Conceptual Model of E-Voting in Indonesia

Dana Indra Sensuse Pandu Bintang Pratama Riswanto

Faculty of Computer Science Faculty of Computer Science Faculty of Computer Science
Universitas Indonesia Universitas Indonesia Universitas Indonesia
Depok, Indonesia Depok, Indonesia Depok, Indonesia
[email protected] [email protected] [email protected]

Abstract— Some countries have adopted the technology of [5]. The high cost of holding the election does not guarantee
electronic voting into their democratic processes. Some of them the implementation of the 2019 elections in Indonesia will run
successfully implemented e-voting in their elections, but some smoothly. Several online news pages reported problems have
countries failed to transform traditional paper-based voting into occurred since before the election process at the stage of
a technology-based automated voting system. Indonesia is a preparing the final voter list [6], reports of fraud committed by
developing country that carries out general elections every five candidates [7], as well as hundreds of officers who died
years. The vote was conducted to elect presidential and vice- suspected due to fatigue after serving at the polling station [8].
presidential candidates, national and regional legislatures, and Some countries have adopted the technology of e-voting into
members of the local representative council. Reducing errors
their democratic processes. Some of them successfully
and speed up the counting process are some of the electronic
implemented e-voting in their elections, but some countries
voting benefits. Electronic voting also eliminates the root cause
of the controversies in every election, which is a vote that is
failed to transform traditional paper-based voting into a
deemed invalid or doubted, thus it still raises doubts, especially technology-based e-voting system. Issues related to data
on the security, confidentiality of voter data, and ease of use of security and confidentiality are crucial factors in the
the system. This research aims to discover the readiness of the implementation of elections with an electronic system.
implementation of the e-voting system in Indonesia by In Indonesia, common mistakes often occur in the
conducting a quantitative survey and propose state-of-the-art electoral process are the registration process, invalid voters,
electronic voting system architecture that is suitable for
slow ballot calculations, slow tabulation processes, delays in
application in Indonesia. Our investigation shows that most of
the respondents are ready to use the e-voting system.
sending, and indications of vote-buying (money politics).
Implementing an electoral system electronically can help
Keywords— electronic voting (e-voting), electronic voting accelerate the calculation process, provide access to the
systems readiness, e-voting system community, save costs, and increase transparency [9]. The
electronic electoral system must be sufficiently resilient to the
I. INTRODUCTION possibility of fraud so that candidates for the poll and the
Voting in various democratic countries is something that public can trust the election results. Previous research
is periodically carried out every few years. Along with conducted by [9] and [10] has also measured the readiness of
information technology development, the implementation of e-voting implementation by combining public surveys and
voting in multiple countries is no longer carried out manually expert opinion on how to make the successful implementation
based on ballot papers. Nowadays, countries are implemented of electronic voting. In this study, we adopt the theory of
an e-voting system that using information technology instead technology acceptance model (TAM) and diffusion of
of traditional voting system that using ballot paper. innovation (DOI) into a model for measuring the readiness of
Governments in various countries have been prompted to use the implementation of electronic elections, while proposing a
an electronic system known as e-voting. Electronic voting system architecture that supports the implementation of
refers to any system where a voter casts his or her ballot using electronic voting by prioritizing the principles of data security
an electronic system, rather than a paper ballot (or a and confidentiality for voters.
mechanical machine to punch a paper ballot). Once recorded, II. METHODOLOGY
an electronic vote is stored digitally and transferred from each
electronic voting machine to a counting system [1]. Indonesia Several theories in information systems [11] can be used
is a developing country that carries out general elections every as a reference in measuring user readiness to accept
five years. The vote was conducted to elect presidential and technology. Theories that are often used in accordance to
vice-presidential candidates, members of the national and measure the acceptance of technology are:
regional legislatures, as well as members of the local
representative council [2]. Elections are also made to A. Technology acceptance model (TAM)
candidates for regional heads and deputy regional heads, as This theory has long been known widely used in
well as mayors or regents in district/city regions [3]. The measuring readiness to predict and explain user acceptance of
traditional paper-based voting system has been implemented information technology [12]. User trust about the relationship
since 1955 until the last election that was just held in 2019. between the technology's usefulness and the attitude and
With the number of voters registered on the permanent voter intention to use the technology influenced by factors
list spread by domestic and foreign countries at 192,866,254 Perceived usefulness (PU) and Perceived ease-of-use (PEOU)
people [4], the election requires a lot of money and cause [12]. Perceived usefulness (PU) is defined as the degree to
many problems. According to data released by the Ministry of which the user believes that using a specific system that will
Finance and the Cabinet Secretariat of the Republic of improve the performance of the job. Whereas Perceived ease-
Indonesia, the costs of holding elections in 2019 reached 25.9 of-use (PEOU) is defined as a condition when a user uses a
trillion rupiahs. While the budget for election supervision and particular system, it would be effort-free [13].
security is 4.85 trillion and 3.29 trillion rupiahs respectively

