Representation of the People Bill — Written evidence submitted by Democracy Club (RPB44)
Parliament bill publication: Written evidence. Commons.
Democracy Club is a Community Interest Company, founded in 2015. We build and run digital voter information tools for UK elections, providing voters with the
core
details
they
need:
what’s
up
for
election,
who
the
candidates
are,
and
where
the
polling
station
is.
We’ve
worked
with
the
Electoral
Commission,
Welsh
Government,
and
Electoral
Office
for
Northern
Ireland,
and
have
partnered
with
every
local
authority
in
the
UK
to
collect
the
data
which
powers
our
tools.
We have some specific suggestions for amendments, focused on voter
information
and
elections
data.
These
are:
Electoral Commissionpower to request election information ● The power to request election information from councils should not be
reliant
on
a
Statutory
Instrument.
Instead
this
should
follow
the
precedent
of
the
Elections
and
Elected
Bodies
(Wales)
Act
2024
,
and
allow
the
Electoral
Commission
or
Secretary
of
State
to
request
information
without
the
need
for
a
new
Instrument
every
time
something
needs
to
be
changed.
Election notice reform ● The Statement of Persons Nominated should specify which emblem each
candidate
will
use.
● Individuals who will not appear on the ballot paper should not be included
in
the
Statement
of
Persons
Nominated.
● The Declaration of Result should include the total electorate and turnout,
preferably
by
ballot
box.
We also have some more general observations on the consequences of the bill,
specifically
on
the
Electoral
Commission’s
new
power
to
regulate
individual
candidates,
candidate
identification,
and
the
challenges
presented
by
automatic
voter
registration.
1
Proposed amendments
Section 49 and election information
Section 49 of the bill allows for the Electoral Commission to ‘require the provision
of
information
that
the
Secretary
of
State
or
the
Electoral
Commission
reasonably
requires
for
the
purpose
of
helping
people
to
understand
the
election,
referendum
or
recall
petition
process’.
For the last 10 years, Democracy Club has become experts in collating information
from
every
local
authority
in
order
to
run
national
information
services.
We
welcome
any
attempt
to
improve
this
complex
process,
and
the
ability
to
request
information
is
a
huge
step
in
the
right
direction.
As
a
result,
we
strongly
support
this
new
power,
which
will
allow
the
Commission
to
run
a
comprehensive
voter
information
service.
However, we are concerned legislation (primary or secondary) could be too rigid
for
these
needs.
It’s
very
possible
that
MHCLG
and
the
Commission
could
unintentionally
write
a
statutory
instrument
requiring
councils
to
publish
data
in
a
format
that
turns
out
not
to
be
useful,
or
is
quickly
proven
outdated.
The complex interplay of software vendors, changes proposed in this bill, and the
ever
changing
nature
of
technology
means
that
there
are
risks
in
linking
secondary
legislation
with
digital
delivery.
We suggest that the Commission be given powers of direction that they can
adapt
to
their
changing
needs
over
time.
This
would
mirror
the
powers
given
to
the
Welsh
Democracy
and
Boundary
Commission
in
the
Elections
and
Elected
Bodies
(Wales)
Act
2024
.
Section
20B(2)
gave
the
Democracy
and
Boundary
Commission
powers
to
request
information
via
Directions
rather
than
an
SI.
Election notice reform
The power for the Electoral Commission to collect data is excellent. However, this
does
not
mean
that
we
should
ignore
issues
with
more
traditional
publishing
methods.
The
new
Commission
powers
should
complement,
not
replace,
election
notices.
We strongly support moving the deadline for candidate nominations from 4pm
to
midday,
which
we
hope
will
result
in
most
councils
publishing
their
Statements
of
Persons
Nominated
on
the
same
day.
At
present,
each
year,
about
half
of
councils
publish
at
close
of
nominations,
and
the
remainder
the
following
day.
2
The official notices published by councils remain the main method of putting
election
information
into
the
public
domain.
We
do
not
think
these
are
currently
fit
for
purpose,
as
they
omit
several
key
pieces
of
pre-
and
post-
election
information.
Modifying
these
notices
would
increase
the
amount
of
information
published
by
councils,
without
requiring
any
additional
data
collection
on
the
part
of
electoral
services
teams.
The suggestions below relate to two notices as defined in the Representation of People Act, 1983, the Statement of Persons Nominated and the Declaration of Result (both in Schedule 1).
1. Ballot emblems. Political parties can register up to three different ‘emblems’ (logos) for use by
candidates
on
ballot
papers.
However,
if
a
candidate
has
chosen
to
use
an
emblem,
this
is
not
mentioned
on
the
Statement
of
Persons
Nominated.
Consequently
we
do
not
know
with
confidence
which
emblem,
if
any,
will
appear
next
to
each
candidate.
Knowing
this
information
in
advance
would
help
many
voters,
especially
those
with
visual
impairments,
prepare
for
polling
day.
