Page 1 of 12

Electronic Voting and the

2014 Namibian General

Elections

Page 2 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

Page 3 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

CONTENTS

Preface ……………………………………………………………………………………………………..

Introduction ………………………………………………………………………………………………

Electronic Voting in Namibia …………………………………………………………………………

Preparation of EVMs before Election Day ……………………………………………………….

EVMs and Election Day …………………………………………………………………………………

Observer findings on EVM performance on Election Day …………………………………..

Conclusion and Recommendations ………………………………………………………………..

4

4

5

6

6

7

8

Page 4 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

PREFACE

EISA has been interested in electronic voting for the

past six years, mainly due to its involvement in the

management and delivery of corporate elections in

South Africa, where the use of electronic voting could

provide increased efficiency. EISA’s interest is also based

on the fact that many countries are considering adopting

electronic voting while some are experimenting, with

electronic voting in their national and local government

elections. EISA needs to be prepared to meet the

challenges of observing a largely “invisible” process.

In order to understand how different electronic voting

systems and technologies work, EISA deployed technical

observer teams to three countries that have used different

electronic voting systems, namely the Philippines (2010)

and Brazil (2010), and more recently the 2014 Namibian

general elections where EISA deployed a pre-election

assessment team ahead of and during Namibia’s general

elections. A technical expert was also deployed.

This report focuses on the Namibia general elections

28 November 2014, which were the first such elections

in an African country conducting a national election

using electronic voting. It presents a brief overview of

electronic voting processes, the option used by Namibia,

and a summary of findings based on direct observation

by the EISA team also drew on observations from other

election observer missions deployed to Namibia.

The EISA team is thankful for the access to information

provided by the Electoral Commission of Namibia (ECN)

and the information and materials provided by Namibian

civil society organisations that monitor electoral processes

in Namibia, particularly the Election Watch project of the

Institute for Public Policy Research (IPPR).

INTRODUCTION

Electronic voting (EV) refers to the use of electronic

systems to cast and count votes. It includes punched

cards, optical scan systems, direct-recording electronic

systems (DREs) and electronic voting machines (EVMs). It

also applies to the transmission of votes via telephones,

private computer networks, or the internet. In general,

two main types of electronic voting can be described:

• Supervised electronic voting: in the presence of

representatives of electoral authorities (e.g.

electronic voting machines located at polling

stations);

• Remote electronic voting: takes place within the

voter’s own privacy without direct supervision by

representatives of electoral authorities (e.g. voting

from personal computers, or mobile phones).

Electronic voting systems have been in use since the

1960s when the United States introduced the punched

card system. This system was then followed by Optical

scan voting which allows computers to read voters’ marks

on the ballots. Later, DRE voting machines, which record

and tabulate votes in a single machine, were introduced

especially in countries with very large electorates, such

as Brazil and India. In some countries, they were used

and discarded due to integrity concerns. More recently,

internet-based voting systems have been introduced in

some European and North American countries.

There are also hybrid systems that include an electronic

ballot-marking device (similar to a DRE) with the

capacity to print a voter verified paper audit trail, and

use a separate machine for electronic tabulation.

In the early 2000s, India introduced EVMs designed

and developed by Government-owned equipment

manufacturers. The EVMs comprise a set of two devices

running on batteries. The voter uses one device, the

Voting Unit, and an Electoral Officer operates the other

device called the Control Unit. A cable connects both

units. The Voting Unit has a button for every candidate,

and can hold up to 15 candidates, but up to 4 Units can

be linked to accommodate up to 60 candidates. The

Control Units have three buttons, namely one to release

a single vote, another to see the total number of votes

cast, and another one to close the voting. The results

button is hidden and sealed, and cannot be pressed

until the Close button is pressed.

4

Page 5 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

Electronic voting systems may provide a number of

advantages:

• It may speed up and ensure accuracy of the voting and

counting process

• It may increase voter turnout

• It may reduce fraud

• It eliminates invalid votes, both null and blank votes

• It reduces the use of paper

ELECTRONIC VOTING IN NAMIBIA

In 2014, Namibia became the first African country to

conduct a national election using electronic voting.

