Thailand’s snap election of 2 February 2014 has deepened the rifts dividing the country along socioeconomic, ideological, regional and ethno-linguistic lines. The Democrat Party boycott and the People’s Democratic Reform Committee (PDRC)’s disruption campaign raise the question of whether elections are at all suitable to show a way out of Thailand’s protracted transformation crisis. Through the prolonged ‘Bangkok shutdown’ and ‘Bangkok picnic’ campaigns, the PDRC hindered candidate registration, advanced voting and election-day voting. According to information released by the Election Commission of Thailand, 10.8% of polling stations nationwide (10,139 of a total of 93,952) were not operational on election-day. In nine out of Thailand’s 77 provinces no voting was possible at all. In nine other provinces – including Bangkok – voting was partially disrupted. It is unclear, exactly how many of Thailand’s 48 million eligible voters were disenfranchised, not least because election results – other than turnout – were never released.
Previous scholarship has found elections in Thailand to be increasingly ‘clean’, despite some persisting problems of money politics. Partisan identities are on the rise, and elections have been described as the way to usher politics away from street movements and back into the constitutional arena. Yet, the electorate is starkly divided on the very issue of elections, with one side regarding elections as the legitimate tool to express their numerical majority and as largely clean, and the other side viewing them as an inherently flawed ‘electocracy’ and the failure of any upcoming polls – were they to be held – as a foregone conclusion.
With the possible July 2014 date for a new election in question again, and the government under increasing pressure from the courts, an independent evaluation of the conduct of the February 2014 election could help clear up some of the contested arguments.
How to assess the quality of elections?
There are different ways to measure the integrity of elections: First, public opinion surveys are frequently used to assess citizens’ trust in elections – two examples are the World Value Survey or the Asian Barometer. They are great tools to gauge mass perceptions, but they are expensive and colored by partisanship of the respondents – winners usually evaluate an election much more positively than those who supported the losing side. Second, election monitoring groups ‘grade’ elections and give an assessment of whether they are ‘free and fair’. But one often comes across contradictory statements by different monitoring groups. In the case of Cambodia for example, the civil society coalition Electoral Reform Alliance (ERA) evaluated the election rather negatively, while observers from the International Conference of Asian Political Parties (ICAPP) – invited by the government – gave it a clean bill of health. So, who should one listen to? Furthermore, none of these existing measurements are very fine-grained. They often give the false impression of a good/bad dichotomy. But elections have obviously many different aspects around the whole electoral cycle.
Pooling expert knowledge is a common practice when dealing with difficult and controversial issues where other sources of comparative evidence are lacking. This is the method that Transparency International uses to compile its well-known Perception of Corruption Index. The Electoral Integrity Project (EIP) – based at the University of Sydney and Harvard University, funded mainly by the Australian Research Council – uses a similar expert survey methodology to compare to what degree countries around the world meet international standards of elections.
The EIP’s annual report ‘The Year in Elections, 2013’ evaluates all national parliamentary and presidential contests occurring in countries with a population higher than 100,000 from July 1, 2012, to Dec. 31, 2013. The coverage for this first annual report is 73 elections in 66 countries. Data is derived from a global survey of 855 election experts who have knowledge about elections in their country of expertise. All survey results are published biannually through the project’s Dataverse .
Electoral integrity in Thailand
This measurement instrument was used to gain an independent assessment of the integrity of Thailand’s 2 February 2014 election. Survey requests for Thailand were sent out on 11 March 2014. Out of 40 requests sent, 14 experts completed the survey (response rate of 35%), eight of which were domestic, and six international. An expert in the sense of this survey is a ‘social scientist who, through his/her work, had demonstrated knowledge of the electoral process in a particular country’ (see Norris, Frank & Martinez i Coma 2013). ‘Demonstrated knowledge’ refers to academic or policy-related publications on elections.
