MEDIA ATTENTION AND THE DEMISE OF AGENCY INDEPENDENCE: EVIDENCE FROM A MASS ADMINISTRATIVE REORGANIZATION IN BRITAIN
| Author | ANTHONY M. BERTELLI,HARAM LEE,J. ANDREW SINCLAIR |
| DOI | http://doi.org/10.1111/padm.12190 |
| Date | 01 December 2015 |
| Published date | 01 December 2015 |
doi : 10. 1111/p adm .12190
MEDIA ATTENTION AND THE DEMISE OF AGENCY
INDEPENDENCE: EVIDENCE FROM A MASS
ADMINISTRATIVE REORGANIZATION IN BRITAIN
ANTHONY M. BERTELLI, J. ANDREW SINCLAIR AND HARAM LEE
When administrative agencies are terminated, do they quietly fade from public view? On the one
hand, the terminated agencies may have weak issue networks and agency reputations, allowing
them to lose public salience. On the other, strong issue networks and agency reputations may
mean that termination increases attention to the agencies, making the government pay the cost
of public attention generated by the actors within the issue networks. We assess these competing
claims by using a unique dataset from a recent mass reorganization of independent agencies in
Britain as well as data capturing media attention to agencies in major national newspapers. Using
a difference-in-differences approach, we nd that the agencies subject to termination experience
reductions in the amount of media coverage in major newspapers, disappearing from public view
during the post-decision reform period.
Stimulating an important literature in administrative and executive politics, Kaufman
(1976) asked, ‘are government organizations immortal?’ The answer for US agencies
(Lewis 2002, 2004; Carpenter and Lewis 2004) and the programmes administered by them
(Berry et al. 2010), as well as for agencies in other countries (e.g. Park 2013; James et al.
2015), has been that they are not. While some scholars view agency termination as the
‘ultimate act of political control’ (Carpenter and Lewis 2004, p. 202), we argue that it may
be only the penultimate act of a political drama. Do agencies slated for termination see a
drop-off in media coverage, or do their public reputations generate media salience even
after they are eliminated? We set our analysis in a recent mass reorganization of British
independent agencies. During these reforms, the government made deliberate choices
about which agencies to retain as they are or to structurally alter or eliminate altogether;
these choices have consequences for the government’s democratic accountability in a
wide variety of policy-making domains (Bertelli and Sinclair 2015).
Regardless of the specic details of the government’s choices, these decisions did rescind
some aspect of the independent agencies’ prior relationship to the broader administrative
state. An agency’s reputation, though, can survive the termination decision, keeping the
public perception of its function and purpose, as well as the objectives of its issue networks,
alive as a potentially salient concern in the subsequent general election. We focus on the
immediate and relevant political aftermath of the termination decision as agencies expire;
termination, while dramatic (as noted in Carpenter and Lewis 2004), is not immediate. A
doomed agency, even if nominally open for some period of time or continuing in some
other reformed fashion, may no longer provide the same benets to interest groups, the
government and its own employees. While it is clear that over substantially longer peri-
ods of time the reputation would eventually decay, the question at hand here is whether
that decay happens before the next election cycle even begins in earnest. One could easily
Anthony M. Bertelli is at the Robert F. Wagner Graduate School of Public Service, New York University, USA and the
School of Government andS ociety,University of Birmingham, UK. J. Andrew Sinclair is at the Robert F.Wagner Graduate
School of Public Service, New York University, USA. Haram Lee is at the Price School of Public Policy, University of
Southern California, USA.
Public Administration Vol.93, No. 4, 2015 (1168–1183)
© 2015 John Wiley & Sons Ltd.
MEDIA ATTENTION AND THE DEMISE OF AGENCY INDEPENDENCE 1169
imagine agencies ‘going out with a bang’, rallying interest groups and allies to plague the
government post-termination.1
The literature on independent agencies often focuses on the government’s choice,
treating elected politicians as principals and administrative agencies as their agents.
Studies of agency termination (Lewis 2002, 2004; Carpenter and Lewis 2004; Park 2013)
often employ duration models, an approach that treats termination as a ‘cleanly observed’
binary outcome (Carpenter and Lewis 2004), and test hypotheses about when agencies
face the greatest risk of termination. Other recent scholarship challenges the binary
treatment of termination from an accountability perspective and focuses on the role of
previous media coverage on the decision of the government (Bertelli and Sinclair 2015).
A number of studies of independent agencies present notions of the value agencies pro-
duce for the politicians who create them – expertise (Gailmard and Patty 2013), neutral
competence (Huber 2007) and reputation (Carpenter 2010; Maor 2010, 2011). Reputation
is ‘a set of symbolic beliefs about the unique or separable capacities, roles, and obliga-
tions of an organization, where these beliefs are embedded in audience networks’ (for
Carpenter 2010, p. 45).
Building on this reputational perspective, we argue that agencies with strong networks
can increase their media prole, providing greater public information about their reputa-
tion to the voting public that can be used in retrospective evaluations of the government.
An issue network represents any combination of outside lobbying organizations, private
companies, citizens’ groups, political organizations and other agencies within the govern-
ment who benet in some way from agency action.2If an agency has strong network and
reputation, eliminating the agency may increase public attention to it. Such reputational
stickiness may make the termination of an agency less clearly observable than the liter-
ature suggests. Alternatively, if networks are weak and media attention decreases with
termination decisions, agency reputations do not survive and the government is better
able to use termination as an act of political control. Our argument and measures concern
media salience rather than sentiment, as different segments of the electorate will interpret
diverse pieces of information according to their own predilections, but cannot react to the
absence of information about an agency.
To test these competing claims, we use a difference-in-differences statistical framework
to assess the impact of the treatment, alteration imposed by the government, on media
coverage during the reform period. After the 2010 general election, the new coalition gov-
ernment agreed to a policy of reforming government operations – dubbed the ‘bonre of
the quangos’, these being independent agencies ofcially called non-departmental pub-
lic bodies (NDPBs) that work at arm’s length from the ministers. Since we aim to assess
the political rather than organizational endurance of the agencies, our denition of ter-
mination matches largely with what the government claimed: any reform or alteration
for which it could claim credit.3We collected counts of stories from major British news-
papers from 2005 to 2013 for all 159 executive NDPBs, a type of NDPB that formulates
and implements policy, and which has constituencies interestedin its fate and affected by
its reputation. Our results indicate that terminated agencies received less news coverage
in the reform period in which they faced termination. Most of these agencies did not go
out with a bang, even if a handful did receive national press attention, but rather with a
whimper. Our evidence is thus consistent with terminating agencies as a political control
strategy.
In the next section, we review the literature on agency termination and motivate
our hypotheses. We then describe the data collection and data structure for the
Public Administration Vol.93, No. 4, 2015 (1168–1183)
© 2015 John Wiley& Sons Ltd.
To continue reading
Request your trial