The introduction of EV into the Namibian elections first

started being discussed in 2004. With the challenges

faced in the counting and tabulation processes

in the 2009 elections, which led to a delay in the

announcement of the election results, the use of EVMs

found its way into the 2009 Electoral Act. As a result of

this, the Electoral Commission of Namibia (ECN) began

purchasing batches of Indian-manufactured EVMs.

These EVMs are stand-alone machines and are not

connected to any computer network. They do not

transmit or receive any signal, therefore cannot be

intercepted. EVMs are also powered by batteries, thus

possible to operate in areas with no electricity for the

entire duration of the voting and counting process.

The 2014 Namibian Electoral Act allowed the use of

electronic voting, but introduced the requirement

that the use of voting machines be subject to the

simultaneous utilisation of a verifiable paper trail for

every vote cast by a voter, and any vote cast is verified

by a count of the paper trail. The introduction of this

requirement in the Namibian Electoral Act of 2014

results from a court case in India, in which the Indian

Supreme Court ruled that verifiable paper trail, should

be indispensable for voter confidence in the system.

The Voter-Verified Paper Audit Trail (VVPAT) is intended

as a verification system designed to allow voters to

verify that their vote was cast correctly, to detect

possible election fraud, and to provide a means to audit

the stored electronic results (see Picture 1 and 2 below).

However, VVPAT technology is very recent, and many of

the EVMs for the Namibian elections had already been

purchased when the technology was made available in

India. Therefore, the ECN decided to go ahead and use

EVMs in the 2014 elections without the use of VVPAT.

Instead they made use of a transitional provision in the

Electoral Act that states that different dates may be

determined in respect to the coming into operation of

different provisions of the Act, and deferred the use of

VVPAT to future elections.

Pic. 1: Explanation of VVPAT

Pic. 2: EVM with VVPAT printer

5

Page 6 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

The EVMs were first put to use in four local elections

in August 2014 and in one by-election in November

the same year, before being rolled out in the General

Elections of 28 November.

Since each Ballot Unit can only take up to 15 candidates

/ parties, if an election has more than 15 candidates /

parties, additional Units can be added in tandem with the

first one. In the case of Namibia, there was one Ballot Unit

in one voting booth for the presidential election (nine

candidates), and two Ballot Units in tandem in another

voting booth for the legislative election (sixteen parties).

Pic. 3: Electronic Voting Machine: Control Unit and

Ballot Unit on the left and Tabulator on the right

The Control Unit is the control section of the EVM and

only a polling official operates it. This unit controls the

polling process. The Control Unit consists of four sections:

the display screen, the candidate set, the results button

and the ballot button.

The Ballot Unit is the voter interface and voters only

interact with this unit and not the Control Unit. It consists

of an interconnecting cable which connects the Ballot

Unit to the Control Unit, ready lamp buttons, one register

lamp button, slide switch window, and 15 candidate

buttons. A ballot paper is inserted under the transparent

screen of the Ballot Unit.

The Tabulator is the unit placed, in the case of Namibia,

at the Constituency Results Centre for the tabulation

of results of all polling stations within a constituency.

After the vote counting, the Control Units are physically

transported to the Constituency results Centre, where

they are connected to the Tabulators and the results are

then tabulated.

The EVMs replace the issuing of a paper ballot with the

pressing of a button on the Control Unit by the polling

official and instead of the voter marking the ballot paper;

she/he presses a green button corresponding to the

political party or candidate of his or her choice on the

Ballot Unit. The voter then has to press the red Register

button on the Ballot Unit to complete the voting process.

It is important to note that other processes in the polling

station such as verification of voter identity against the

voter register, checking if a voter has voted before by

verifying the fingers for indelible ink and the inking

process still remain in place.

