The Implications of Not Voting in May

Mock-elections-2013-UPISVoters fill out their forms in the mock elections held at the old UP Integrated School campus on February 2013. File photo by Jun Madrid.

Aside from a sense of duty and love of country, more Filipinos should vote on May 9, 2016. The reason is tied to the very essence of democracy. Statistical analysis indicates that because there are only two effective candidates in local elections, non-voting will make one’s least preferred candidate win. Here are the details of my argument:

Elections are no doubt stressful for candidates and the Comelec, but such stress is healthy for our democracy. Democracy is often measured through the regularity of electoral exercises and the extent to which these exercises are competitive.

So what if one cannot vote? What has voting got to do with competitive elections? Let us consider what election data in 81 provinces in the Philippines from 1998 to 2013 tell us.

Let us assume that the proportion of people who did not, or were not able to vote, corresponds to the quotient when the difference between registered voters and actual voters is divided by total votes cast in the gubernatorial race.

For simplicity, we allow this proportion ([registered voters – actual voters]/total voters) to represent crudely all those who by sheer pessimism did not choose to vote, as well as those who were disenfranchised by reasons such as, but not limited to, violence, intimidation, force, and fraud.

An election is presumably more competitive if there are more candidates competing effectively1. An effective candidate, in this case, is one who can mount a campaign credible enough to drive supporters to go to the polls (without necessarily winning it).

Table 1 summarizes the regression estimates showing the relationship between not voting and the competitiveness of elections. The estimates are based on a fixed effect model treating each gubernatorial election for all 81 provinces from 1998 to 2013 as panel. Here, poverty incidence is added as a crude proxy controlling for socioeconomic conditions.

Politicians, after all, are known to exploit socioeconomic inequalities to mobilize or demobilize voters. The estimates suggest the strong relationship between not voting and the competitiveness of the electoral exercise. This implies that elections become less competitive when people skip—or for some reason, were prevented from going to—the polls.

Table 1. The relationship between voting and competitive elections2

  Model 1 Model 2
  Effective number of candidates (Golosov) Effective number of candidates (Laakso Taagapera)
Non exercise of electoral franchise -0.714** -0.925**
  (0.268) (0.313)
Incidence of poverty 0.00818a 0.00549
  (0.00428) (0.00592)
Constant 1.741*** 2.142***
  (0.201) (0.280)
AIC 675.0 863.3
BIC 687.5 875.8
N 473 473
R Squared 0.348 0.320

Standard errors in parentheses
* p < 0.05, ** p < 0.01, *** p < 0.001
a significant at the 90% confidence level

On average, one in three (33.7 percent) registered voters will not likely, or will never be able to vote. Overall voter turnout in the Philippines, however, is considered high. The national mean voter turnout for the Philippines from 1967 to 2013, according to the Institute for Democracy and Electoral Assistance (IDEA), is 78.49 percent.

The first post-Marcos congressional election in 1987 saw the highest voter turnout at 90 percent. The national election during Arroyo’s succession presidency, on the other hand, has had the most dismal (63.68 percent). The national figures, however, do not reflect the political exclusions that take place in the countryside.

Curiously, for gubernatorial elections held between 1998 and 2013, the average number of effective candidates in the Philippines is only two. The provinces listed in Table 2 below had incumbents who competed for the gubernatorial seat virtually unopposed during these election years. Interestingly, most3 of the unopposed, winning candidates in these provinces happen to have previously occupied the gubernatorial seat as incumbents or succeeded a relative.4

Table 2. Gubernatorial candidates who won virtually unopposed, 1998-2013

Province Candidate Election year
Ilocos Norte Marcos, Ferdinand Jr. 2001
Ilocos Norte Marcos, Ferdinand Jr. 2004
Ilocos Norte Marcos, Maria Imelda R. 2013
Isabela Dy, Faustino Jr. 2001
Quirino Cua, Junie Evangelista 2013
Bataan Roman, Leonardo B. 2001
Bulacan Sy-Alvarado, Wilhelmino 2013
Oriental Mindoro Umali, Alfonso Jr. 2013
Albay Salceda, Jose Ma. Clemente 2013
Leyte Petilla, Carlos Jericho 2010
Leyte Petilla, Leopoldo 2013
Southern Leyte Lerias, Rosete Y. 2001
Zamboanga Del Sur Cerilles, Antonio Herrera 2013
Compostela Valley Uy, Arturo Tao 2010
Compostela Valley Uy, Arturo Tao 2013
Davao Del Norte Del Rosario, Rodolfo 2010
Davao Del Norte Del Rosario, Rodolfo 2013
Davao Oriental Malanyaon, Corazon 2013
Sarangani Solon, Steve Chiongbian 2013
Agusan Del Norte Amante-Matba, Ma. Angelica 2013
Dinagat Islands Villaroman, Geraldine Ecleo 2007
Apayao Laoat, Batara P. 1998
Basilan Wahab, akbar Muhtamad 2004
Lanao Del Sur Adiong, Mamintal Sr. 2004
Lanao Del Sur Adiong, Mamintal Jr. 2010
Maguindanao Ampatuan, Andal Salibo 2007
Tawi Tawi Sahali, Sadikul Adalla 2004

Source: Commission on Elections; Center for People Empowerment in Governance (CenPEG)

The fact that many of these candidates belong to well-entrenched political families is arguably incidental. We certainly do not intend to imply anything about competitive elections and political families. Furthermore, the estimates are crude and admittedly needs more refinements.

What is apparent is that it is us voters who ultimately bear the brunt of our electoral choices, as well as the lack of it. In many instances, votes are not waived but robbed from the people using violence or fraud.

For those who can exercise the right unhindered, the data hints that we are doing our least preferred candidate a favor whenever we waive our franchise.

In the absence of quality candidates, some may be of the impression it is better to boycott than choose between lesser evils. And that may just be what the least preferred candidate wants.


1 Adopting Golosov’s formula for computing effective number of parties, we can derive a similar index equal to Σ(pi/(pi + p12 – pi2)), where pi is the candidate’s vote share (votes received out of the total votes), p12 is the square of the share of the candidate with the highest vote, and pi2 is the square of the each candidate’s vote share.
2 Those interested to run their own estimates may request the dataset compiled by the Center for People Empowerment in Governance (CenPEG).
3 Except for Saranggani in 2013, Dinagat Island in 2007, Apayao in 1998, and Tawi Tawi in 2004.
4 A relative here is one who is related to the winning candidate by consanguinity or affinity up to the third degree.

Alicor Panao
Author: Alicor Panao
Rogelio Alicor L. Panao is an Assistant Professor in the Department of Political Science, College of Social Sciences and Philosophy, University of the Philippines Diliman. He holds an MA in Industrial Relations from UPD and an MA in Public Administration and Ph D in Political Science from the International Christian University, Tokyo. His research interests include quantitative analysis and quantitative research methods.