EXPLAINING UNINTENDED AND UNEXPECTED CONSEQUENCES OF POLICY DECISIONS: COMPARING THREE BRITISH GOVERNMENTS, 1959–74
| Date | 01 September 2014 |
| DOI | http://doi.org/10.1111/padm.12081 |
| Author | PERRI 6 |
| Published date | 01 September 2014 |
doi: 10.1111/padm.12081
EXPLAINING UNINTENDED AND UNEXPECTED
CONSEQUENCES OF POLICY DECISIONS: COMPARING
THREE BRITISH GOVERNMENTS, 1959–74
PERRI 6
The vulnerability of policymaking to unintended and unanticipated consequences has been doc-
umented since Thucydides. Yet we still lack integrated conceptual and explanatory accounts of
their variety and aetiology. Adequate consideration of putatively unintended and unanticipated
consequences requires evidence about policymakers’ prior intentions and anticipations, the factors
affecting their cognition, and the forces bearing upon responses to attempted execution of policies.
This study uses archival evidence about three post-war British governments to examine hypotheses
derived from neo-Durkheimian institutional theory. It compares relationships between policymak-
ers’ informal social organization and their biases in framing anticipations and intentions in three
policy fields. It shows that, contrary to widely made claims about a ‘law’ of unintended conse-
quences, neither unintended nor unexpected consequences are random, but reflect basic patterns in
variation and aetiology which the neo-Durkheimian theory explains well.
INTRODUCTION
Policy researchers commonly argue that better understanding of unanticipated and
unintended consequences can help policymakers to minimize their incidence or severity,
usually by improving their prior anticipation (e.g. King 1995; Grabosky 1996; Streets and
Glantz 2000). By contrast, many political scientists (e.g. Rhodes 2000, 2005, 2011; Hood and
Peters 2004) and complexity theorists (Geyer and Rihani 2010) believe that unintended
and unanticipated consequences are unavoidable, suggesting that neither their incidence
nor severity can be reduced. Some write of a ‘law’ of unintended consequences, suggesting
that their variation is indefinite and without pattern and their occurrence random and
inevitable. Yet these claims are rarely examined empirically.
Studies often classify unhappy consequences as unintended or unanticipated without
examining evidence about what policymakers did anticipate or intend (6 2010): absence of
anticipation or intention is often inferred from the fact that a policy’s outcomes are unwel-
come (6 2010; Newberry 2002 and Rao 2002 are exceptions). Most studies concern fiascos
and failures, even though some unintended or unanticipated outcomes and surprises can
be welcome to policymakers (Cortell and Peterson 2001 is an exception). Furthermore,
most studies in this vast literature study single cases. Some studies use comparative
analysis to demonstrate the importance of a particular causal factor (e.g. Scott’s 1998
critique of hubristic governmental programmes of technological modernization). Induc-
tive comparison can provide valuable insights about different dynamics for under- and
over-shooting intentions (e.g. Anderson 2004). Margetts et al. (2010) use case-comparative
analysis to develop a theory of causation for unintended consequences but only for cases
of ‘modernization’.
We therefore need better understandings of variety and aetiology in unintended and
unanticipated consequences of policy decisions. Is variety limited? Are there patterns in
Perri 6 is Professor in Public Management, School of Business and Management, Queen Mary, University of London,
London, UK.
Public Administration Vol. 92, No. 3, 2014 (673–691)
©2014 John Wiley & Sons Ltd.
674 PERRI 6
aetiology? What contribution do policymakers’ biases in anticipation and intention make
to causation?
This article shows how neo-Durkheimian institutional theory can answer these ques-
tions. Using archival sources, the study compares cases from several policy fields in
British governments between 1959 and 1974. It shows that, by contrast with one version
of a ‘law’ of unintended and unanticipated consequences, hypotheses derived from this
theory perform well in explaining how intentions and anticipations might be skewed and
policy fields structured to cause these outcomes.
This is a well-documented period. The aim, therefore, is not to use theory to retrodict
previously unknown facts and then reveal previously unknown historical evidence to test
those retrodictions. Rather, if the theory’s retrodictions are supported, its contribution
will be in providing an integrated explanation for the variation in, and patterns among,
particular empirical factors previously emphasized in historiographical studies. Historical
work has identified particular empirical factors to explain unintended consequences, but
scientific progress consists not only in solving empirical problems but also in resolving
conceptual ones (Laudan 1977; Johnson 2003). For the first time, this article offers an
integrated, comprehensive, elegant, parsimonious explanatory theory within which these
particular empirical facts play clear roles and in which their variation can explained.
UNINTENDED CONSEQUENCES AND SURPRISES IN GOVERNMENT
This article uses Hirschman’s (1991) distinction among unintended consequences between
welcome surprises, cases of futility (null effect), of perversity (undermining the goal
policymakers intend to achieve), and of jeopardy (undermining other things policymakers
care about).
There are different understandings of what a ‘law’ of unintended consequences would
claim, but none is entirely clear. The least controversial, least precise, and least interesting
is the simple generalization that they are common (e.g. Fine 2006, p. 7). The most extreme
is Roots’ (2004) claim that statute lawmaking is more vulnerable to unintended, perverse
consequences than any other form of human action.
Some writers interpret such a law as claiming that unintended effects of policies
typically outweigh intended ones, or that the principal outcomes will be unintended (e.g.
Rhodes and Bevir 2003, p. 76). If a policy has any consequences at all, then in the long run,
most of them must be unintended, because people cannot frame intentions far into the
future. Such a ‘law’ is therefore only interesting if it claims that unintended effects will
dominate in the short run. Borrowing Hennessy’s (1992, p. 453) phrase, Rhodes (2000, p.
353) describes this short run claim as one of the ‘sour laws of unintended consequences’,
implying that these consequences arise randomly and without pattern. Yet even this is a
descriptive empirical generalization rather than a law, because it provides no explanatory
mechanisms.
Lacking explanatory theory, studies that seek generalization tend to use descriptive
typologies. This article argues that the widely cited taxonomic frameworks are of greatest
value when combined with an approach capable of demonstrating causation. Classification
of policies and outcomes must address both forces biasing decision-makers’ intentions
and anticipation, and forces affecting policy execution. Intention and anticipation can
vary independently (6 2010). Unintended consequences may be anticipated. Sometimes
unwelcome outcomes are anticipated as possibilities but not expected. These are ‘risks
knowingly run’. Conversely, when a hoped-for outcome comes as a welcome surprise
Public Administration Vol. 92, No. 3, 2014 (673–691)
©2014 John Wiley & Sons Ltd.
To continue reading
Request your trial