Electronic Voting System in Mexican Elections Figueroa K. Universidad Michoacana Faculty of physics and mathematics Email: karina@fismat.umich.mx opez E. Universidad Michoacana Faculty of physics and mathematics Email: elopez@fismat.umich.mx Garc´ ıa-Garc´ ıa J. M. Instituto Tecnol´ ogico de Morelia Email: juanmanuelgarciagarcia@gmail.com Abstract—The use of technology in supporting elections offers many benefits in addition to faster tallying, cost savings, etc. Every country has its own voting system legislation. Particularly in Mexico it is possible to modify the actual voting legislation with a new proposal presented in this paper. To a large extent, our principal contribution is to avoid human errors throughout the voting process (i.e. counting). In addition, our system also issues all the necessary reports at the end of the Election Day. The counting system could decide immediately the outcome of the election. It could avoid excessive costs with empty or void ballots. In addition it could have all the necessary ballots at temporary booths (for voters who are not in their home town). Our proposal is a public-network DRE, using a client-server protocol. We propose to encrypt every vote sent using an RSA scheme in a system with a public-key infrastructure. I. I NTRODUCTION The use of technology in supporting elections offers many benefits in addition to faster tallying, cost savings, etc. How- ever, a major challenge is to maintain confidence in the election process using an Electronic Voting system (e-voting). E-voting has four dimensions as described in [1]: political, legal, technological and social. In this paper we only discuss the technological dimension and indicate which portions of it are in accordance to the Mexican Electoral Procedure Manual (COPIFE, in Spanish) [1]. Basically, there are three kinds of e-voting systems: paper- based electronic voting, Direct Recording Electronic (DRE) voting machine, and public-network DRE. In particular, DRE voting machines record the votes without transmitting them over the Internet or another network. The interface of a DRE machine often can be a touch screen or a paper ballot scanner. In this paper we present a public-network DRE system. The design of the system ensures the principles required for democratic elections, namely the guarantee of the freedom to vote, the secrecy of the vote, the non-modification of the expressed intention of the vote, and the reliability and security of the voting process. These principles are very important to ensure that the most vulnerable issue of the election process can be addressed: counting of votes. In a system without e- voting, the human error is involved and it generally has a great impact on the results. In the Mexican election process [1], voters have cast a vote in the paper ballot by marking with a cross the name of the candidate for whom they wish to vote. The ballot has a box to the right of the name of each candidate for that purpose. If they vote for more than one candidate, their ballot paper will not be counted. The election process is divided into four main parts: 1) Selecting the representatives at polling stations, 2) Polling day, 3) Proclamation of results by representatives at polling stations (i.e. reports), 4) Proclamation of results by electoral authorities. The first part is selecting the representatives at polling stations. Electoral authorities appoint some citizens as polling station officials, and train them about the process. The second part is the polling day. Polling station officials authenticate citizens, and give them a paper ballot. Every citizen must cast their vote in secret and put it at the ballot box. The third part is the proclamation of results by polling station officials. These appointed citizens should count the votes, and report the outcome. In [2] it states that the local results should be hanged at the wall outside the polling place. And finally, the electoral authorities tally all the polling station results and report the final outcome of the election day. Currently in Mexico, the electoral process is extremely expensive and full of human errors. Last year, 2012, the presidential election budget was 18 times larger than the average spent in Latin America. It is estimated that the real cost could have reached 40,000 million Mexican pesos (about 3 billion US dollars). Mexican system spent a lot of money on printing paper ballots and at the end of election day; the unused paper ballots are nullified. Report forms should be filled out with outcome by hand (which is very prone to human error). The outcome is taken to the electoral authorities and they capture the received data (again, human error is probably at this stage and with a major impact on the final result). In this paper we propose a novel Mexican voting system, even when authors in [2] state that citizens cannot trust a code that they did not totally create by themselves. In order to give some kind of certainty to the citizens, we propose to keep some components of the actual process and renew others. It is important to say that our proposal does not violate the electoral laws regulations, and no modification to these laws is required. This paper is organized as follow, in section II we present the previous work, and the most prevalent voting system around the world. Then, in section III we show our proposal and, finally in section V we discuss this proposal and add some conclusions. II. STATE OF THE ART Basically, three aspects of verifiability are considered in a voting system [3], [4]: individual, universal, and eligibility. Individual verifiability allows a voter to check that his or her own vote is included in the election outcome (this is a 2013 Mexican International Conference on Computer Science 1550-4069/13 $26.00 © 2013 IEEE DOI 10.1109/ENC.2013.6 1