Suggested solution: The Electoral Commission already provides each emblem
with
an
ID
number
in
its
parties
database.
The
Statement
of
Persons
Nominated
could
be
modified
to
include
this
ID,
perhaps
alongside
an
image
of
the
emblem
itself.
2. Withdrawn or disqualified candidates. The Statement of Persons Nominated is currently defined in the Representation
of
the
People
Act,
1983,
as
a
list
of
‘the
persons
who
have
been
and
stand
nominated
and
any
other
persons
who
have
been
nominated,
with
the
reason
why
they
no
longer
stand
nominated
.’
The inclusion of people who are no longer nominated is confusing. While some
councils
publish
these
withdrawn
candidates
separately,
they
are
commonly
included
alongside
the
other
candidates,
and
identified
only
by
a
very-easy-to-miss
note.
As
well
as
potentially
misleading
voters,
this
causes
frustrating
issues
for
anyone
attempting
to
collect
candidate
data
at
scale.
Suggested solution: The Statement of Persons Nominated should only include
validly
nominated
candidates;
this
is
already
the
case
in
Scottish
council
elections.
1
The
Association
of
Electoral
Administrators
has
in
the
past
suggested
that
1
The Scottish Local Government Elections Order 2011, Schedule 1, Part VII Form 7.
3
councils should be required to publish a ‘statement of rejected nominations’, to
enhance
the
transparency
of
the
nomination
process.
2
This
projected
additional
notice
could
easily
incorporate
withdrawn
or
disqualified
candidates.
3. The Declaration of Result. After an election is concluded, councils are required to publish a Declaration of
Result,
which
is
defined
as
“[the]
name
and
of
the
total
number
of
votes
given
for
each
candidate
together
with
the
number
of
rejected
ballot
papers”.
This
declaration
does
not
require
councils
to
include
the
total
electorate
on
the
day
of
the
poll,
nor
does
it
require
them
to
publish
the
number
of
ballot
papers
issued
(that
is,
the
number
of
people
who
voted).
Consequently
it
is
often
impossible
to
know
the
turnout
for
an
election
without
asking
the
council,
creating
more
work
for
everyone
involved.
The
number
of
ballot
papers
issued
is
important,
because
the
turnout
cannot
be
worked
out
from
votes
in
cases
where
each
elector
had
more
than
one
vote
(in
a
three
seat
ward,
for
example).
Suggested solution: The Declaration of Result should be modified to include the
electorate
and
number
of
ballot
papers
issued.
Better
still,
it
might
be
modified
to
require
electorate
and
turnout
by
ballot
box
,
which
would
instantly
unlock
a
fantastic
new
resource
for
researchers,
campaigners,
and
the
general
public.
This
information
is
already
collected
as
part
of
the
verification
process,
so
it
would
simply
be
a
matter
of
including
it
in
the
Declaration.
3
3
Representation of the People Act 1983, Schedule 1, Part III, para. 45
.
2
Association of Electoral Administrators, Blueprint for a Modern Electoral Landscape (2021), p.39.
4
Observations
Electoral Commission regulation of candidates
The Bill extends the Electoral Commission’s regulatory role to include
enforcement
of
candidates’
activities.
For
this
to
happen,
the
Commission
will
need
a
list
of
candidates.
Currently,
there
is
no
single
authoritative
source
of
candidate
data
held
centrally.
This creates challenges for the Commission, as any regulation will require a
comprehensive
list
of
candidates.
Manually
collecting
this
information
from
individual
local
authorities
would
be
time-consuming
and
have
an
inherent
risk
of
inconsistency
and
error.
One option might be to rethink how we do candidate nominations. Instead of
candidates
needing
to
submit
their
nomination
forms
to
their
council,
the
UK
could
introduce
a
centralised
candidate
nomination
and
verification
process,
administered
by
The
Electoral
Commission.
This
could
operate
in
a
manner
analogous
to
the
Individual
Electoral
Registration
(IER)
service,
with
candidates
submitting
a
digital
nomination
form
through
a
central
system,
which
is
then
passed
to
the
relevant
local
authority.
This
would
also
enable
the
new
rules
on
candidate
identification
to
be
more
easily
enforced
and
take
pressure
off
electoral
service
teams,
without
fundamentally
changing
our
decentralised
electoral
system.
Such an approach would offer several potential benefits● A single, authoritative dataset of candidates for each election, enabling the
Commission
to
perform
its
expanded
regulatory
duties
● Improved identity verification, ensuring that candidates are who they claim
to
be
at
both
nomination
and
reporting
stages
● Greater efficiency and reduced administrative burden for local authorities,
by
standardising
aspects
of
the
nomination
process
● Enhanced data analysis, including the ability to link candidacies across
elections
(for
example,
where
an
individual
stands
in
both
local
and
parliamentary
elections
on
the
same
day)
There may also be an opportunity to align nomination and reporting processes
through
the
use
of
existing
digital
identity
infrastructure,
such
as
GOV.UK
One
Login.
