Socio-demographic Change and Age-Crime Patterns in Taiwan and the United States, 1976-2015
Restricted (Penn State Only)
- Author:
- Lu, Yunmei
- Graduate Program:
- Sociology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 28, 2018
- Committee Members:
- Darrell J Steffensmeier, Dissertation Advisor/Co-Advisor
Darrell J Steffensmeier, Committee Chair/Co-Chair
Jeremy Staff, Committee Member
Liying Luo, Committee Member
Guangqing Chi, Outside Member - Keywords:
- Age-crime Relations
Cross-cultural Comparison
Age-period-cohort Analysis
Social Change
Life Course
Asian Criminology - Abstract:
- Age and its relationship with crime has been discussed in sociology and criminology for two hundred years. In 1983, Hirschi and Gottfredson (HG hereafter) proposed one of the most influential and controversial projections about the age-crime relationship. They argued: (1) crime is always and everywhere heavily concentrated among teenagers and young adults; (2) crime always declines after reaching the adolescent peak. They further asserted that the reverted J-shaped age-crime curve is universal across social contexts. In criminological writings over the past few decades, HG’s invariance thesis is often viewed as “uncontested.”. One of the most prominent challenges to the age-crime invariance thesis is the recent publication in Criminology (Steffensmeier, Zhong, and Lu 2017) that revealed substantial divergence in the age-crime patterns of Taiwan as compared to the U.S. from 2008 to 2010. However, the study only covers three years of data and does not take into account the impact of period and cohort on the age-crime patterns. Another study conducted by Zhong (2005) also revealed some variation in the age-crime patterns of Taiwan during the period of 1970s to 1990s, but it also failed to tease out age effects from both cohort and period influences. Besides these two studies, most of the age-crime analyses are based on crime data from the U.S. and a few prototypical western countries. Notably, very little is known about the age-crime relationship in non-western societies, especially the age-crime patterns of these societies across various historical periods. Therefore, the first objective of the current project is to build on the research conducted by Steffensmeier and colleagues in 2017 and provide a systematic assessment to the age-specific crime data from both Taiwan and the U.S. over the past four decades. Through comparing the age-crime patterns between the two societies, across different historical periods, and across offense types, the current project provides a multi-faceted evaluation of the HG invariance thesis. The age distribution of crime observed in a particular period is not only shaped by age effects but also by social change and cohort-specific experiences. Prior sociological perspectives (e.g. modernization theory) suggest that social transformations, such as economic development, may lead to disproportionate increases in youth crime and shifts toward more adolescent-peaked age-crime distributions as a society becomes more industrialized. Theoretical perspectives in demography also highlight the importance of cohort replacement in understanding the impact of social change on age-specific outcomes across time. Specifically, Easterlin’s (1987) cohort-size hypothesis proposed that large cohorts are more likely to be economically disadvantaged and have higher propensity to engage in crime than other cohorts. Existing sociological research on age and crime based on cross-sectional data that did not differentiate the age effect on crime from the influence of social change and cohort experience. Thus, the second objective of the project is to investigate the impact of cohort-specific experiences resulted from social changes (e.g. economic development, westernization and demographic transformation) on the age-specific crime rates in Taiwan as compared to the United States. To meet these objectives, my dissertation draws on mainstream theoretical and methodological frameworks in sociology, criminology and demography. Theoretically, the study emerges from sociological theorizing about age effects on crime, the relationship between social change and crime, the demographic conceptualization of cohort replacement and cohort size effects on crime. Empirically, I address the above objectives using census-based population age-specific data and official age-specific crime statistics of Taiwan and U.S. from 1976 to 2015. Crime data from each society covers seven different offense types, including three summary categories of total, property, and violent offenses. The first part of the analysis tests the invariance thesis through comparing various measures of the age-crime curves developed by criminologists. The second part of the analysis applies age-period-cohort models and age-period-cohort-size models developed by demographers and criminologists, examining the impact of social change and cohort experience on the age-crime curve. Two important sets of findings emerge from my analysis. First, I found that the age-crime distributions in Taiwan demonstrate considerable divergence across time and across different offense types, which is in contrast to the robust concentration of crime among young people observed in the U.S. across time. These findings suggest that strong claims about the invariance of the age-crime relationship should be abated and cast doubt on the biological approaches that attribute teenage crime to neurobiological changes. The age-period-cohort analysis also reveals two important findings. First, age-crime curves in both countries are shaped by period and cohort effects, but period and cohort effects are more prominent in predicting Taiwan’s age-period-specific arrest curves than those of the U.S. Moreover, young cohorts who were born in the 1980s and 1990s are less crime prone than the older cohorts, at least during the teenager and young adult periods. I conclude that cohort replacement, which is often ignored in the age-crime literature, plays a role in shaping the age-crime curves, especially in Taiwan and for property offenses. Second, relative cohort size, which is considered a mechanism of cohort effect, has significant positive effects on age-specific crime rates in Taiwan and the effect is greater and more robust for property crime than violent crime. These findings are consistent with the U.S. findings of the past two decades and provide new evidence to support Easterlin’s cohort size thesis. The current project offers three contributions to criminology and allied social science fields. First, the project contributes to the longstanding debate on the age-crime relationship by bringing in new evidence from a non-Western country across multiple decades and with multiple analytical techniques. Second, since age is an important variable and status marker in sociology, criminology, and demography, the current study helps researchers further understand the puzzling effects of age on social outcomes. If we are interested in the universality of age effects on certain outcomes, we may need to adopt the comparative approach used in the current study to test the universal claims. Third, the current project underscores the importance of the cross-fertilization benefits of applying the methods and theories of one substantive area to another substantive area: in this case, applying the methods and theories of demography to criminology. Through incorporating the concept of cohort and the age-period-cohort modeling developed by demographers, the current project provides a new assessment of the age effects on crime by taking into account the impacts of social changes and cohort experiences. This kind of assessment is rarely seen in the age-crime analysis of a non-Western country.