978-1-7281-7071-8/20/$31.00 ©2020 IEEE 13-14 August 2020

2020 International Conference on Information Management and Technology (ICIMTech)
387
Authorized licensed use limited to: University of Canberra. Downloaded on October 04,2020 at 16:29:52 UTC from IEEE Xplore. Restrictions apply.
B. Diffusion of Innovation (DOI) Theory conducted a literature search related to the measurement of
The Diffusion of Innovation theory describes a model readiness in implementing e-voting and literature about e-
which is used in information system research to describe the voting technology in various countries. Next, step, we
user’s adoption of new technology. Diffusion is a process in formulated ways to measure e-voting readiness with existing
which innovation, such as new technology, is channeled supporting theories, namely the technology acceptance model
through specific channels within a specified period among (TAM), diffusion of innovation (DOI) theory, and trust in
members of the social system. Innovation can be defined as an government theory. After that, we formulated ways to
idea, practice, or object that is considered new or different measure e-voting readiness with existing supporting
from an individual or group. [14]. If the concept seems unique methods, namely the technology acceptance model (TAM),
to the individual, it is an innovation. The DOI theory consists diffusion of innovation (DOI) theory, and trust in government
of five main characteristics of innovation, which are relative theory.
advantage, complexity, compatibility, trialability, and
observability. According to previous studies [13], we use two
constructs relative advantage and complexity, which were a
critical factor in the implementation of the e-voting system.
Relative advantage is the degree to which an innovation is
seen as better than the idea it supersedes, and complexity is
the degree to which an innovation is perceived as challenging
to understand and use.
C. Trust in Government
Previous research conducted by [15] states two types of
trust need to be outlined to ensure the adoption of e-voting,
namely, trust in the internet and trust in government. Trust in
government as one of the constructs that can be used as a
critical factor in implementing e-voting in Indonesia.

Fig. 2. Research procedures.

E. Data Collection and Analysis

We had surveyed seventy-eight citizens from any
backgrounds using a questionnaire using a Likert scale 1-4,
which was taken based on a sample of the Indonesian
population with the criteria of age more than equal 17 years
who had ID cards. Next, in the fourth stage, we do data
analysis using quantitative methods to test the data collected
with predetermined hypotheses. At the last step, we designed
Fig. 1. Conceptual model for measuring e-voting readiness
a development architecture of the e-voting system that was
This model is still conceptual and needs more quantitative suitable for Indonesia's implementation. We have surveyed
testing for validation. However, since the model is based on seventy-eight Indonesian citizen respondents who have the
a theory that is strong enough to measure the readiness of right to vote with a questionnaire developed from the theory
technology use as described previously, we believe that the described in the previous chapter. The factors that we use in
model can be used to measure the readiness of technology the survey are Perceived usefulness, Perceived ease-of-use,
adoption, in this case, the readiness to use e-voting. Relative Advantage, Complexity, and Trust in government.
We developed twenty-four questions in the questionnaire to
D. Research Design determine the readiness of Indonesian citizens if they used
To guide research more structured and systematic, we electronic voting systems. Data collected from the survey
compile a research design and methodology divided into five distributed were seventy-eight, consisting of 56.4% of
stages. The first step is to identify literature that is relevant to respondents were male, and 43.6% of respondents were
the topic discussed, namely e-voting. At this stage, we female. The majority of respondents are in the age group

