Biology Reference
In-Depth Information
conceptualizing extrapolation that combines elements from Steel ( 2008 ) and Pearl
and Barienboim ( 2011 ) and explain its relevance to the Donohue and Levitt study.
Section 4 explains how macro-level data can be used to support an extrapolation
made at the level of mechanisms and links this to the final of the three philosophical
themes listed above.
1 Donohue and Levitt's Argument
In this section, I examine the structure of Donohue and Levitt's argument for their
hypothesis that the legalization of abortion in 1973 is the primary cause of the
abrupt, nationwide, and persistent drop in crime rates that began in the early 1990s
in the United States. Their case can be usefully organized into three main
intertwined sub-arguments. First, they argue that the decline in crime rates is not
adequately explained by alternative causes, such as increased incarceration or the
decline of the crack epidemic. Secondly, tracing a mechanism through which
legalized abortion in 1973 could result in a drop in the crime rate 18-24 years
later when the first post-legalization birth cohort entered its peak crime age. Finally,
a statistical argument focused on differing timing of declines in crime rates between
earlier and later legalizing states and on the timing of which types of crime dropped
first.
Nationwide statistics of rates of violent and property crimes in the USA peaked
in the first half of the 1990s and then began a steady decline that has continued into
the first decade of the new century (Truman and Rand 2010 ). The abrupt, wide-
spread, and persistent nature of this decline constrains possible explanations. Any
explanation must be able to explain why the decline began when it did, why it
occurred throughout the USA, and why it has persisted, now, for over 15 years.
Donohue and Levitt's hypothesis directly explains each of these features. First, the
impact of abortion legalization on crime would begin to be felt approximately
18 years after the Roe v. Wade decision in 1973—that is, in 1991—when the first
post-legalization cohort entered its peak crime years. Secondly, Roe v. Wade
invalidated legal bans on abortion in all jurisdictions in the USA and hence is a
cause of nationwide scope. Finally, abortion legalization is a cause that exerts its
impact over an extended period of time. The effect of legalization would continue
to be felt as prelegalization generations age out and post-legalization generations
age in. Thus, Donohue and Levitt predict that the crime-reducing impact of
legalized abortion would continue to be felt until around 2020 ( 2001 , p. 415).
Donohue and Levitt argue that a number of alternative explanations of the
decline fail to account for one or more of the three basic features listed above.
For example, consider one of the most popularly cited explanations, innovative
policing strategies. These were instituted in New York City only after the decline in
crime had already begun there and were not implemented in many other cities, such
as Los Angeles, that also experienced significant crime reductions (Levitt 2004 , pp.
172-173). Another explanation points to the crack cocaine epidemic, which struck
Search WWH ::




Custom Search