Memorandum submitted by Professor Patrick
Dunleavy (ID02) (London School of Economics and Political Science)
WHAT IMPACT
WILL THE
INTRODUCTION OF
IDENTITY CARDS
HAVE ON
PUBLIC SERVICES
IN GENERAL?
1. One of the reasons why a national identity
card often appears attractive to policy-makers is that it may
appear to hold out the prospect of uniquely identifying people
in the same way across many different policy fields. National
ID numbers are certainly used in this way in some European countries,
notably Finland. Here a past tradition of authoritarian government
meant that the ID card concept was introduced early on and acquired
historical acceptance. But in addition modern Finnish citizens
have a great deal of trust that their government will use information
about them in ways that respect their human rights, civil liberties
and personal freedoms, and strong and comprehensive enforcement
mechanisms backing up these protections.
2. The UK in common with the United States
and some other large liberal countries in the Anglo-American tradition
has previously lacked any such foundation for a national ID number
to be used pervasively by government to conduct its relations
with citizens. In these countries there has been a concern to
keep citizens' relations with governments compartmentalized and
to prevent information submitted for one purpose becoming available
to other departments or agencies for other purposes. Not having
a national ID number has been seen in the past as a key support
of civil liberties by making it less easy for personnel within
government agencies to bring together many different information
pots relating to the same citizen.
3. The main consequence of this approach
has been that across UK central government there has been a complex
set of identifiers in use by different departments and agencies
for tracking their relations with citizens. LSE Public Policy
Group researched this picture for a National Audit Office "value
for money" study called Difficult Forms, published
in October 2003. Our researchers accessed a wide range of forms
(519 in all) used by central government organizations for dealings
with individual citizens (that is, excluding all forms dealing
with businesses or other civil society organizations). The data
set was also confined to "first contact" forms, those
sent in by citizens as the first stage of making a claim on or
securing an authentication from government, and hence covers all
the most important and general forms submitted by millions of
citizens annually. There are in addition many thousands of follow-on
or further information forms that are sent out by departments
and agencies to selected citizens in order to clarify or supplement
information gathered by the first-contact forms—we have
not covered these follow-on forms in our data here. The survey
was conducted in early 2002, but we have no reason to believe
that the picture involved will have changed very substantially
since then.
Figure 1 overleaf shows the identifiers asked
for across the citizen forms we surveyed. The most common identifiers
used are name, address, telephone number and date of birth. The
only identity number commonly used across more than one department
or agency at present is the National Insurance number, but it
is included in only 19% of government forms used by citizens.
4. Figure 2 overleaf also shows that use
of the National Insurance number is highly concentrated in the
social welfare field, where it is asked for on four fifths (81%)
of citizen-facing forms. The second most popular field is for
taxation where it is asked for on just over a quarter of forms.
In education somewhat over a sixth of forms include the NI number
and in health just over a tenth of forms. In other policy areas
it is rarely used, even in immigration.
5. Taken together Figures 1 and 2 show the
existence of a complex ecology of identifiers currently asked
for by government agencies (the demand side for identifiers) and
supplied by citizens (the supply-side). In thinking about the
impact that the introduction of a National Identity card may have
on this situation it will be important to look at both demand-side
and supply-side factors. I give below two lists of factors that
may stimulate or inhibit use of an ID number.
6. On the demand side, the following factors
are likely to influence whether government organizations are likely
to ask for a national ID number on forms and in their other dealings
with citizens:
— Whether legislation or government
policy has required them to use ID numbers generally.
— Whether legislation or court
judgements have made it clear that it is legitimate for a national
ID number to be requested in the specific context of an individual
form or service.
— How much it would cost agencies
to have their IT systems adapted to incorporate reference to ID
numbers. Inherently it may be likely that central government organizations
will wait for a major redesign of their IT systems to happen anyway
before incorporating ID card numbers in complex systems. This
may reduce (or blur) their costs but it will also delay implementation
and postpone the point where any "critical mass" of
agencies or services is using ID cards numbers or biometric identifiers.
— How much departments or agencies
have to pay to the Home Office or its component agencies in order
to consult the National Data Register. Agencies often dislike
paying for access to identifiers. For instance, Inland Revenue
(now Her Majesty's Revenue and Customs) encouraged taxpayers to
use an IR-specific taxpayer number rather than National Insurance
numbers, which they had to pay to look up.