978-1-7281-7071-8/20/$31.00 ©2020 IEEE 13-14 August 2020

2020 International Conference on Information Management and Technology (ICIMTech)
388
Authorized licensed use limited to: University of Canberra. Downloaded on October 04,2020 at 16:29:52 UTC from IEEE Xplore. Restrictions apply.
between 17-25 years, with a percentage of 50%, the 26-35 No Item r count r table result
year group is second largest with 20.5%. The third and fourth 6 PEOU1 0,586 0,227 Valid
groups were 11.5%, respectively, aged 46-55 years and 56-
65 years. There is no respondent group above 65 years in the 7 PEOU2 0,622 0,227 Valid
collected data. The majority of the respondents' education is 8 PEOU3 0,744 0,227 Valid
holding a bachelor's degree; the second largest group of
9 PEOU4 0,792 0,227 Valid
respondents in a senior high school degree, and the third
group of respondents is magister degree education. Based on 10 PEOU5 0,815 0,227 Valid
work background, the student or university student are the 11 RA1 0,714 0,227 Valid
majority of respondents with 38,5% percentage. The second
group is 25.6% of respondent are a civil servant. 10,3% of 12 RA2 0,732 0,227 Valid
respondents are entrepreneurs, and 20, 5% of respondents 13 RA3 0,759 0,227 Valid
chosen other occupations.
14 RA4 0,791 0,227 Valid
TABLE I. DEMOGRAPHIC DATA OF RESPONDENTS 15 RA5 0,743 0,227 Valid
Number of respodents 78 persons 16 CPL1 0,839 0,227 Valid
Category by Sub category percentage 17 CPL2 0,914 0,227 Valid
Male 56,40% 18 CPL3 0,862 0,227 Valid
Sex
Female 43,60% 19 CPL4 0,778 0,227 Valid
17-25 years old 50,00% 20 TIG1 0,884 0,227 Valid
26-35 years old 20,50% 21 TIG2 0,861 0,227 Valid
Ages 36-45 years old 11,50% 22 TIG3 0,861 0,227 Valid
46-55 years old 11,50% 23 TIG4 0,757 0,227 Valid
56-65 years old 6,50% 24 TIG5 0,852 0,227 Valid
Bachelor degree 64,10% a.
PU = Perceived usefulness, PEOU = Perceived ease-of-use, RA = Relative advantage,
CPL = Complexity, TIG = Trust in government
Senior High School 19,20%

Education level Master Degree 10,26% TABLE III. RELIABILITY TEST

Diploma 3,85% Cronbach's

No Item r table 5% result
Alpha
Doctoral 2,56%
1 PU 0,807 0,277 Valid
Students 38,50%
2 PEOU 0,749 0,277 Valid
Civil servant 25,6%
Profession 3 RA 0,794 0,277 Valid
Entrepreneur 10,3%
4 CLP 0,871 0,277 Valid
Others 25,6%
5 TIG 0,899 0,277 Valid
F. Validity and Reliability Test
III. RESULT AND DISCUSSION
The Statistical Product and Service Solutions (SPSS)
application software to verify the validity and reliability of the A. Respondent’s Perceptions
distributed questionnaire with seventy-eight samples to test The table below shows the results of data processing with
the validity and reliability of the research instrument. After descriptive statistics about respondents' perceptions of the
testing the validity, then we conducted a reliability test to administration of elections with e-voting:
determine the consistency level of the questionnaire used to
ensure that the survey was reliable for measuring the research No Item D (%) N (%) A (%) SA (%) Result
variables.
1 PU 1.3 17.9 48.7 32.1 (+)
TABLE II. VALIDITY TEST (+)
2 PEOU 1.3 20.5 56.4 21.8
No Item r count r table result (+)
3 RA 2.6 11.5 53.8 32.1
1 PU1 0,835 0,227 Valid
4 CLP 3.8 24.4 47.4 24.4 (+)
2 PU2 0,808 0,227 Valid
5 TIG 10.3 41.1 30.8 17.9 (+)
3 PU3 0,853 0,227 Valid a.
D = disagree, N = neutral, A = agree
4 PU4 0,601 0,227 Valid b.
( + ) = positive perception, ( − ) = negative perception
5 PU5 0,629 0,227 Valid Based on five variables measured by the conceptual model
of e-voting readiness, the majority of respondents gave a