Respondents were asked to assess the quality of the election based on 49 indicators. The responses are then clustered into eleven sub-dimensions occurring during the electoral cycle and summed to construct an overall 100-point expert Perception of Electoral Integrity (PEI) index and ranking. Missing values were substituted.
The experts evaluated the 2 February 2014 election with an overall PEI score of 61 – which places this contest firmly at the global average of electoral integrity (average PEI score of 63.8 for all 73 contests assessed in the 2012/2013 annual report). Yet, the disaggregation of the 11 sub-dimensions of electoral integrity reveals some surprising findings, as Figure 1 shows.
(Click on image for larger version.)
A flawed legal framework and fraudulent elections?
Contrary to the belief that Thailand’s electoral laws need to be reformed and electoral fraud needs to be eradicated first, the experts evaluated the election rather positively in these dimensions. The electoral laws were assessed well above the global average (81 compared to the global mean of 62). In particular, the experts judged that the legal framework did not favor the governing party nor did it disadvantage smaller parties or restrict citizens’ rights.
At the same time, the overall score for the voting process closely matched the global average (Thailand: 64; global mean: 63). This dimensions encompassed for example whether fraudulent votes were cast. Yet most items in this sub-dimensions are about more technical aspects of the voting process such as whether expatriates could vote, whether postal ballots were available, or whether the voting process was easy. This suggests that Thailand conducted an election of average quality in terms of the voters’ election day experience. In fact, it did not differ from the global average in a statistically significant way. This is all the more surprising, since this sub-dimension included the item ‘some voters were threatened with violence at the polls’ (see below).
The same is true for the sub-dimension of campaign finance – which included items on whether ‘rich people buy elections’ or ‘state resources were improperly used for campaigning’. Thailand’s score is slightly above-average, (Thailand 60; global mean: 49). In a global comparison, campaign finance is the most problematic dimension of electoral integrity, yet money politics seem to be somewhat less of a problem in Thailand than elsewhere. It must be noted however, that this difference in means is not significant in a statistical sense.
The experts’ assessments in these three dimensions thus suggest that the discourse of routinely stolen elections, a skewed legal framework and the huge impact of vote-buying is exaggerated. Although not in the top tier of countries, Thailand compares favorably with the global average on these issues.
The electoral authorities
On the other hand, huge institutional problems become apparent in the experts’ evaluation of the electoral authorities. Given the Election Commission of Thailand’s (ECT) refusal to publish election results, it is not surprising that Thailand achieves its lowest score in the sub-dimension of results (42 compared to a global mean of 71). This is in fact the lowest score of all countries surveyed in the PEI to date.
Similarly poor are the assessment for vote count (Thailand: 62; global mean: 74) and electoral authorities (Thailand: 46; global mean: 67). The latter encompasses questions about the impartiality, professionalism and performance of the ECT. The low score likely speaks to the ECT’s failure to ensure safe and convenient voting, as well as their non-committal or even adversarial stance towards the election date. In a regional comparison, the Thai election management body ranks about the same as its much-criticized Malaysian (‘electoral authorities’ score 46) and Cambodian (score 42) counterparts. This finding is in line with Thailand’s poor performance in the electoral procedures sub-dimension (Thailand: 54; global mean: 72), which asks such things as ‘are elections well managed?’ or ‘are election officials fair?’ All deviations from the global average in these four sub-dimensions are statistically significant.
The emerging pattern seems to suggest that although there is comparatively less bias towards the ruling party in terms of electoral laws, campaign finance or election fraud, the quality of the Thai election was significantly reduced by the electoral authorities’ failure to conduct the election in a professional and impartial manner. The danger in this regard is that such failures are likely to erode public trust and the legitimacy of elections even among those societal segments that support elections as an institution.
