Evidence and liberty
| Author | Lawrence W. Sherman |
| DOI | 10.1177/1748895808099178 |
| Published date | 01 February 2009 |
| Date | 01 February 2009 |
| Subject Matter | Articles |
5
Evidence and liberty:
The promise of experimental criminology
LAWRENCE W. SHERMAN1
University of Cambridge, UK
Abstract
Evidence and liberty are two great ideas in British history. One conse-
quence of both ideas is experimental criminology, which applies
research designs developed in Britain to matters of liberty affecting
the entire world. The promise of experimental criminology is to gen-
erate better evidence about how to increase liberty. The crucial chal-
lenge to experimental criminology is the means by which research
results, even when accepted as true, may be translated into wide-
spread practice. Two models for applying experimental research are
possible. One is ‘bottom–up discretion’, in which crime victims,
police, judges and probation officers and others take experimental
results into account when making their decisions one case at a time.
Another model is ‘top–down guidance’, in which either Ministers or
an independent evidence-assessment agency like ‘NICE’—the
National Institute of Health and Clinical Excellence—appoint commit-
tees of practitioners and researchers to develop succinct operational
guidance based upon systematic reviews of experimental evidence.
Bottom–up discretion will grow steadily from the sheer persistence of
criminologists and practitioners doing experiments. Liberty could be
advanced much faster by top–down guidance, however, if govern-
ment chose to invest in ‘prospective meta-analysis’ a method of multi-
site randomized trials that would assess generalizability of policy
effects as well as their magnitude and cost-effectiveness.
Key Words
cost-effectiveness • evidence • experimental criminology • liberty
• policy • prospective meta-analysis • slavery
Criminology & Criminal Justice
© The Author(s), 2009. Reprints and Permissions:
http://www.sagepub.co.uk/JournalsPermissions.nav
www.sagepublications.com
ISSN 1748–8958; Vol: 9(1): 5–28
DOI: 10.1177/1748895808099178
DEBATE AND DIALOGUE
Liberty in pursuit of evidence
Evidence and liberty are two great ideas in British history. They are also closely
connected. From the Magna Carta (Lackland, 1215) to John Stuart Mill
(1859), from Isaiah Berlin (1969) to the Lords’ rejection of 42 days’ detention
of suspected terrorists without evidence to support a charge (BBC News,
2008), Britons have required evidence of harm as the basis for deprivations of
liberty. From the making of laws to the enforcing of laws, and even to morally
defiant breaking of laws, the substantive commitment to liberty has been cul-
turally supported by a procedural commitment to evidence. These commit-
ments inevitably include arguments over the very meaning of the words, let
alone how they are related. Yet the core consensus over their importance cuts
right across British society, from families to medicine to finance to justice.
From the standpoint of criminology, the most important meaning of
evidence is scientific rather than forensic. That is, evidence is not central to
criminology (as in criminal law) in the sense of determining whether a
particular suspect is guilty. Evidence is central to criminology in determin-
ing whether a hypothesis is false. Criminology has long been rich in theory,
but poor in evidence. And without evidence, liberty has suffered at the
hands of theory. From the dogmatic jurisprudence that claims punishment
will prevent crime to the Information Commissioner’s recent order that old
criminal records should be destroyed (Information Tribunal, 2008), both
governments and criminals may reduce or increase our liberty on the basis
of theory, not evidence. Where those theories are normative statements of
value, this is perhaps as it should be. But where even normative judgements
are made upon untested factual assumptions about cause and effect, liberty
remains vulnerable without evidence.
In criminology, as in medicine, assertions of cause and effect are often
made on the basis of experience, hunches, peer opinions and occupational
cultures (Chalmers, 2003). Only in very recent years has medicine begun a
shift to more radically evidence-based practice that is far from completed
(Millenson, 1997). In doing so, it has relied upon research designs devel-
oped over the past century in Britain, especially the randomized controlled
trial (RCT; see, for example, Fisher, 1926, 1935; Doll, 1992). This reliance
is now so great that it literally determines matters of life or death in the pro-
vision of specific medical treatments by the National Health Service on the
basis of experimental evidence of their effects. Not everyone is happy with
this development, to be sure, but it may still be described modestly as the
most advanced and comprehensive application of hypothesis-testing evi-
dence to government policy in the history of the world.
The example of medicine looms large in experimental criminology, a
field whose rapid growth (Farrington and Welsh, 2005) has been fuelled by
growing demand for more reliable evidence of cause and effect. While
experiments in criminology date to the 18th century and randomized trials
to the mid-20th (Sherman, 2005), a learned society in the field was only
organized in 19982and its journal in 2005.3In this article written to mark
Criminology & Criminal Justice 9(1)6
To continue reading
Request your trial