AGE AND ITS RELATION TO CRIME IN TAIWAN AND THE UNITED STATES: INVARIANT, OR DOES CULTURAL CONTEXT MATTER?*

• DOI: http://doi.org/10.1111/1745-9125.12139
• Author: YUNMEI LU,DARRELL STEFFENSMEIER,HUA ZHONG
• Published date: 01 May 2017
• Date: 01 May 2017

AGE AND ITS RELATION TO CRIME IN TAIWAN AND

THE UNITED STATES: INVARIANT, OR DOES

CULTURAL CONTEXT MATTER?∗

DARRELL STEFFENSMEIER,1HUA ZHONG,2and YUNMEI LU1

1The Pennsylvania State University

2The Chinese University of Hong Kong

KEYWORDS: age–crime relations, life course/developmental, international, theory,

adolescent risk taking

Current empirical and theoretical understanding of the relation between age and

crime is based almost entirely on data from the United States and a few prototypi-

cal Western societies for which age-specific crime information across offense types is

available. By using Western databases, Hirschi and Gottfredson (1983) projected that

the age distribution of crime is always and everywhere robustly right-skewed (i.e.,

sharp adolescent peak)—a thesis that is both contested and widely accepted in crimi-

nology and social science writings. In the study described here, we tested this age–crime

invariance thesis by comparing age–crime patterns in Taiwan (a non-Western Chinese

society) with those in the United States. In light of Taiwan’s collectivist culture versus

the U.S. individualist gestalt, we anticipated more divergence than homogeneity in their

age–crime schedules. Our findings show robust divergence in Taiwan’s age–crime pat-

terns compared with U.S. patterns and the reverted J-shaped norm projected by Hirschi

and Gottfredson. Implications for research and theory on the age–crime relation and

for studying human development or life-course topics more broadly are discussed.

In this research, age–crime data from Taiwan and the United States were used to in-

vestigate the influential thesis of Hirschi and Gottfredson (1983: 554) that an adolescent

peaked age–crime relationship is invariant to social and cultural conditions. “For all of-

fenses, at all times, and in all places and for all races and both sexes,” they wrote, “in-

volvement in crime reaches its peak in the middle to late teens and then declines rapidly

throughout life” (Gottfredson and Hirschi, 1990: 219). This “brute fact of criminology,”

they argued further (Hirschi and Gottfredson 1983: 552), is manifest regardless of data

[Correction added on April 27, 2017 after online publication: the text “(N=1,695)” and “(N=

8,076)” were deleted in the third sentence of the abstract.]

∗Additional supporting information can be found in the listing for this article in the Wiley Online

Library at http://onlinelibrary.wiley.com/doi/10.1111/crim.2017.55.issue-2/issuetoc.

We thank Steven Cuvelier, Miles Harer, Charles Hou, Susyan Jou, Yung-Lien Lai, Ping Li,

Li-Chen Lin, Cheng-Hsien Lin, Daniel McCarthy, Chuen-Jim Sheu, Ivan Sun, Jason Thomas, Ar-

land Thornton, and the Criminology editor and anonymous reviewers for comments and advice.

We also thank the many helpful scholars and law enforcement officials who graciously agreed to

be interviewed for this age–crime project.

Direct correspondence to Darrell Steffensmeier, The Pennsylvania State University, 1016 Oswald

Tower, University Park, PA 16802 (e-mail: d4s@psu.edu).

C2017 American Society of Criminology doi: 10.1111/1745-9125.12139

CRIMINOLOGY Volume 55 Number 2 377–404 2017 377

378 STEFFENSMEIER, ZHONG, & LU

source and exists for both aggregate and individual levels. It cannot be explained by

extant sociological variables or theories. Instead, they argued that age has a direct effect

on crime, presumably through biological maturation and natural aging of the organism.

Many scholars today agree with the Hirschi and Gottfredson (HG hereafter; 1983) the-

sis of a ubiquitous age–crime schedule and consider it to be “uncontested in the crimi-

nological literature” (Kanazawa and Still, 2000: 434; Steinberg, 2008; Walsh, 2009). But

other analysts consider the matter to be unsettled because the claim is based largely on

age–crime statistics for a few Western nations (e.g., Australia, England, and Sweden)

and particularly for the United States, which are among the few nations to collect and

publish age–crime statistics (arrests) that cover several crime types over a long period

of time (Greenberg, 1985; Pridemore, 2003; Steffensmeier and Allan, 2000). U.S. age–

crime statistics are notable for demonstrating skewed age curves in which involvement in

crime peaks early (middle to late teens) and then declines rapidly or slowly over subse-

quent years. In contrast, little if anything is known about the age–crime distribution in

non-Western countries, most notably in Chinese societies like Taiwan, with collectivist

cultural heritages (extending at least as far back to Confucius circa 500 BC) differing con-

siderably from Western societies with individualist cultural heritages (Greco-European

influence extending back to Greek philosophers such as Socrates circa 400 BC).

Despite whether it is correct, the age–crime invariance thesis has been alternately influ-

ential and contentious. In addition to its influence on criminologists’ thinking about the

effects of age on crime, the invariance thesis is central to an enormous body of research

and theory on adolescent risky behaviors and the dynamics of the adolescent–adulthood

transition that now crisscross several areas of interest in the social and behavioral sci-

ences, including life-course sociology, evolutionary and developmental psychology, and

most recently, the nascent field of neurobiology. With their emphasis on establishing uni-

versal forms caused by internal states, the notion of a ubiquitous adolescent-driven age–

crime relation has gained traction in particular among scholars favoring evolutionary or

biological interpretations of human development (Kanazawa and Still, 2000; Steinberg,

2008; Walsh, 2009). An invariant teen-driven age–crime schedule may raise doubts about

the adequacy of numerous social theories that not only imply an ability to account for

age variations but also typically perceive social context as playing a major or even the pri-

mary role, and make it easier to argue that high levels of adolescent crime are a natural

or normal part of the human developmental process.

Surely, contended prominent sociobiologist Anthony Walsh (2009), if social influences

shorn of evolved developmental patterns were what really mattered, “we would have ac-

counts of some cultures and some times when the age-crime curve did not exist, or was

even reversed, but we do not” (p. 156). Rather than social influences, a better explanation

for the ostensibly much higher teen rates of crime is to be sought in underlying biological

causes (e.g., hormonal and neurological changes) that accompany adolescence. “Age is

an index of a series of developmental stages from prenatal to senescence that we all go

through if we live long enough with each having its own characteristic physiology. There

must be something special requiring its own explanations going on during adolescence

that dramatically, albeit temporarily, greatly increases the probability of antisocial behav-

ior” (Walsh, 2009: 156, emphasis added).

An international perspective is especially useful here because there is no reason to

believe that innate biological factors involved in determining age effects will vary across

countries. The degree to which age differences in crime result from biological or fixed

developmental factors implies that there should be no international variation in these