Preparation Of Evms Before Election Day

By law, all EVMs and Tabulators must be checked by a

manufacturer representative or by an expert person

designated by the ECN, to ensure that they are in working

order. This verification process takes place prior to any

election at the ECN offices in the presence of election

agents. After the verification process, the EVMs must be

prepared for Election Day.

6

Page 7 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

Returning Officers prepare the ballot units at the ECN

offices in the presence of the election agents by recording

the serial number of the ballot unit, stamping the ballot

paper to be inserted and displayed on the ballot paper

screen with an official ECN mark, inserting the ballot

paper in the ballot paper screen, ensuring that the name

of candidate or party corresponds with the candidate

button and sealing the ballot paper screen with the seal

clip provided by the ECN. The seal tag is then signed and

election agents who are present are allowed to sign the

seal tag.

The returning officer may connect more than one ballot

unit to a single control unit if the candidates or parties

partaking in an election are more than fifteen by setting

the slide switch to the position marked “1” if a single

ballot unit is used or to the position marked “2” if two

ballot units are used and so on.

After having completed the preparation of a ballot

unit, the returning officer prepares the control unit by

connecting the ballot unit to the control unit using the

interconnecting cable, switching on the control unit of

the voting machine, recording the serial number of the

control unit, and pressing the candidate set button to

configure the voting machine.

After the returning officer has completed the processes

referred to above, the control unit requests the setting

of the number of elections and the type of election to

be held, and the returning officer sets the number of

elections followed by the type of election by pressing the

first candidate button on the ballot unit if it is a candidate

election, or the second candidate button on the ballot

unit if it is a party election.

Afterwards, the returning officer closes the inner cover

of the candidate set section on the control unit and seals

it; closes the outer cover of the candidate set section on

the control unit and seals it; places the control unit and

ballot unit in their respective carrying cases; affixes a seal

to the carrying cases and allows the election agents who

are present to affix their own seals and record the seal

numbers.

EVMs and Election Day

On 28 November 2014 (Election Day), the ECN deployed

2,080 sets of EVMs: 1,255 at fixed, full-time polling

stations and 825 with the mobile teams that covered 2,711

temporary polling stations. The ECN also established a

technical support system for the EVMs, with 121 Namibian

IT technicians (one technician per constituency), and 31

engineers from the EVMs’ manufacturer.

Before voting started, the EVMs were prepared by the

polling station presiding officers and a pre-test was

then conducted. All these procedures took place in

the presence of the party agents and other authorised

persons such as observers who were present at the

polling station.

Presiding officers verified the seals on the carrying cases

of the voting machine, removed the ballot unit and the

control unit from their carrying cases, verified the seals

and the serial numbers on the control unit and the ballot

unit, connected the ballot unit to the control unit with the

interconnecting cable and the control unit was switched

on. Thereafter, they opened the outer and inner cover of

the “result section” of the control unit and pressed “total

button” to indicate that there were no votes recorded

on the voting machine. If by any chance votes had been

recorded on the voting machine, the presiding officer

would press the “clear button” to clear any recorded

votes.

After demonstrating that there were no votes recorded

in the voting machines, presiding officers conducted a

pre-poll test to ensure that all candidate buttons were

functioning properly, and that votes cast were properly

recorded and aggregated by the EVM.

The pre-poll test was conducted by allowing each election

agent present at the polling station to select his or her

candidate or party by pressing the “candidate button”

corresponding to the name of his or her candidate or

party and pressing the “register button” to cast his or her

vote. After all election agents cast their vote, presiding

officers pressed the “close button” to close the pre-poll

test. Presiding officers then pressed the “results button”

7

Page 8 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

to show the result of the pre-poll test which corresponded

to the votes cast. After the results of the pre-poll were

shown to all present, presiding officers pressed the “clear

button” to clear the votes recorded during the pre-poll

test.

Once the pre-poll test had been concluded, presiding

officers closed and sealed the inner and outer cover of

the results section with a clip seal, recorded the clip seal

numbers and allowed election agents to verify the clip

seal number. The EVMs were then ready for the beginning

of voting.

