test the association between values and demographic features. The study identifies secularization and individuation as especially important in value changes for European countries and breaks its operationalization into three parts. Part one of the operationalization proves the connection between specific values and either lower order or higher order needs. Using the EVS for European Economic Community members, it uses data from questions divided into the following categories. 政治大. and items: Religion and morality (traditional beliefs, institutional religion, personal religiosity, Ten Commandments morality, permissiveness), qualities to be transmitted. 立. through education (conformism, pluralism, independence, imagination and leadership),. ‧ 國. 學. political values (lower order versus higher order needs based on the Inglehart scale, leftism, protest proneness, confidence in institutions, political involvement, national pride, surpranationalism), work values (personal satisfaction from and self-fulfillment in. ‧. work, material condition) and marriage and family (traditionalism, freedom in sexual. sit. y. Nat. matters, meaning of parenthood). Each of these items is linked with either institutional regulation and lower order needs or individualization and higher order needs. Correlation. io. n. al. er. coefficients were found between each of the items and individualization and higher order needs with the expected results.. Ch. engchi. i n U. v. Part two of the operationalization looks at value changes and the role cohorts and education play in how those changes occur. Theoretically, values are formed in early life from either parental values or education and peers are then held throughout life. This allows distinguishing between cohorts that conform to their parents’ value systems and those that break with the previous generation. To prove the validity of using an education-cohort model, the study uses average factor values for the items listed in part one for two age groups with two levels of education. The results show that items associated with lower order needs increased with age and decreased with education while items associated with higher order needs decreased with age and increased with education. Supporting proof of a pure cohort effect with respect to values are findings of. 18.

(30) cohort-specific, nonintersecting scores for Inglehart’s index of postmaterialism for six European Economic Community countries measured from 1970 to 1986. The last part of the operationalization looks at family formation values and other values listed in part one. Here the study notes the limitation of using EVS data because it lacks sufficient demographic information. This section finds the correlation coefficient of the three main family formation values (tolerance of nonconformism, meaning of parenthood, preference for three or more children) with variables measuring religious, moral, and political values. The study found that religiosity and morality were most important in determining family values for all ages and genders. The political values. 政治大 Surkyn and Lesthaeghe (2004) explore the connection between “value 立 orientations and the life course choices concerning living arrangements and family were also related as expected, but with less strength than religion and morality.. ‧ 國. 學. formation,” and assert that this connection is a “crucial element” of the Second Demographic Transition (p. 45). The data is drawn from the 1999-2000 European Values. ‧. Surveys and compares three groups of European countries: Western Europe (Belgium, France, Germany), Iberia (Portugal, Spain), and Scandinavia (Sweden, Denmark). The. Nat. sit. y. analysis is divided into two parts.. er. io. Part one analyzes the connection between 80 value items and type of household.. al. iv n C single or returned home after partnership), single (living h e n g c h i U alone, never married, not in a partnership, no children), cohabitating (unmarried, no children), cohabitating with n. Households are broken into eight types: single and living with parents (either always. children, married without children, married with children but never previously cohabitated, married with children and previously cohabitated, and formerly married or cohabitating and divorced or separated but not yet in a new union. Values are related to the following topics: marriage and family, religion, civil morality, politics, identification, retreat (from organizations, people, politics), socialization, work qualities, and social distance. Multiple Classification Analysis is used to control for other factors, the value items are coded as more or less non-conformist, and net deviations from the item mean are calculated. Household types are then compared for number of deviations indicating that a type is more or less associated with non-conformist values. 19.

(31) Part two uses Correspondence Analysis to plot the 80 value items on a plane to show the proximities of each household type and their associated values. The three country groups are then compared for similarities. Though the results show some variation between countries, overall the trends are very similar. The final conclusions show that in all countries analyzed, values and household type are statistically associated. In particular, all countries show that cohabitating without children is linked most strongly with non-conformist values, being married with children and never having cohabitated is linked most strongly with conformist values, and having previously cohabitated results in stronger non-conformist values in those. 政治大. married with children relative to those married with children who have never cohabitated. In most but not all cases, being single is also linked to more non-conformist values and. 立. those who are divorced or separated but not yet with a new partner have more non-. ‧ 國. 學. conformist values than those who are married. Surkyn and Lesthaeghe refer to these household value profiles as “footprints” of the SDT, “found in all the regions studied so far, and hence in the SDT-precursors as well as in the SDT newcomers” (2004, p. 75).. ‧. Lesthaeghe (2010) gives empirical evidence of the SDT in the Asian countries of. Nat. sit. y. Japan, South Korea, and Singapore. It is the only known study to date that attempts a. io. er. statistical analysis linking value changes and demographic traits in Asia. This study uses data from the World Values Survey from 1995 and 2000 (pooled) for Japan, 1994 and. n. al. Ch. i n U. v. 2001 (pooled) for South Korea, and 2002 for Singapore. Lesthaeghe admits to one. engchi. significant limitation of the data: the only demographic information that was obtained is marital status and number of children, excluding information on divorce and cohabitation. As such the study is limited to determining if delayed fertility is related to the same values as in the West. The values included in the analysis were those related to family and gender, socialization traits, work characteristics, political orientation, ethics and morality, and religion and secularization. The study utilizes binary logistic regression, giving exponentiated regression coefficients for each value item. Then the number of coefficients in the correct direction (i.e. conformist values resulting in earlier fertility and non-conformist values resulting in delayed fertility) are totaled for each value category for the three countries. In Japan 83 20.

