. 19
( 53 .)


“mediocre and fragmented recordkeeping”22 ; the complaint was
echoed in several other rich states.
The problems are substantially worse in developing countries.
Anne Thurston, founder of the International Records Management
Trust, says that the record-keeping systems of many developing coun-
tries are “in decline, and in some cases total collapse” for lack of
proper policies, trained staff, or adequate facilities.23 An early Trust
study of government personnel records in Uganda reported dire
conditions: “No temperature, humidity or pest control exist, so paper
is rotting, metal is rusting and there are layers of insects on or in
¬les (termites have damaged shelving and wasps have nested among
¬les).”24 In Tanzania, the Trust found that the ability to monitor ¬nan-
cial and personnel systems was compromised because “the system is
overwhelmed by huge volumes of unmanaged paper. For example, it is
very dif¬cult to audit the payroll because the relevant documents are
scattered in different ¬les in a variety of locations. . . . [Personnel] ¬les
are frequently incomplete, missing or misplaced.”25 Another study
found that the Ecuadorian court system™s archives had accumulated
2.5 million ¬les “ 500,000 on shelves, and the other 2 million on the
archive™s ¬‚oors. File retrieval depended on the “knowledge and mem-
ory of the Director.”26
In Kenya “ which will have a constitutional right to information
if a Bill of Rights drafted in 2004 is put into force “ missing govern-
ment ¬les were found to be a major problem. A government archivist
believed that a major reason for the loss of ¬les was corruption;
of¬cials were simply destroying incriminating documents.27 This
highlights one of the perversities of disclosure law as an anticorrup-
tion tool: It operates on the premise that the administrative system is,
in large part, not corrupt. One of the remarkable features of disclosure
systems in advanced democracies is the frequency with which they
result in the release of documents that citizens did not know existed in
the ¬rst place. That these documents are released, and not destroyed,
is a testament to the professionalism of the civil service. Even in the
cleanest civil services, however, cases of document destruction are
occasionally uncovered. The problem is likely to be more severe in
weakly professionalized civil services, particularly when missing doc-
uments can be blamed on notoriously poor record-keeping systems.
A decent system of record keeping and a reasonably profes-
sional civil service are likely to be two prerequisites for an effective

Soft States

Documents stored in the archive of the municipality of Independencia, Peru,
August 2004. Photograph by Helen Darbishire.

disclosure law. A third will be adequate resources for administering
the law. Proponents of disclosure laws have sometimes been reluctant
to discuss the potential burden of administering new laws, but they
can be substantial if the law is to be applied properly. Staff need to
be trained so that they know how to receive and respond to requests.
In departments likely to receive a large number of requests, special
of¬ces may need to be established. The processing of requests will
require some of¬cials to retrieve records and review them to sepa-
rate those that are sensitive from those that are not. Lawyers may
be needed to give advice on the interpretation of exemptions in the
law. Copies of documents containing sensitive material will have to
be made, and the sensitive portions blacked out.
In short, a functioning disclosure law will spawn its own admin-
istrative routines and bureaucracy, and this will impose signi¬cant
costs on government. In 2004, the Australian government estimated
that the annual cost of administering its Freedom of Information Act
was about $14 million, or about US$330 for each request.28 In 2000,