After the end of polling, closing and counting procedures

in the polling stations were strictly followed, and all EVM

control units produced aggregated results that were

accepted by all election agents. After the announcement

of results at the polling stations, the control units were

taken to the constituency results centres where they

were then plugged into the Tabulators1

to aggregate all

the polling stations results for further transmission to the

national results centre.

Observer findings on EVM performance on Election Day

Most election observation reports agree that polling

station staff were generally well-prepared to conduct

the EVM preparatory procedures and pre-poll tests,

with some inconsistencies that are to be expected when

systems are used for the first time. All the observed

pre-poll tests produced the expected results and no

complaints were recorded from election agents.

Five main questions were in the minds of stakeholders

regarding the use of EMVs on Election Day: (i) would they

accurately record and aggregate votes? (ii) would they

function without major breakdowns? (iii) in the case of

breakdowns, would the ECN be able to respond timeously

to assistance requests? (iv) would voters use the EVMs

with ease?, and (v) would the final polling station results

produced by the EVMs be accepted by all stakeholders?

The paragraphs to follow, explain the motivation behind

questions in more detail.

Accuracy: Given that Namibian EVMs so far have no VVPAT

capability, the only assurance that EVMs record and

aggregate votes accurately is the pre-poll test conducted

immediately before the beginning of vote. Since the EVMs

do not transmit or receive any signal and are therefore

immune from hacking, it is fair to assume that, following

a successful pre-poll test, they will continue to record and

aggregate votes accurately.

Breakdowns and technical support: There were a

few cases of EVM breakdowns that were reported by

observers. In all those instances, technical support from

the ECN was swift and managed to resolve the problem

without any major impact on the voting process. Most

breakdowns at the polling stations happened in relation

to the voter identification machines, not the EVMs.

Voter-friendliness: With the exception of voters who had

voted in the few local or by-elections prior to the general

elections, this was the first time the overwhelming

majority of voters operated an EVM. On average voters

took less than 30 seconds with each EVM to make their

choice and cast their vote. There were many reported

cases of voters who were unsure about which buttons

to press, and in which order. However, with the prompt

assistance from polling officials, the vote selection and

casting process remained very swift.

Results acceptance: Despite the absence of a manual

count that all present can visualize, follow and monitor,

which was the norm under manual voting, party agents

at the polling stations did not challenge the results

CONCLUSION AND RECOMMENDATIONS

There is a general consensus among observer groups that

electronic voting was largely successful in Namibia, with

one observer mission even recommending its adoption

by other Southern African countries, and the permanent

adoption of electronic voting in Namibia .

1 The tabulators worked like calculators that collated results from a number of EVMs

8

Page 9 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

Some domestic stakeholders, particularly opposition

political parties, still alleged that the system was

manipulated in favour of the winning party, though no

evidence was presented. In this regard, the ECN clarified

that the EVM had no preprogrammed or computerised

systems attached to them. They were standalone

units that worked independently based on on-site

programming on election day.

Many stakeholders however rightly feel that the electronic

voting system was largely imposed on them without

enough opportunities for them to familiarize themselves

with it and “own” it.

EISA commends the ECN for taking the bold step to

introduce EVM and noted that the use of the EVM

was largely successful during the 2014 elections albeit

the process experienced delays with the failure of the

electronic voter identification system. Overall, the

introduction of the EVM addressed the issue of spoilt and

rejected ballots as all ballots were valid at the end of the

process. It also contributed to the quick tabulation of the

results as they were released on 1 December, less than 72

hours after the close of polling.

There are some areas for improvement, both in the

system, its deployment and management, and in the

public perception about it. Below we present a number

of recommendations to address those issues:

1. Introduce a VVPAT component to the EVMs in order

to improve transparency and increase public trust in

the system;

2. Create opportunities for regular interaction between

stakeholders and the system, through presentations

and simulations, to increase their familiarity with

electronic voting and improve their trust in it;

3. Allow space for stakeholders to present suggestions

for improvements to the system in order to strengthen

their sense of ownership;

4. Provide continuous voter education about electronic

voting and the mechanics of using the EVM to make a

choice and cast a vote in order to make them familiar

and comfortable with both the concept and the

practice;

5. Strengthen the knowledge of polling station staff

about preparatory procedures, pre-testing and how

to assist voters without violating secrecy of voting;

this is necessary in order ensure consistency in the

application of procedures for future elections.