— How much it will cost to equip
the organization's field offices or citizen-contact offices or
facilities with equipment for testing biometrics and contacting
the National Data Register and to train staff to operate them.
— How effectively biometrics and
the National Data Register work in practice. If these systems
do not succeed, or prove more costly in operation than current
Home Office estimates suggest, other organizations will hold back
in adopting them. Similarly if biometrics do not work well in
practice, generating many exceptions or false negatives that are
troublesome and expensive to handle administratively, or if citizens
need to have extensive reassurance before offering biometrics,
then their use may be inhibited.
— How far ID cards and biometrics
prove resistant to forgery and other forms of misuse. The more
"gold standard" and reliable the ID card proves the
more its use may spread to other departments and agencies.
— How securely the National Data
Register proves to operate. In the past there has been a high
level of leakage of information from UK government databases to
unauthorized users, principally via private detectives, credit
checking and other professionals having contacts amongst and paying
bribes to public service workers (for instance, police officers
or civil servants). If this pattern turns out to be replicated
with the joined-up information held on the National Data Register
then there could be legal liability implications for agencies
sufficient to discourage them from using it.
— How far administrative processes
have moved to phone-based processes rather than personal contacts,
since biometrics require personal contact with citizens to work
and cannot currently be given by phone. (For instance, in order
to gather biometrics for the new passport the Passports Agency
has reintroduced personal interviews with all applicants, a provision
that has not applied hitherto. This adds substantially to administrative
costs.)
7. On the supply side, the following factors
will influence the behaviour of government organizations (beyond
the Home Office):
— How many citizens always or
regularly carry the ID card with them and feel happy to offer
the card and biometrics when asked to do so. If citizens commonly
leave their ID cards at home in a drawer for fear of losing it
or having it stolen, then in many single-contact contexts agencies
will have difficulties in requesting to see the ID card. If large
numbers of people turn up to offices or facilities without their
cards it will be difficult to process them or to ask that they
all return home to fetch the card.
— One key influence on whether
people will carry ID cards or not will be how hard it becomes
to establish your identity if you have lost it or had it stolen,
especially if someone else is using it (in non-biometric contexts).
There is a danger that if ID cards become widely used as identifiers
then the existing "slime trail" methods of establishing
who you are will progressively atrophy, considerably increasing
the costs and difficulties of re-establishing your identity without
an ID card. The more these costs rise the more reluctant citizens
may be to carry their cards with them. High costs may also people
off replacing their ID cards when lost, unless compelled to do
so.
— So it will be crucial whether
it becomes compulsory for citizens to show their cards for access
to all or some public services or to carry their cards when they
are out and about.
— Even if some general legal compulsion
to carry cards is introduced, the impact may depend upon the extent
to which the police enforce such provisions—it is as yet
unclear that they would have the personnel resources to do so
and there might be practical difficulties in enforcement (for
instance, in souring police relations with black or minority ethnic
groups).
— If a significant group of citizens
resist carrying or using ID cards or biometrics then the costs
involved for all public agencies will tend to increase appreciably—for
instance, in terms of many appeals or exceptions-based cases clogging
up administrative processes. There tend to be very high "gearing
ratios" in such processes—that is to say, the withdrawal
of "quasi-voluntary compliance" by only a small proportion
of citizens has a severely adverse effect on administrative costs.
A recent example of this phenomenon was the reaction to the poll
tax in 1989-91 when over a million people removed themselves from
electoral rolls and non-payments or delayed/disputed payments
of poll tax imposed significant costs both on local authorities
and on compliant taxpayers.
— Citizens' reactions to being
asked to show an ID card, to checking of ID numbers with the National
Data register, and to being asked for biometrics will all have
a final pervasive impact on how the system operates in practice.
If there is widespread resistance to any of these stages then
there may be incentives for "street level bureaucrats"
to minimize their checks. For instance, if most police officers
or other officials just ask to see an ID card and perhaps quickly
check the photo ID to establish that someone is who they say they
are, then the biometric protections which the card is supposed
to offer may not be activated. Hence there will be some utility
for criminals or others in having a fraudulent ID card that would
not survive closer inspection or offering of biometrics. If implementation
works out in this way then official trust and public trust in
the ID card being offered will tend to decline.
In general we should expect that the more problems
surround the public's acceptance of and ready use of the National
ID card the slower and more partial will be any take-up process
by other departments and agencies.
24 November 2005
|
