. 24
( 53 .)


York City and Chicago police departments refused to disclose their
MOU; the Chicago police department argued that the agreement was
federal property and that requests for the agreement should be sent
to the federal government.78 The Houston Police Department was
more forthcoming, however, and released its 2003 agreement with
the JITF. The agreement af¬rms that “to the greatest extent possible,
there should be transparency between and among” members of the
JRIES network. But JRIES members also promise that received infor-
mation will be treated as “the property of the originating agency” and
that requests for such information under state or local laws will be
This limitation on transparency is potentially troubling. Critics
have already expressed concern that JRIES will provide the Depart-
ment of Defense with a way of circumventing restrictions on the
collection of domestic intelligence, and the MOU™s conditions will
complicate efforts to determine whether this new capacity is being
abused. Other JRIES members have also been criticized for the mis-
use of network capabilities. One of the three founding members of
JRIES, the California Anti-Terrorism Information Center (CATIC),
is itself the hub of an information-sharing network of law enforce-
ment agencies within California. In 2003, CATIC was reorganized
following the revelation by the Oakland Tribune that it had spread
information on the activity of nonviolent political groups to local
The restrictions contained in the JRIES agreement are replicated
in several other new information-sharing schemes. The Homeland
Security Act of 2002 authorizes the Department of Homeland Security
to develop a new policy for the distribution of “sensitive homeland
security information” to state and local of¬cials that also blocks the
disclosure of shared information. The law stipulates that information
sent to state and local governments remains under the control of the
federal government, and that state or local open government laws
will not apply to it. For added protection, the law anticipates that
state and local of¬cials will sign nondisclosure agreements before
receiving information under the policy. New policies for the sharing of

Opaque Networks

“critical infrastructure information” with state and local of¬cials also
include provisions to override state and local transparency laws.81

Intractable secrecy
Conventional wisdom tells us that the twentieth century was the age
of strong states and large bureaucracies. Governments managed their
own problems, largely without intensive interaction with neighbor-
ing governments; and they did this work through large departments
or agencies over which they exercised total control. By the end of the
twentieth century this mode of governance was dying. Many prob-
lems require now the combined efforts of agencies in many differ-
ent jurisdictions. In a real sense, the administrative capability of one
state “ its organizational ability to diagnose and solve problems “ is
determined in large measure by the capabilities of departments or
agencies in other sovereign states.
We had begun to shift from the age of bureaucracy to the age of
networks. “Networks,” said the noted analyst Manuel Castells in 1996,
“are the fundamental stuff of which new organizations are and will be
made.”82 (Similarly, Ronald Diebert and Janice Gross Stein suggest
that the network will be “the dominant form of social organization in
postindustrial society.”83 ) Increasingly, Castells argues, work will be
accomplished through “network enterprises” whose “system of means
is constituted by the intersection of autonomous systems of goals” “
in other words, by the collaboration of units that are independent
and formally integrated in other organizational structures that may
well have distinct priorities. The expansion of security networks “ the
new “security architecture,” in Etzioni™s terms “ may be a harbinger
of how governing structures will evolve in other areas of government,
just as the emergence of large defense and police forces marked the
advent of the age of bureaucracy a century earlier.
The threat to open government is straightforward. As networks
become larger, and information ¬‚ows among network members
become deeper, the inventory of information held by any one agency
at a particular point in time will change. That inventory of informa-
tion will increase; but the proportion that has been received from
agencies in other jurisdictions will also increase, and the number
of agencies whose information is represented in that inventory will

Blacked Out

increase as well. We expect that agencies with a larger and more
diverse inventory of information will make better decisions. But
sometimes they will make mistakes, and when they do, we will want
to know what information they had available when they made their
Unfortunately, conventional open government laws will do little
to reveal what information was available to decision makers in a par-
ticular jurisdiction. The stock of information that was produced by
an agency on its own might be subject to the usual test on disclo-
sure, balancing the bene¬t and harm from release of information.
But the increasing proportion of information that is received from
other agencies will not be subject to this test. This shared informa-
tion will be subjected to con¬dentiality rules; the agency that provided
the information will continue to “own” “ and therefore control “ it.
The public record (and perhaps even the archival record84 ) will have
a large hole where the shared information once had been.
There might seem to be a simple solution to this problem: Have
the resident of jurisdiction A (which received the information) simply
ask jurisdiction B (which sent the information) to provide it directly.
There are several reasons why this will not be an effective remedy.
Jurisdiction A may be unwilling to reveal the source of its informa-
tion, or may have received its information from so many sources that
further searching is impracticable. Or the citizen in jurisdiction A
may have no right to the information from jurisdiction B. (A resident
of New York cannot use Pennsylvania™s public records law, and a resi-
dent of the United States cannot use the European Union™s.) Or there
may be no effective right to information. (A citizen of the United King-
dom who applies to the FBI for information will ¬nd that a complex
request will require more than a year for a decision.85 ) Underlying
all of this is a more fundamental question: Why should a citizen™s
capacity to hold his own government accountable hinge on the trans-
parency rules adopted by another government?
There is an alternative approach, one that acknowledges the
right of agencies within a network to make their own judgment
about the release of shared information, or that (a variation on
the theme) establishes effective procedures to ensure that intergov-
ernmental information-sharing agreements conform to the spirit of
domestic transparency laws. The ease with which this might be done
should not be overestimated. As networks grow larger, the dif¬culty

Opaque Networks

of coordinating practice on the disclosure of information by net-
work members increases. Current practice “ essentially a ¬‚at rule
against disclosure of shared information “ has the virtue of being
easily applied.
Another dif¬culty is political. The challenge of renegotiating
information-sharing rules also increases as the population of agencies
within a network increases, and existing rules become entrenched
in practice. Outsider agencies face the reality that the bene¬ts of
joining an established network are large, while the prospects of
successfully persuading existing members to rede¬ne the terms on
which the network operates are very poor. This is the predicament
that confronted the countries of Central and Eastern Europe, whose
leaders saw the substantial bene¬ts of joining NATO and the futil-
ity of attempting to engage NATO about the reasonableness of its
information-sharing policy. The same predicament will be confronted
by other states on the edges of emerging transnational and domestic
security networks. Opaque networks tend to stay opaque.


