Polls, Politics and Predictions in the 14th General Election

Here's an excerpt from my upcoming book chapter on Media and GE14 (14th General Election). The gist of the chapter is basically in the final paragraph.

--

GE14 was reportedly to have scored a series of records in the history of Malaysian election, such as the first ever electoral win by the opposition, the oldest Prime Minister elected – and re-elected – into office, the first woman Deputy Prime Minister and the youngest cabinet minister. Apart from that, the other record was the rise of opinion polling and political forecasting “industry”.

Unlike the United States, where one can easily find dozens of opinion polling results and forecasts by the media and research institutions, the trend of conducting opinion polls during election picked up rather late in Malaysia or that some of the polling conducted were only meant for internal consumption, such as those prepared by government agencies and political parties.

According to Malaysia’s leading scholar in public opinion research, Syed Arabi Idid (2012), the first public opinion survey during election year was known to have started in 1986. He added that it was an initiative that was drawn mainly from the activities of local academics (Idid, 2012). Today, there are several institutions and academic units in Malaysian universities that conducts electoral research such as the University of Malaya’s Centre for Democracy and Elections (UMCEDEL) and the Electoral Studies Unit at the International Islamic University Malaysia (Idid, 2012). 

In addition to the university-based electoral research units, there are also other private and non-governmental public opinion firms. One of the leading private opinion research firms is Merdeka Center, which has been tracking the Malaysian public opinion since 2004 and other social and policy issues (Merdeka, 2018).

The last few elections have witnessed the rise of new players in the political forecasting and opinion polling “market”. Apart from Merdeka Center, among those which have made their forecast publicly available were the Malaysia-based think tank, Asian Strategy & Leadership Institute (ASLI), National Council of Professors (MPN), the Selangor state government-owned Institut Darul Ehsan (IDE), Invoke Malaysia, Ilham Centre, Politweet, KAJIDATA Research, CIMB Research and University of Malaya. 

Fascinatingly – or perhaps mysteriously – the global series of unfortunate polls, from the 2015 UK election, the Brexit referendum, to the US presidential election in 2016, seemed to have struck and “infected” Malaysia too. Academics and pollsters who researched and forecasted the 14^th General Elections must have been seriously taken aback by their missed predictions.

Out of ten polls and forecasts that were reported by the media, only two polls appeared to be favourable to Pakatan Harapan; namely, Invoke Malaysia (Surendran, 2018) and Ilham Center (Ruban, 2018). While most pundits and pollsters adopted a more cautionary tone in making an electoral forecast, Invoke Malaysia, which is led by PKR Vice President Rafizi Ramli, chose to be bold and vocal. Invoke Malaysia was the only one who had consistently forecasted a Pakatan Harapan win way back since January 2017 (Lim I. , 2018). 

Apart from research firms, one individual that is worth mentioning is former Finance Minister, Tun Daim Zainuddin. After correctly predicting the outcomes of the previous two elections, GE12 and GE13, he was again spot on when he predicted that Pakatan Harapan will win GE14 (Hilmy, 2018). The other pollsters from the established outfit like Merdeka Center, state-owned Institut Darul Ehsan (IDE) and University of Malaya, had all predicted a BN win in GE14 (Malaysiakini, 2018; Sulong, 2018; Edward, 2018).

So how did 80 percent of the predictions get it wrong? There were several factors. To many pollsters, the electoral redelineation was the single biggest factor that would increase BN’s chance in GE14 (Zurairi, 2018). Professor Dr Syed Arabi Idid (2018) had predicted that, with the redelineation, at best the result would be a hung parliament. The other factors listed were the reluctance of the survey respondents to share their sentiment, particularly the Malays. According to Ibrahim Suffian, “A lot of people don’t want to talk. They can talk about the issues, but when it comes to their choice, they don’t want to talk,”.

Apart from local firms and institutions, the intensity of GE14 had also attracted a strong interest among foreign media, investors and observers. Globally, the financial markets were all but unanimous in their view of preferring the political status quo and stability than political change and uncertainty (Bloomberg, 2018). Despite the growing competitiveness in the election, the prevailing negative sentiments and a host of other political issues that plagued Najib Tun Razak and the BN, Edwin Gutierrez, head of a London-based investment firm, told Bloomberg that it was almost “the overwhelming consensus from the market” that Najib Tun Razak would win the election (Liau & Shukry, 2018).

“I expect this (GE14) to be a non-event with Najib coming away with a resounding victory,”

Edwin Gutierrez, Aberdeen Standard Investments.

Likewise, the London-based international daily newspaper, Financial Times (FT) and Japanese financial company, Nomura Group too had predicted a BN win (Star, 2018; Rosli, 2018). The prevailing negative sentiment towards the economy and politics were well documented in the FT survey, which was done across several states in the Malaysian peninsula. However, it was concluded that the recent redelineation of electoral boundaries in several parliamentary seats was thought to be one of the major factors that would increase BN’s chance of winning (Star, 2018). 

Alas, like a global pandemic of missed predictions, they got it all wrong. So, what should pollsters and researchers do to solve this issue? Is there a better public opinion methodology? Can internet and social media data be the new solution?