Blacked Out

the Canadian government estimated that the annual cost of admin-
istering its Access to Information Act was US$19.4 million, or about
US$1,340 for each information request received that year.29 The dif-
ference in the two estimates is likely a result of the relative complex-
ity of requests being handled in the Canadian system30 ; the Canadian
estimate also included the budget of the independent Information
Commissioner, who investigates complaints about the handling of
AIA requests.
In 2003, the cost of administering the U.S. Freedom of Informa-
tion Act was estimated to be $323 million, or about $100 per request.31
This is signi¬cantly lower than either the Australian or Canadian esti-
mates. Again, part of the difference is attributable to a variation in
the kind of requests received by the three governments. Eighty per-
cent of the 3.2 million FOIA requests received by the U.S. government
in 2003 consisted of requests made by clients of the Veterans Health
Administration or the Social Security Administration for personal
information ¬les; these requests are simple and highly standardized,
and can be processed quickly and at low cost.32 The average process-
ing cost for requests received elsewhere in the U.S. government in
2003 was $405. The U.S. estimate also excluded some costs “ such
as the time taken by bureaucrats outside FOI of¬ces to gather ¬les “
that were counted in the Canadian estimate. Nor is there an informa-
tion commissioner in the U.S. system, as there is in Canada. In the
United States, much of the enforcement cost “ consisting largely of
the cost of FOI litigation “ is shifted to the federal court system and
individuals who litigate FOIA cases.
These estimates are rough, but they suggest that the cost of pro-
cessing an information request in these three countries is likely sev-
eral hundred dollars, unless the process is highly routinized. In prin-
ciple, it is possible to levy charges on citizens to recoup much of this
cost. Indeed, a model right to information law drafted by ARTICLE 19
and the Commonwealth Human Rights Initiative in 2001 would allow
fees up to “the actual cost of searching for, preparing and communi-
cating” information, with waivers for personal information and pub-
lic interest requests.33 In practice, however, the Australian, Canadian,
and U.S. governments rarely charge actual costs; revenue from fees is
equal to roughly 1 or 2 percent of each government™s expenses. There
is evidence that a fee policy based on full cost recovery would cause
the demand for information to collapse entirely.34

Soft States

A working disclosure system cannot sustain itself ¬nancially; it
requires almost complete subsidy from government coffers. There are
sound arguments for such a subsidy in af¬‚uent nations. For exam-
ple, governments in these countries spend a vastly larger amount on
advertising and other promotional activities aimed at conveying infor-
mation in ways that favor of¬cial priorities. In these countries, how-
ever, the subsidy for disclosure is explicitly recognized in government
budgets: To put it another way, an allowance is made for the burden
that will be put on departments by new disclosure laws. It is not clear
that countries now adopting disclosure laws have done this; on the
contrary, the assumption appears to be that the cost of implementa-
tion will be absorbed within existing budgets.
The disparity between the approach to implementation in richer
and poorer countries was clear in 2004. The British government was
in its fourth year of planning for the roll-out of its new Freedom of
Information Act; adopted in December 2000, the law was scheduled
to go into effect in January 2005. There had been heavy investment
by government agencies in training, promotional material, and com-
puter systems to handle anticipated requests. Parliamentary commit-
tees had scrutinized the implementation effort, as had the National
Audit Of¬ce. In 2004 alone, the new Of¬ce of the Information Com-
missioner was given US$7 million to prepare for the law; the of¬ce
within the central government™s Department of Constitutional Affairs
that had responsibility for providing guidance and overseeing imple-
mentation had, in that year, a budget of US$15 million.35 Consultants
had been hired; advisory boards had been appointed; “project risks”
associated with the roll-out had been carefully monitored. Consti-
tutional Affairs™ public relations of¬ce announced that it would dis-
tribute complimentary coasters and pens bearing a new FOI logo36
(thus raising, for FOI requesters, the perverse possibility that their
denial letter would be signed with a pen celebrating the government™s
Jamaica, a poorer relation of the United Kingdom, put its Access
to Information Act into force in January 2004, after only ¬fteen
months of preparation.37 This was, admittedly, a smaller effort than
in the United Kingdom, but scale alone could not account for the
disproportion in resources available for implementation in Jamaica.
The budget for central guidance of the British FOI implementation
effort exceeded the budget of the Jamaican Access to Information Unit