9

Page 10 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

10

Page 11 of 12

Electronic Voting and the 2014 Namibian General Elections | A Brief Assessment Report

19

ABOUT EISA

INSTITUTIONAL BACKGROUND

EISA has since its inception in July 1996 established itself as a leading institution and influential player dealing

with elections and democracy related issues in the African continent. It envisions an African continent where

democratic governance, human rights and citizen participation are upheld in a peaceful environment.

The Institute’s vision is executed by striving for excellence in the promotion of credible elections, citizen

participation, and the strengthening of political institutions for sustainable democracy in Africa.

Having supported and/or observed over 70 electoral processes in Africa, EISA has extensive experience in

formulating, structuring and implementing democratic and electoral initiatives. It has built an internationally

recognised centre for policy, research and information and provides this service to electoral management

bodies, political parties and civil society organisations in a variety of areas, such as voter and civic education

and electoral assistance and observation. Besides its expanded geographical scope, the Institute has, for

the past several years, been increasingly working in new in-between election areas along the electoral and

parliamentary cycle, including constitution and law making processes, legislative strengthening, conflict

management and transformation, political party development, the African Peer Review Mechanism (APRM)

and local governance and decentralisation.

EISA provides assistance to inter-governmental institutions, like the African Union, and the Pan-African

Parliament, to reinforce their capacity in the elections and democracy field. The Institute has just signed an

MOU with the Economic Community of Central African States (ECCAS); the East African Community (EAC);

and the Common Market for East and Southern Africa (COMESA). Within the framework of these recently

signed memoranda, the Institute will also provide similar assistance respectively these intergovernmental

institutions. Its MoU with the African Union was also renewed in 2014.

With its headquarters in Johannesburg (South Africa), EISA has had field offices across the African continent

and currently has offices in Central African Republic, Democratic Republic of Congo, Kenya, Madagascar, Mali,

Mozambique and Somalia, and a regional liaison office at the secretariat of the ECCAS in Libreville, Gabon.

Election observation activities

EISA has deployed continental witness missions for the past ten years including missions to Angola (2008),

Botswana (1999, 2004, 2009), Central African Republic (2010, 2011), Democratic Republic of Congo (2005

referendum, 2006 elections), Egypt (2011, 2012, 2014), Ghana (2008, 2012), Guinea Conakry (2010), Lesotho

(1998, 2002, 2007, 2012,2015), Liberia (2011), Madagascar (2005, 2007, 2013), Malawi (1999, 2004,2009),

Mauritius (2000, 2005, 2010, 2014), Mozambique (1999, 2004, 2009, 2013, 2014), Namibia (1999, 2004,

2009), Senegal (2012), Seychelles (2011), South Africa (1999, 2004, 2009, 2014), Tanzania (2005, 2010),

Uganda (2011), Zanzibar (2005, 2010), Zambia (2005, 2008, 2011,2015), and Zimbabwe (2000, 2002, 2008),

Reports on these missions can also be found on our website.

Page 12 of 12

T +27 11 381 60 00 · F +27 11 482 61 63

14 Park Rd · Richmond · Johannesburg

PO Box 740 · Auckland Park 2006 · South Africa

About EISA

EISA is a not for profit organisation established in 1996 based in Johannesburg (South Africa) with field

offices in Central African Republic, Gabon, Kenya, Madagascar, Mali, Mozambique and Somalia.

Our vision

An African continent where democratic governance, human rights and citizen participation are upheld

in a peaceful environment.

Mission statement

EISA strives for excellence in the promotion of credible elections, citizen participation, and the

strengthening of political institutions for sustainable democracy in Africa.

Funded by