Violence and media bias
Disaggregating the above findings further allows to ‘zoom in’ on the parts of the electoral cycle in which Thailand performed best and those where the worst problems were noticed. Figure 2 shows 48 of the 49 PEI survey items and how Thailand scored relative to the world average. Survey items were sometimes phrased positively and sometimes negatively. Hence, Figure 2 splits the results in two graphs that show the deviation from the global mean in percent. Deviations from the global average that signal a lesser quality of the election are shaded in red, while deviations that signify a better-than-average quality are shown in green. In addition, survey items are ordered from best to worst performance, breaking up the ordering along the electoral cycle.
(Click on image for larger version.)
The figure reiterates the previous findings. Several items on the performance of the electoral authorities have significantly lower scores than the global average. The figure further highlights the detrimental impact of the PDRC’s violent campaign to hinder the polls. The survey items ‘some voters were threatened with violence at the polls’ and ‘the elections triggered violent protest’ are among those indicators that clearly drag Thailand’s overall score down, with 55% and 64% worse than the global average, respectively. This places Thailand in the vicinity of much more violent elections such as Pakistan or Bangladesh in terms of violent intimidation.
Another low point is the 45%-worse-than-average rating on how candidates are selected. It has long been pointed out that Thailand lags behind in party institutionalization, since candidates are selected by top party leaders instead of primaries or caucuses.
On the other hand, Thailand’s election was evaluated much better than the global average on issues of voting district boundaries, electoral laws and media bias. A highly better-than-average deviation from the global mean was recorded on the item ‘TV news favored the governing party’. This may in fact highlight the limitations of some of the PEI survey items, which implicitly are geared toward malpractices committed by the government. As the other media indicators show, the media landscape was in fact tilted against the governing party, a fact likely reflected in the exceptionally good score on this survey item.
Elections not fraudulent, but facing stark opposition
In sum, the results of the PEI survey suggest that elections in Thailand are by and large well-administered routine affairs in terms of the technical aspects, and that the existing legal framework does not significantly favor incumbents. Election fraud, vote buying or other forms of manipulation seem to be less problematic and on par with other countries. The predominant problems in the 2 February 2014 election were rooted first and foremost in the conduct of the opposition street movement and its campaign to make voting impossible – including violent intimidation of voters. And second, in the conduct of the electoral authorities and their fainthearted enforcement of the mandate to hold elections, the failure to ensure safe voting, and negligence in the aggregation and announcement of results.
It appears that elections as a mechanism for the legitimate transfer of political power are at the same time deeply rooted and fragile in Thailand. The capacities to hold elections with integrity are clearly there, and some of the worst problems of fraud or a skewed playing field as seen in other countries are not present in Thailand. Yet, it is highly worrying that the election management body itself is evaluated so negatively.
Several vignettes asked in the conclusion of the survey are meant to weight experts’ answers based on how serious they regard three hypothetical electoral problems. The hypothetical scenario of an opposition boycott was regarded as ‘very serious’ (score 8-10 out of 10) by nine of the 14 surveyed experts. This result strongly suggests that the opposition movement needs to re-enter the electoral arena in order to restore trust in elections and to avert an escalation of the country’s deep crisis.
Naturally, the question must be asked whether the experts’ perceptions are also reflected in mass perceptions of electoral integrity. Furthermore, there is reason to believe that in such a deeply divided society as Thailand, popular perceptions of election quality are more colored by partisanship than elsewhere. Future research by the author based on the Asian Barometer and the World Values Survey will address these questions and also assess the impact of media consumption on public perceptions of election quality.
Max Grömping is a PhD candidate with the Electoral Integrity Project at the University of Sydney. His current research focuses on the role of social media and domestic observers on electoral integrity. Prior to this, he lectured at Thammasat University, Thailand. Contact: max.groemping[at]sydney.edu.au Twitter: @MaxGroemping
 Note that for the annual report multiple imputation was used to substitute missing values. Yet, for the discussion of the Thai results presented in this article, mean substitution was used to fill in missing values.
 The differences in means reported here are statistically significant at the 95% level in a two-tailed t-test for equality of means, except where noted otherwise.
 The survey item on internet voting is dropped, because of non-applicability in the Thai case.