(32) percent of items, in South Korea 89 percent of items, and in Singapore 86 percent of items “are in the expected direction, indicating an overwhelming concordance with what is found in the West (cf. Surkyn and Lesthaeghe, 2004): non-conformist or more libertarian attitudes correctly predict postponement of parenthood” (2010, p. 241).. 2.3 Existing Research on Causes of Demographic Change in Taiwan There are relatively few studies that have attempted to explain the reasons behind all of the demographic changes described in section 2.1 for Taiwan. Of the research that does, the majority discuss Taiwan in the context of trends in East Asia as a whole. This. 政治大. section summarizes three studies that consider the full range of demographic changes and. 立. Taiwan.. ‧ 國. 學. Atoh, Kandiah & Ivanov (2004) take a fertility-focused analysis of whether Second Demographic Transition changes exist in Japan, Singapore, Hong Kong, South Korea, Thailand, China, and Taiwan. The study includes the following demographic. ‧. features: fertility changes, age at childbirth, changes in marriage, changes in marital. y. Nat. fertility, and the connection between delayed marriage and fertility. Fertility changes are. sit. then framed in the context of the following social and economic contexts: modernization. er. io. and family planning programs, women’s emancipation, and family and gender equality.. al. n. iv n C phenomenon and that policy has h done little to ameliorate e n g c h i U the situation. Additionally, the study notes that, “the family system was, unexpectedly, able to coexist with fertility. The study concludes by confirming that the fertility decline is not an exclusively Western. transition to the replacement level” despite the fact that “some argue that the traditional family system is rather favorable to fertility” (p. 54). The proposed explanation for this is that traditional family values only increase fertility in married couples. At the same time, traditional values “[seem] to be a component that is rather unfavorable to partnership formation in the age of women’s emancipation, since it has such emotional connotation as women’s obedience to men and the predominance of parent-child relationship over conjugal relationship” (p. 54.).. 21.

(33) Ochiai (2011) and Raymo, Park, Xie & Yeung (2015) also raise the question of whether East Asian countries are going through the Second Demographic Transition in the same way as Europe. Both conclude that the demographic features are nearly the same, aside from low levels of cohabitation and little to no childbearing outside of wedlock in Asian examples, but the features are not caused by ideational change as proposed by Lesthaeghe (2010). While each study uses different theoretical reasoning to reach its conclusion, both connect the demographic changes to a combination of rapid modernization and rigid family values. Ochiai (2011) traces the development histories of Japan, Singapore, Korea,. 政治大. Taiwan, and Hong Kong through the lens of “compressed modernity,” defined as rapid pace modernization. Unlike the West which had a long period in which fertility stayed at. 立. replacement level, in these countries the First Demographic Transition was immediately. ‧ 國. 學. followed by the Second Demographic Transition due to the speed of industrialization. During industrialization states need to develop welfare systems and in all countries welfare was initially based on “familialism,” defined as “the idea that the family should. ‧. have the greatest welfare responsibility towards its members, both in income distribution. sit. y. Nat. and care provision” (Esping-Andersen, 2009, as cited in Ochiai, 2011, p. 230). Development occurred too quickly to later transition to state-provided welfare like many. io. n. al. er. countries in Europe did. As a result, familialism is much more strongly entrenched in the. i n U. v. countries studied. While marriage has become deinstitutionalized in Europe:. Ch. engchi. Marriage as an institution of duty and responsibility rather than intimacy is still intact in East Asia. Because of that, when family relationships changed from being social resources to being risks during the economic crisis, risk-aversive individualization occurred to avoid the burden of a family…it was not individualism but familialism that is causing the current demographic and family changes in East Asia (Ochiai, 2011, p. 235). Ochiai concludes that the consequences of familialism show that is has “failed in constructing a sustainable social system…In the near future, when other societies are as. 22.