978-1-7281-7071-8/20/$31.00 ©2020 IEEE 13-14 August 2020

2020 International Conference on Information Management and Technology (ICIMTech)
389
Authorized licensed use limited to: University of Canberra. Downloaded on October 04,2020 at 16:29:52 UTC from IEEE Xplore. Restrictions apply.
positive perception to adopt e-voting technology if the local server at the polling station. Then the voting data is
election was carried out using the system. However, the transferred through a Virtual Private Network (VPN) network
variable trust in the government has no significant value. This to ensure data security goes to the data center. The entire
factor indicates that the level of public confidence in the process of data communication must be encrypted with
election organizers is still quite weak. This is related to the cryptographic protocols [10] or blockchain technology [16]
electronic voting system that will be designed and developed. that aim to maintain the confidentiality of voter data,
The e-voting system designed must implement high data safeguard unwanted things such as changing the appearance
security standards to prevent the potential for fraud so that it of applications, changing identities, entering dangerous
can increase public confidence in the election organizer. applications, attacking servers, and identity theft.

TABLE IV. VALIDITY TEST

No Factors Index score

1 Perceived usefulness (PU) 3.10
2 Perceived ease-of-use (PEOU) 2.97
3 Relative advantage (RA) 3.09
4 Complexity (CPL) 2.82
5 Trust in government (TIG) 2.60
Cumulative index score 2.92

From the results of the assessment of e-voting readiness

using the model previously described, the results of the
readiness assessment obtained with a value of 2.92 by using
index assessments using a scale of 1-4 where the higher the
benefit obtained, the more ready the community users of
e-voting systems for use the e-voting.
B. E-Voting System Development
The main requirement in e-voting is to have an Indonesian
Citizenship Identity Card and to be registered on the final
voter list. By the rules of the Indonesian republic's
government, citizens whose names are not registered on the
final voter list can still carry out voting as long as they have a
resident card that matches the address of the polling place. If
Fig. 3. E-Voting process activity.
the address on the voting card of the voter does not match the
address of the polling station, it can still conduct voting while
The data is combined with data from other polling stations
having C5 form issued by the local district/city election
spread throughout the polling area on the election commission
commission. The verifier officer then checks the national ID
data center. Then finally, the general election commission will
card for the voter and matches the voter list. If the voter data
hold a plenary meeting based on the national recapitulation of
is on the list of voters, the verifier officer will print the receipt
voting data that is entered into the e-voting system database as
and give it to the voters. Election participants then enter NIK
the basis for determining the results of the general election.
and tokens into the system to log in. If the address of the voter
The stages of electoral vote calculation in the e-voting system
is the same as the address of the polling location's location, the
are carried out automatically by the system to prevent fraud in
voters can make an election against the candidates for
the vote counting process that may occur in the manual
president and vice president, members of the legislature,
process. When voting was completed on polling day, the
members of the regional parliament, and members of the local
polling group organizers sent a recapitulation of the results of
representative council. However, if the voter's address is not
vote counting through the e-voting system through secure
the same as the address of the polling place, then the voter can
VPN data communication and encryption protocols to ensure
only make an election against the presidential and vice-
the security of voting data to the data center. These stages are
presidential candidates. Figure 3 show the activity processes
carried out to protect the data and minimizing the possibility
at the polling station. The next process after voting is doing a
of fraud so that the candidates participating in the election, as
vote count. If in the manual system, the vote-counting process well as all relevant parties, can receive the results of the vote.
was done by counting the ballot papers that were cast for each In several recent studies [17], both encryption protocols and
candidate. In the e-voting system, the vote calculation will be blockchain technology are the most up-to-date approach to
automatically carried out by the system. After the voting securing related data [18]. The electronic electoral system's
process, the leader of the polling group prints the results of safety and reliability must be handled in such a way that the
vote counting on C1 form, then is signed by the chairman and electronic electoral system can be the main alternative to
members of the voting group along with the witnesses’ replacing the manual method with ballots to increase public
present. After the signing process is complete, the C1 form is trust in the process of holding elections.
then scanned and uploaded into the e-voting system and
submitted to the election commission data center. The results
of the vote calculation are displayed on a large screen so that
witnesses and the people present can witness the
recapitulation of vote counts. The voting data is stored in a

