CONTROLLING RISK INSIDE MODERN GOVERNMENT: DEVELOPING INTERVAL MEASURES OF THE GRID‐GROUP DIMENSIONS FOR ASSESSING SUICIDE RISK CONTROL SYSTEMS IN THE ENGLISH AND JAPANESE PRISON SERVICES
| Author | AYAKO NAKAMURA |
| Date | 01 December 2016 |
| Published date | 01 December 2016 |
| DOI | http://doi.org/10.1111/padm.12265 |
doi: 10.1111/padm.12265
CONTROLLING RISK INSIDE MODERN GOVERNMENT:
DEVELOPING INTERVAL MEASURES OF THE
GRID-GROUP DIMENSIONS FOR ASSESSING SUICIDE
RISK CONTROL SYSTEMS IN THE ENGLISH AND
JAPANESE PRISON SERVICES
AYAKO NAKAMURA
Mechanisms for controlling government organizations have attracted major interest from public
administration researchers. Cultural theory has been used as a tool to identify the core control
approaches of individual organizations. Whereas major existing studies have applied the theory
based on the cultural types using nominal-level measures, this article builds a novel set of interval
measures, focusing on two fundamental factors of government control proposed by cultural theory:
grid-group dimensions. The measures are applied to assess suicide risk control systems in the
English (HMPS) and Japanese prison services (JPS). The results highlight the fatalistic approach in
HMPS and egalitarian approach in JPS, as well as demonstrating the structural characteristics of
each risk control system beyond nominal lists of control tools.
INTRODUCTION
The internal mechanisms of government control have been a major area of government
studies over the last few decades (Hood 1995, 1998; Hood et al. 1999; James 2003; Peters
and Pierre 2003; James and Nakamura 2015). Although several frameworks have been
used to assess government control approaches, this article sheds new light on this area
by utilizing the degrees of group integration and division of labour proposed by cultural
theory (hereafter CT). CT, also known as the grid-group theory or cultural theory of risk,
was originally a typology of culture associated with the social dimensions observed in
social institutions. It was developed by Mary Douglas (Douglas 1970, 1992), and has since
been enlarged by many other scholars (Douglas and Wildavsky1983; Thompson et al. 1990;
Coyle and Ellis 1994; 6 2014).
In government studies, CT provides a coherent comparative framework which has been
used to shed light on the various forms of bureaucratic control associated with the rise of
New Public Management (NPM) (Hood 1995, 1998; Hood et al. 2004; Lodge et al. 2010; 6
2014). Hood et al. (2004), in their book Controlling Modern Government, applied CT to the
governance of public services in three policy domains across eight OECD countries. The
comparative framework of CT provided a jointly exhaustive and mutually exclusive set
of categories of control types. This framework enabled them to demonstrate the consid-
erable variety in ways of controlling bureaucracies, despite supposedly global pressures
for convergence, and enabled them to distinguish trends towards more use of particular
control types across different countries and public service domains.
The dominant approach of previous empirical applications of CT to control of bureau-
cracies, as exemplied by Hood et al.’s (2004) study,has been to apply CT using the cultural
types mainly at the nominal level of measurement. This article develops the empirical
application of CT through attention to the social dimensions by applying measures at
Ayako Nakamura is at the Department of Political Science, Musashino University, Japan.
Public Administration Vol.94, No. 4, 2016 (1077–1093)
© 2016 John Wiley & Sons Ltd.
1078 AYAKO NAKAMURA
the interval level which allow quantication of the degree of difference between objects
measured on this scale in contrast to nominal measures which only identify categories
of control. The fundamental theoretical advantage of measuring CT at this level in this
eld of study is that it captures the idea of social dimensions in response to the two most
fundamental factors of control over bureaucracies: control by degrees of use of rules and
regulation (i.e. grid), and group integration between government ofcials (i.e. group).
This article aims to develop a set of interval measures to assess risk control systems
inside government institutions based on the grid-group dimensions, and thus to also iden-
tify mechanisms of control over bureaucracies. Tomeet those ends, it extends the analysis
of Hood et al. (2004) and work that more generally uses categorical measures by develop-
ing interval-level measures of cultural types in government control studies. In so doing,
the article shows how the proposed measures based on the grid-group dimensions can be
applied to empirical case studies to help reveal important differences; specically in this
article, suicide control systems in the very different prison systems of England and Japan.
The rst section below provides an overview of how CT has been used to analyse gov-
ernment control mechanisms, placing special emphasis on the operational method used in
Hood et al. (2004). The second section explains the methodological strategies developed by
the present author, and includes a discussion of case-specic strategies for suicide control
systems in HMPS and JPS. The third section applies the proposed method to assess those
suicide control systems and, nally, the last section analyses the results, linking them with
existing discussions in CT,in particular the work of Hood et al. (2004) and showing the con-
tribution to understanding control systems provided by the interval-level measurement
approach.
CULTURAL THEORY AND CONTROLLING MODERN GOVERNMENT
CT was originally proposed as a means of understanding different attitudes towards risk
embedded in social organizations (Douglas and Wildavsky 1983; Thompson et al. 1990).
Whereas classic risk assessment studies focused on assessing purely the objective like-
lihood of undesirable events and the nancial losses and gains represented therein, CT
focuses on institutional attitudes toward specic risks in the context of individual social
systems (Douglas and Wildavsky 1983; Gross and Rayner 1985; Dake 1991; Rayner 1992;
6 2014). The core theoretical innovation of CT is its identifying of two social dimensions:
that is, grid and group. The grid dimension denotes rules circumscribing an individual’s
life in each social system (Douglas and Wildavsky 1983; Gross and Rayner 1985; Dake
1991; Rayner 1992; Hood 1998; 6 2014); and the group dimension denotes group pressures
shared exclusively between members within a particular institution, while strictly distin-
guishing them from non-members outside of the institution who are not affected by the
same pressures (Douglas and Wildavsky 1983; Gross and Rayner 1985; Thompson et al.
1990; Dake 1991; Rayner 1992; Hood 1998; 6 2014). CT assumes that institutional attitudes
to risk are determined by the interaction between these social dimensions (Douglas and
Wildavsky 1983; Gross and Rayner 1985; Thompson et al. 1990; Dake 1991; Rayner 1992;
Hood 1998; 6 2014), the two axes of which cross-tabulate to form four quadrants: hierar-
chist, individualist, egalitarian and fatalist (see gure 1).
Government studies shed light on CT as an organizational control theory (Hood 1995,
1998; Bellamy et al. 2008; Löfstedt and 6 2009; Lodge et al. 2010; Swedlow 2011; 6 2014;
Wouters and Maesschalck 2014). In particular, the theory is used to capture the vari-
ous forms of control over bureaucracies that dramatically changed following so-called
Public Administration Vol.94, No. 4, 2016 (1077–1093)
© 2016 John Wiley& Sons Ltd.
To continue reading
Request your trial