(34) aged as Japan is today, the social reproduction of East Asia will become an even harder challenge without radical and dynamic policy changes” (p. 235). Raymo, Park, Xie & Yeung (2015) also assert that there has been very little change in the concept of family in East Asia and that “trends in marriage and fertility reflect tension between rapid social and economic changes and limited change in family expectations and obligations” (p. 471). The study analyzes China, Japan, South Korea, and Taiwan from the perspective of various theoretical frameworks to demonstrate how this conflict has resulted in the demographic changes in the region mistakenly attributed to individualism. The underlying belief is that the social and economic conditions for the. 政治大. Second Demographic Transition are similar for both the West and East, but the reasons behind demographic changes differ because of cultural conditions in Asia, including. 立. factors like a strong connection between marriage and fertility, unequal gender roles in. ‧ 國. 學. marriage, expectation of heavy role in childrearing for mothers, expectations for women to marry a man of higher status, and obligations to support other generations within the family or live with extended family.. ‧. In the analysis of possible explanations for demographic change, Raymo, Park,. Nat. sit. y. Xie & Yeung consider the following potential causes: ideational change, women’s. io. er. economic independence and the opportunity costs of marriage and parenthood, economic stagnation and growing inequality, marriage markets, and continuity and change in. n. al. Ch. family organization/living arrangements.. engchi. i n U. v. They reject the idea of ideational change due to lack of evidence because most singles in East Asia still express the desire to get married and most married couples express the desire to have two children. In discussion of women and opportunity costs, the study finds it very likely that delaying or avoiding marriage is a tactic to avoid all of the roles and obligations attached. Intensive mothering is also known to negatively affect women’s careers in Asian job markets because job and family expectations make it incredibly difficult to fulfill both roles, especially in Japan and Korea. The theory of economic stagnation and growing inequality contributing to delayed marriage is supported by the expectation for men to be the main earner of a household. Under poor economies, men with reduced economic prospects are unable to get married or have 23.

(35) children. Marriage markets may also affect the ability to find an appropriate spouse. No mechanism has replaced arranged marriage for finding partners. Trends of higher education for women have also resulted in groups of women with high levels of education and men with low levels of education being unable to find a suitable partner due to the expectation that women must marry “up.” Finally, changes and continuity in living arrangements mean that couples generally begin married life in their own household but live together with elderly parents later in life. This makes starting one’s own household more expensive and may force some couples to postpone marriage. Women are also more likely to be able to keep their jobs if they live in extended households early in marriage. Expectation to care for elderly parents may also make. 政治大. eldest sons have difficulty finding a partner willing to accept that role.. 立. Overall, Raymo, Park, Xie & Yeung conclude that:. ‧ 國. 學. Tensions generated by rapid social and economic changes similar to those observed in the West and limited change in family expectations and. ‧. obligations…may contribute to later and less marriage and childbearing by increasing the opportunity costs of marriage for women (especially for. Nat. sit. y. women with higher levels of education), decreasing the ability of men to. io. er. fulfill the provider role (especially for men with lower levels of education), and exacerbating marriage market mismatches (p. 485).. al. n. iv n C A notable similarity in all three of these studies h e n g c h i U involving Taiwan and the Second Demographic Transition is that they reject the use of a Western model for explaining Asian circumstances. All of the studies confirm the existence of similar demographic changes in East Asia as in the West, but dispute that the cause is the same. Each study separately concludes that Asian family values are the likely reason for delayed marriage, reduced overall marriage, and subreplacement fertility. SDT theory does not reject the possibility of family systems postponing certain features of the transition, such as cohabitation or fertility outside of marriage, but it assumes those features will occur eventually. Also, SDT value shifts include greater tolerance toward alternative lifestyles. In Asia, marriage and childbearing are still tightly linked and cohabitation and sex outside of marriage are still taboo subjects. However, a major limitation of the 24.

(36) conclusions reached by these three studies is that they are purely theoretical and have not been tested with empirical data. To conclude the true cause of the demographic changes, evidence proving a connection is required.. 2.4 Theoretical Framework This research utilizes Second Demographic Transition theory in its analysis of Taiwan’s demographic changes. The research design adapts previous scholars’ SDT approaches to Taiwan’s situation. In order to account for the objections of authors in the previous section to using Western SDT theory to explain behavior in Asia, family values. 政治大 independently of each other. In this way, this study aimed to determine if the Second 立 Demographic Transition is occurring or partially occurring in Taiwan. Whether and other value categories were analyzed for their connection to union formation. ‧ 國. 學. Taiwanese society is undergoing the SDT is of particular importance to the status of women. Value changes creating a more tolerant society with less rigid family values. ‧. allow women greater freedom of choice. Conversely, a partial value change may create the desire for self-realization without loosening family obligations, forcing women to. y. Nat. sit. choose between individual desires and family expectations. The next chapter describes. n. al. er. io. the research approach in detail.. Ch. engchi. 25. i n U. v.