“The era of big government is over,” President Bill Clinton told
Americans in his 1996 State of the Union Address, promising a
“smaller, less bureaucratic government in Washington.”1 This was
both true and untrue, depending on how one thought about the ques-
tion. The number of employees working within the federal bureau-
cracy had undoubtedly declined substantially during Clinton™s ¬rst
term as President. On the other hand, the level of government spend-
ing had not radically changed, nor had the catalogue of functions
for which government was responsible been reduced. What had cer-
tainly changed was the Clinton administration™s attitude toward the
means by which government work was to be done. The volume of
work that was being transferred to private contractors and nonpro¬t
organizations was growing steadily.
Clinton™s statement represented a liberal concession to the new
realities of governance. Almost two decades earlier, rising conser-
vative forces began to threaten the structure of government as it
had developed in most advanced democracies in the preceding half-
century. The conservative challenge was pointed: State-owned busi-
nesses were to be divested “ or privatized, to use a phrase popular-
ized by the Economist magazine. The power of regulatory agencies
would be sharply reduced. Government agencies would retreat from
the business of directly producing health, education, and other social
services. Private enterprise, working in lightly controlled markets,
would take up responsibility for producing these services instead.2
In the 1990s, many liberal policymakers attempted to absorb this
conservative challenge by proposing a new way of thinking about
government. Liberal purposes, they conceded, might not necessarily

The Corporate Veil

mean the expansion of bureaucracy. Governments had to be clever in
¬nding other ways of achieving their goals, which might include the
use of contractors, or the delegation of functions to nonpro¬t organi-
zations, or the creation of new markets for the production of services.
Americans called this “reinvented” government, in which policymak-
ers retained the responsibility for “steering” the ship of state but
were indifferent about the means by which the “rowing” was accom-
plished.3 British activists called this a “third way” of thinking about
the role of the state, with a point of view that could be described as
“structural pluralism”: bound neither to the monisms of bureaucracy
or market, and pragmatic in the choice of methods for advancing the
public interest.4
The result of this debate over the role of government “ of conser-
vative thrust and liberal parry “ is, in most advanced democracies, a
public sector whose structure has been fundamentally transformed.
Large state-owned industries have been transferred to private hands.
Major utilities “ responsible for the provision of water, electric power,
telecommunications, and other services “ have been sold off. Private
enterprise has entered areas that were once regarded as the core of
the public sector.
In the United States, one company, Edison Schools, boasts that
it operates so many elementary and secondary schools that it could
be counted as one of the largest school systems in the United States.
Around the world, the business of providing water and sewer systems
is now dominated by three French and German ¬rms “ Ondeo, Veolia,
and RWE Thames Water. A Danish ¬rm, Group 4 Falck, operates a
network of prisons and detention centers spanning four continents.
An Australian business, Macquarie Infrastructure, has developed a
lucrative business in building and operating toll highways and bridges
around the world.5 Britain™s Labour government, once the main pro-
ponent of an expansive state sector, now has a policy of encouraging
private businesses to build hospitals and schools on its behalf.
Even the defense sector “ surely the most basic state function “ has
been laid open for business. It is estimated that the private military
industry earned $100 billion in global revenue in 2003.6 So many con-
tractor employees were at work in occupied Iraq in 2004 “ by some
estimates, 20,000 or more “ that analysts suggested it was the pri-
vate military industry, and not the United Kingdom (with only 10,000
troops in the ¬eld), that should be counted as the second-largest

Blacked Out

contributor to the war effort.7 Contractors engaged in combat in Iraq,
took heavier casualties than some regular combat forces, and played
a controversial role in collecting intelligence.8

Blurred boundaries
The transformation of the architecture of the public sector over the
last two decades has caused confusion about the applicability of dis-
closure laws, most of which were drafted with the purpose of improv-
ing transparency within government agencies staffed by government
employees. As work left government departments “ to go to contrac-
tors, privatized utilities, and nonpro¬t organizations “ the principle
of access to government documents began to break down.
The simplest problem was that of gaining access to the contracts
that governments signed when they decided to shift functions (the
operation of a prison, for example) to a private operator. The terms
of a prison contract are critically important: This single document
de¬nes the conditions under which prisoners will live, as well as the
incentives that shape the behavior of their keepers. The contract can
be very speci¬c on these points. One prison contract between the
state of Western Australia and Corrections Corporation of Australia,
a subsidiary of the American Wackenhut Corrections Corporation,
tied payment to twelve performance targets, including a maximum
number of attempted suicides (twenty-¬ve) or prisoner-on-prisoner
assaults (thirty) each year, and imposed a ¬‚at $100,000 ¬ne for each
escape or death in custody from unnatural causes.9
In principle there is no reason why contracts such as these cannot
be put in the public domain. Because a copy of the contract is held
by a government agency, it is unambiguously a government record.
In practice, however, many governments who began relying heavily
on contractors throughout the 1990s also resisted demands for dis-
closure of contracts. Contractors pressured policymakers to keep the
documents secret, to avoid disclosure of information that would have
been valuable to competitors and other prospective clients; but gov-
ernments also had their own reasons to keep contracts secret “ for
example, to obscure evidence that might compromise their claims
about the success of highly controversial privatization programs.


. 24
( 53 .)