978-1-7281-7071-8/20/$31.00 ©2020 IEEE 13-14 August 2020

2020 International Conference on Information Management and Technology (ICIMTech)
390
Authorized licensed use limited to: University of Canberra. Downloaded on October 04,2020 at 16:29:52 UTC from IEEE Xplore. Restrictions apply.
 voting users, but the aspect of e-voting organizers is also
crucial in the success of implementing e-voting system.
Since this work was conducted before the COVID-19
outbreak, the architectural design used still adopted the polling
station's election process. We suggest that further research
design an e-voting system using smartphone technology to
avoid the potential risk of transmitting COVID-19 through the
tools or objects used in the polling station's voting process
[19].
ACKNOWLEDGMENT
This work supported by hibah Publikasi Terindeks
Internasional (PUTI) Universitas Indonesia 2020 with
contract number: NKB-881/UN2.RST/HKP.05.00/2020.
REFERENCES

Fig. 4. Proposed technology architecture framework. [1] M. K. Alomari, “E-voting adoption in a developing country,”
Transform. Gov. People, Process Policy, vol. 10, no. 4, pp. 526–547,
IV. CONCLUSION 2016.
From the obtained value analysis it can be concluded that [2] Komisi Pemilihan Umum, “KPU - Portal Publikasi Pemilihan Umum
in general Indonesian people are prepared to use the e-voting 2019.” [Online]. Available: https://infopemilu.kpu.go.id/. [Accessed:
system, only trust factors are a concern in this study because
17-May-2019].
it turns out based on answers from respondents that using e-
voting systems does not directly increase public trust towards [3] Wikipedia, “Pemilihan umum di Indonesia,” Wikipedia.Org, 2011.
the government, so that the e-voting system developed must [Online]. Available:
take into account factors such as system security, data http://id.wikipedia.org/wiki/Pemilihan_umum_di_Indonesia.
security, communication network security, and applicable [Accessed: 04-May-2019].
standards with best practices so that the e-voting system built [4] KPU, “Keputusan Komisi Pemilihan Umum Republik Indonesia
will be able to increase public trust as its users. The readiness
Nomor : 860/PL.02.1-Kpt/01/KPU/IV/2019 Rekapitulasi Tentang
of the people who use the e-voting system is one of the
Daftar Pemilih Tetap Dalam Negeri dan Luar Neger Hasil Perbaikan
determining factors for the success of the implementation of
the system. Besides, the readiness factor for supporting Ketiga Tingkat Nasional Pemilihan Umum Tahun 2019.”
infrastructure and election organizers is another factor that can [5] S. K. RI, “Naik 61% Dibanding 2014, Anggaran Penyelenggaraan
be further investigated. Issues related to data security and Pemilu 2019 Capai Rp25,59 Triliun | Sekretariat Kabinet Republik
confidentiality are also the main topics that are often Indonesia.” [Online]. Available: https://setkab.go.id/naik-61-
questioned in the implementation of e-voting systems. In this dibanding-2014-anggaran-penyelenggaraan-pemilu-2019-capai-
paper, we investigate the readiness of the Indonesian people rp2559-triliun/. [Accessed: 05-May-2019].
in accepting the e-voting system. The results of these
investigations in general, most of the Indonesian people can [6] Tribunnews, “Permasalahan Pemilu 2019, Mulai dari Penyusunan
allow the adoption of new e-voting technology and believe DPT Hingga Penyebaran Hoaks - Tribunnews.com.” [Online].
that the e-voting system can help the implementation of Available:
elections can be better in terms of time and trust We propose http://www.tribunnews.com/nasional/2019/02/28/permasalahan-
the model of e-voting implementation that we consider pemilu-2019-mulai-dari-penyusunan-dpt-hingga-penyebaran-hoaks.
in this paper. [Accessed: 05-May-2019].
The results of this study have shown the value of the [7] kompas.com, “TKN Terima 14.843 Laporan Dugaan Kecurangan
readiness of the implementation of the e-voting system and the yang Untungkan Kubu Prabowo.” [Online]. Available:
conceptual model of the e-voting system in Indonesia. https://nasional.kompas.com/read/2019/05/01/21010231/tkn-terima-
However, there are still shortcomings that can be improved in 14843-laporan-dugaan-kecurangan-yang-untungkan-kubu-prabowo.
the next study. From the respondents 'data obtained as many [Accessed: 05-May-2019].
as seventy-eight respondents used to measure e-voting
readiness using respondents with higher education levels [8] merdeka.com, “Jumlah Petugas KPPS Meninggal Bertambah,
ranging from high school to doctoral degree, while lower Totalnya Jadi 440 Orang | merdeka.com.” [Online]. Available:
education levels did not include the respondents' data obtained https://www.merdeka.com/politik/jumlah-petugas-kpps-meninggal-
in this study. Respondent data collected does not represent bertambah-totalnya-jadi-440-orang.html. [Accessed: 05-May-2019].
remote areas in Indonesia where internet connection is a [9] M. Lubis, M. Kartiwi, and Y. Durachman, “Assessing privacy and
common problem that is common in these remote areas. This
readiness of electronic voting system in Indonesia,” 2017 5th Int.
positively affects the value of people's readiness to accept new
Conf. Cyber IT Serv. Manag. CITSM 2017, 2017.
technologies where technology uses the internet as one of the
conditions. Then finally, the preparedness factor of the e- [10] M. Maletić, D. Barać, V. Rakočević, T. Naumović, and A. Bjelica,
voting organizer also needs to be measured to ensure that the “Scaffolding e-voting in developing countries,” Manag. Sustain. Bus.
readiness of delivery is not only from the perspective of e- Manag. Solut. Emerg. Econ., pp. 1–16, 2019.
[11] D. Aizstrauta, E. Ginters, and M. A. P. Eroles, “Applying theory of