This
could
provide
a
consistent
mechanism
for
verifying
candidates
at
the
point
of
nomination
and
subsequently
authenticating
them
when
submitting
campaign
finance
returns.
We recognise that this is a complex area, particularly given the scale of local
elections,
where
there
may
be
over
30,000
candidates.
Any
changes
would
5
require careful design to ensure resilience, accessibility, and compatibility with
existing
electoral
management
systems.
However,
we
consider
that
a
more
integrated,
digital-first
approach
to
candidate
nominations
and
verification
could
form
a
valuable
foundation
for
the
Commission’s
evolving
role.
Automated and automatic registration
Democracy Club is neutral on the policy intent of AVR. If the Government wants to
move
the
responsibility
of
registration
away
from
the
individual,
we
support
the
current
strategy
of
the
initial
focus
on
data
sharing
and
the
legal
frameworks.
More automation is not simple. Matching people across different databases is
hard.
It
will
produce
both
false-positives
and
false-negatives.
We are very sceptical that existing government datasets can be automatically
merged
without
a
significant
margin
of
error.
Both
names
and
addresses
can
take
many
forms
while
still
being
valid.
Middle
names
or
second
lines
of
addresses
can
be
omitted
from
one
database
and
not
the
other.
This
requires
some
element
of
‘fuzzy’
matching,
and
there
is
no
way
to
guarantee
accuracy
using
this
form
of
automation.
Errors at this stage risk putting the wrong people on the register, creating millions
of
duplicates,
or
leaving
eligible
voters
off.
Any
of
these
outcomes
risks
damaging
trust
in
the
electoral
system.
This
work
should
move
slowly
and
only
when
there
is
a
high
level
of
confidence
in
the
data
and
the
process.
Without
care,
the
electoral
register
could
be
corrupted
and
in
turn
core
trust
in
elections
could
be
damaged.
Even with extreme care, we expect the burden on Electoral Registration Officers
(EROs)
to
increase.
Automated
systems
are
likely
to
create
duplicate
or
conflicting
records
that
will
need
to
be
resolved
by
hand.
This
will
require
additional
funding
for EROs at least. The Electoral Commission guidance to EROs says:
Automated matching would still require an element of decision making by
[EROs]
to
ensure
that
matching
undertaken
is
to
the
expected
standard
and
to
resolve
queries.
To this day, the Passport Office rejects names with accents. IT system errors forced
North
Yorkshire
Council
to
remove
apostrophes
from
their
street
names
in
2024.
The
NHS
has
no
single
system
for
changing
someone’s
name.
The
Government
Digital
Service
pilots
of
automated
matching
in
2013
used
matching
systems
that
biased
towards
western
English
names.
There
are
countless
cases
like
this
that,
when
added
together,
can
only
lead
to
automation
errors.
There is also a question about public messaging. At present, people are told to
register
themselves.
If
that
continues
alongside
automation,
duplication
will
6
increase. If campaigns stop, some people may miss out if they are not captured by
automated
processes.
This
needs
a
clear
and
consistent
approach.
Automation must be transparent. A simple software bug could systematically fail
to
register
some
groups
of
people.
A
malicious
actor
could
also
attempt
to
interfere
with
the
process.
These
risks
are
new,
and
they
go
to
the
heart
of
the
system.
In order to minimise these risks, we think consideration should be given to the
creation
of
an
“Am
I
on
the
register?”
checker.
This
would
let
voters
confirm
their
status
directly.
This
would
be
difficult
to
build,
but
no
more
so
than
the
matching
systems
required
for
automation.
It
would
give
voters
confidence
that
they
can
vote,
and
reduce
some
duplicate
registrations.
If people assume they have been added to the register automatically, they might
not
check:
this
increases
the
chances
of
people
finding
out
they
are
not
registered
only
on
polling
day.
In
this
scenario,
it
may
be
worth
considering
introducing
same-day
registration.
This
could
involve
strict
checks,
such
as
presenting
documents
at
council
offices.
But
there
should
always
be
a
route
to
correct
errors.
No
one
should
lose
their
vote
because
of
a
failure
in
an
automated
system.
We also urge caution in using pilots for testing. This is generally a good idea, but
unless
carefully
thought
through
there
is
a
very
high
chance
of
the
perception
of
a
sort
of
data-gerrymandering.
There
is
an
obvious
party
political
advantage
to
registering
more
people
in
one
area
over
another,
but
there’s
a
similar
risk
in
targeting
one
national
dataset
over
another.
For
example,
automatically
registering
all
drivers
would
benefit
a
national
pro-car
lobby,
or
all
small
business
owners
would
benefit
a
small
business
lobby.
We suggest that the 2013 work to test automated matching is continued and
tooling
is
developed
to
help
EROs
identify
possible
electors
who
are
not
on
the
register.
Then
a
manual
process
of
registering
using
IER
can
be
attempted.
This
is
a
safer
way
of
piloting:
with
human
intervention
in
the
process.
7