Blacked Out

(with its staff of four), the government™s Archives and Records Depart-
ment, the other parts of the Prime Minister™s Of¬ce, and the Jamaican
Houses of Parliament “ combined.38
In September 2004 the Open Society Justice Initiative issued a
report on the results of a test in which requests for information
were submitted to government agencies in three countries that had
recently adopted disclosure laws “ Bulgaria, Peru, and South Africa.39
The ¬ndings were discouraging. Of¬cials frequently refused to accept
requests for information, particularly if they were submitted by mem-
bers of “vulnerable and excluded groups,” while over 40 percent
of requests that were accepted by of¬cials were simply ignored. In
South Africa, 70 percent of requests were either rejected or ignored.
The problem was not simply the “enduring re¬‚ex toward secrecy,”
the report concluded; of¬cials were often ignorant of the law, and
agencies often lacked clear procedures for handling requests. The
promise of the law had been defeated by failures in implementation.40
A similar 2004 study of the Moldovan law found that one-quarter of
state bodies ignored requests entirely, while another quarter violated
deadlines for response.41
In the af¬‚uent democracies, of¬cial recalcitrance is often reme-
died by an appeal to the courts. But this makes another assumption
about governmental capacity: speci¬cally, that the courts are able to
make a timely and independent appraisal of bureaucratic compliance
with the law. This may not always be the case. In some poorer coun-
tries, court systems are overburdened and incapable of handling cases
promptly. (In a 2004 assessment, American human rights specialists
observed that the Pakistani court system was plagued with backlogs
due to “archaic and inef¬cient court procedures”; the Indian court
system was subject to similar criticism.42 ) Court systems may also
be subject to political interference or governments may simply fail to
comply with judgments “ both of which are signi¬cant problems in
the Ukraine, for example.

Civil society capacity
To say that the right to information is a citizen™s right is, in a cer-
tain sense, misleading. There are many circumstances in which
the disclosure of information helps to protect a citizen™s important
interests. However, it is unlikely “ at least based on the experience

Soft States

of laws already in operation “ that individual citizens will, on their
own behalf, make requests for information under a disclosure law.
Even if they seek personal information (about a health bene¬t, for
example, or an adverse decision on school admission or immigra-
tion status), individuals may rely on an advocate to make a request
for them. And individuals are even less likely to make requests for
other kinds of documents “ sometimes known as “general records”
requests. Requests such as this might also protect important interests
such as the right to be informed about government decision mak-
ing, but it is more likely that a nongovernmental organization “ an
advocacy group, media outlet, union, or business “ will ask for the
The reasons for this are straightforward. Even in countries with
long-established disclosure laws, making a request for information
requires knowledge about the bureaucratic routine by which infor-
mation requests are processed and about the legal provisions that
should govern decisions on the release of information. Often it is use-
ful to have a good understanding of the organization of ¬les within
the bureaucracy “ to know where the bodies are buried, so to speak.
The act of requesting information also requires a strong sense of polit-
ical ef¬cacy and persistence, due to the long delays that may arise in
the handling of requests. Finally, asking for information may require
money, particularly if the request is novel or complex or if the law
lacks a mechanism by which appeals can be lodged at little or no cost
to the requester.
The practical impediments to the use of any disclosure law are
evident in the stories that are told about the use of new disclosure
laws. In Rajasthan, it was not the villagers of Kelwara who pursued
the request for ration dealers™ registers; it was MKSS, an organiza-
tion created by a former employee of the prestigious Indian Admin-
istrative Service (see Chapter 1). In Thailand, the case against the
Kasetsart University Demonstration School was brought by a parent
who also happened to be a public prosecutor. (In fact, most of the
individuals who ¬led complaints about noncompliance with the Thai
law in 2002 were government employees seeking information about
disciplinary actions.43 ) It was also lawyers (a high-status occupa-
tion in Japan) who constituted the Zenkoku Shimin Ombudsmen, the
organization that routed out evidence of corruption in Japanese gov-
ernment. Disclosure laws are wielded by knowledgeable, empowered

Blacked Out

professionals, even if they are used to protect the interests of a larger
Expertise and a strong sense of political ef¬cacy, while critical,
may not be enough. The U.S. Freedom of Information Act works as
it does because the federal government is surrounded by nongovern-
mental organizations and media outlets with the resources to use the
right to information aggressively. Many of these nongovernmental
organizations also take a special interest in the principle of openness.
These include groups such as Public Citizen (“We ¬ght for openness
and democratic accountability in government”), the National Secu-
rity Archive (“a counter-institution to the U.S. government™s secrecy
system”), or OMB Watch (“Our objective is to improve access to
government decision-makers and energize citizen participation”).44
Others, such as the Reporters™ Committee for Freedom of the Press,
treat the right to information as one of several issues that are impor-
tant to their core constituency. This “transparency lobby” depends in
turn on contributions from a broad community of philanthropies, as
well as favorable treatment under federal tax law.
The af¬‚uence of public interest groups within the United States,
as well as the country™s major media outlets, is remarkable even when


. 19
( 53 .)