978-1-7281-7071-8/20/$31.00 ©2020 IEEE 13-14 August 2020

2020 International Conference on Information Management and Technology (ICIMTech)
391
Authorized licensed use limited to: University of Canberra. Downloaded on October 04,2020 at 16:29:52 UTC from IEEE Xplore. Restrictions apply.
 diffusion of innovations to evaluate technology acceptance and [16] H. Yi, “Securing e-voting based on blockchain in P2P network,”
sustainability,” Procedia Comput. Sci., vol. 43, no. C, pp. 69–77, Eurasip J. Wirel. Commun. Netw., vol. 2019, no. 1, pp. 1–10, 2019.
2015. [17] Y. Abuidris, A. Hassan, A. Hadabi, and I. Elfadul, “Risks and
[12] F. D. Davis, “Perceived Usefulness, Perceived Ease of Use, and User Opportunities of Blockchain Based on E-Voting Systems,” 2019 16th
Acceptance of Information Technology,” MIS Q., vol. 13, no. 3, p. Int. Comput. Conf. Wavelet Act. Media Technol. Inf. Process.
319, 1989. ICCWAMTIP 2019, pp. 365–368, 2019.
[13] E. Aljarrah, H. Elrehail, and B. Aababneh, “E-voting in Jordan: [18] A. Olumide S., B. Olutayo K., and S. E. Adekunle, “A Review of
Assessing readiness and developing a system,” Comput. Human Electronic Voting Systems: Strategy for a Novel,” Int. J. Inf. Eng.
Behav., vol. 63, pp. 860–867, 2016. Electron. Bus., vol. 12, no. 1, pp. 19–29, 2020.
[14] E. M. Rogers, Diffusion of innovations - 3rd Edition. 1995. [19] G. Ye et al., “Environmental contamination of SARS-CoV-2 in
[15] A. Powell, C. K. Williams, D. B. Bock, T. Doellman, and J. Allen, “E- healthcare premises,” J. Infect., 2020.
Voting intent: A comparison of young and elderly voters,” Gov. Inf.
Q., vol. 29, no. 3, pp. 361–372, 2012.

978-1-7281-7071-8/20/$31.00 ©2020 IEEE 13-14 August 2020

2020 International Conference on Information Management and Technology (ICIMTech)
392
Authorized licensed use limited to: University of Canberra. Downloaded on October 04,2020 at 16:29:52 UTC from IEEE Xplore. Restrictions apply.