Slow ballot counting

Mon, 21 Jun 1999, 00:00 WIB | Source: JP

In addition to the frequent explanation given in the media
about the slow ballot counting by the General Elections
Commission, which uses a computer network system with a
centralized and closed architecture, the following explanation
needs to be mentioned:

1. The reason for the tardiness in vote counting is the manual
path from, respectively, the polls station (TPS), to the local
elections committee (PPS), from the PPS to the subdistrict
elections committee (PPK), from the PPK to the district/mayoralty
elections committee (PPD II), where the data is fed into a
computer terminal.

2. In the latest concept of information system design, it is
necessary to distinguish the flow of information and the flow of
control or the flow of validation. These two flows, depending on
the context, may constitute one path or separate ones.

3. In vote counting, in which accuracy and high speed are very
much expected while tardiness is, theoretically, often identical
with decreasing accuracy, the flow of information should be
separated from the flow of validation.

4. What has been made a scapegoat in point 1 above should be the
flow of control or the flow of validation, while the flow of
information can always be separated by providing a TPS with a
cellular phone or an SSB radio; or in the absence of these
telecommunications devices, by making use of the closest
telecommunications nodes such as a local telecommunications stall
(Wartel) (of which there about 80,000), or mobile or stationary
postal outlets (about 27,000), etc., so that one second after the
counting of ballot papers at each TPS has been completed, the
result can be reported to the PPS. In this way, the possibility
of manipulation in vote counting by state apparatuses may be
minimized because the witnesses, the monitoring personnel and the
local community are still present when the result is reported to
the PPS.

5. Separate paths will enable mutual checking between these two
paths. In addition, the flow of information will usually be
faster and more accurate because the participation of the local
community is still intense, the source of primary data and the
data are still fresh and are unlikely to be manipulated, while
the use of different tabulation forms may lead to a deliberate or
inadvertent error.

Around March 1999, we, of the National Coordination Office of
Popular Economic Posts (PER), along with our colleagues from the
Community Concerned Over Elections (Mapelu), made an offer that
the network of the community of PER, made up, among others, by
the Wasantara-Net network, which covers 107 cities from Aceh to
Irian Jaya plus 27,000 mobile and stationary outlets, the Telkom
Multimedia network, which reaches 12 cities with 1,140 ports, the
Bukopin Swamitra network, which goes down to subdistrict in rural
areas, BRI, BNI, BTPN, the network of some 80,000 Wartel and
2,500 PER nodes, of which about 500 nodes are e-mail-connected,
could participate in vote counting.

At that time, our delegation was received by Adnan Buyung
Nasution, accompanied by, among others, Hendra Dharsono, who is
responsible for the vote counting information system. Dharsono
told us then that UNDP could not accept a new proposal because
the fund allocated for the information system was very limited
and that new participation was always possible on condition that
the fund was made available by the aspiring new participant.

In the meantime, through a number of colleagues, who represent
a number of political parties, we have also suggested that vote
counting monitoring should be conducted both in the manual path
and in the electronic expertise path. It is always likely that a
computer expert will move the votes given to a particular party
to another party through a simple programming algorithm which is
made to look complicated. This is usually known as "computer
fraud".

JUSTIANI