Electoral conflict and the maturity of local democracy in Indonesia: testing the modernisation hypothesis

Abstract

This paper constructs an electoral hostility index for 282 local direct elections (PILKADA) of district heads during 2005–2007 and examines the socio-economic determinants of local democratic maturity in Indonesia. There are 67 PILKADAs (out of 282) categorised as having medium, high or very high levels of electoral hostility. The picture is dominated by hostilities directed towards the local elections commission after voting day. The large sample quantitative analysis employs ordered logistic regression. The results show some evidence in support of the modernisation hypothesis in the context of Indonesia's local democracies. Higher PILKADA hostility or less mature local democracy tends to be experienced by districts with lower income, higher poverty incidence and less urbanised. The results also imply that democracy cannot be deepened in the absence of economic development.

Keywords:

• electoral conflict
• Indonesia
• local democracy
• modernisation hypothesis

JEL classifications:

• P16
• O10

Acknowledgements

The author is grateful to Anis Chowdhury and two anonymous referees for helpful and constructive feedback; however, any remaining errors are mine.

Notes

1. The term of PILKADA that stands for Pemilihan Kepala Daerah (head of region election) was widely used during 2009–2010. It refers to the elections of both district heads as well as provincial heads. However, a new terminology of PEMILUKADA (Pemilihan Umum Kepala Daerah – general election of region head) has been introduced since late 2010. I use the PILKADA terminology throughout this paper.

2. This is not to suggest that the national elections were entirely free of violence. In fact there were some. During the campaign period of the 1999 elections, several clashes occurred between the Islamist party of PPP and the Islamist-inclusive party of PKB in Central Java: the most serious one was the clash in Jepara district that claimed five lives. When Megawati Soekarnoputri, the chairman of the winning party PDIP, failed in her bid to be elected as the president by the MPR in October 1999, riots broke out in Bali and Solo that killed one, burned dozens of buildings and hundreds of vehicles (Tadjoeddin 2002). Although minor violence incidents also occurred in the 2004 and 2009 national elections, especially during the campaign periods, they are negligible compared with incidents of electoral violence during 2005 and 2010 rounds of local direct elections.

3. The Jakarta Post, 4 August 2011.

4. DPD has no legislative function; however, it may propose legislations relating to regional development to the Indonesian Parliament (DPR – Dewan Perwakilan Rakyat). Both DPR and DPD form the Indonesian People's Consultative Assembly (MPR – Majelis Permusyawaratan Rakyat).

5. Fragmentation refers to the division of an entity into many fragments, while complete polarization occurs when an entity is divided into two segments of equal size.

6. The chairman of Nahdatul Ulama Hasyim Muzadi called for the abolition of direct elections for local executives and a return to the old system where heads of district or province were elected by local parliaments (Kompas, 28 January 2008). The chairman of Muhammadiyah Din Syamsuddin suggested that local direct elections should be seriously re-evaluated with a view to finding a better solution (Suara Karya, 3 March 2008).

7. This was voiced by the chairman of DPD in his official speech at a plenary session of DPD on 22 August 2008 (http://www.tempo.co.id/hg/nasional/2008/08/22/brk,20080822-131863,id.html, accessed on 26 April 2009).

8. Comment made by Professor Ryaas Rasyid, a key designer of Indonesian decentralization laws and a former Indonesian minister for regional autonomy (Kompas, 5 September 2009).

9. Kompas, 10 December 2010. (http://cetak.kompas.com/read/2010/12/10/04550653/gubernur.dipilih.dprd, accessed on 20 December 2010).

10. Among others, some key studies are Londregan and Poole (1996), Przeworski and Limongi (1997), Przeworski et al. (2000), Barro (1997, 1999), Boix and Stokes (2003), Epstein et al. (2006), Glaeser et al. (2004), Acemoglu et al. (2009) and Papaioannou and Siourounis (2008).

11. Patronage politics are not limited to immature democracies, as they are also frequently practiced in advanced democracies. However, in Indonesia, it has been argued that the legacies of the old regime have survived and succeeded in reinserting themselves in the newly set-up democratic political system (Robison and Hadiz 2004).

12. The military was not ordered to brutally crush student demonstrations, as one observed in the Tiananmen Square in 1989.

13. The South Africa path is one of four scenarios of dictatorship and democracy offered by Acemoglu and Robinson (2006). The other three are the UK, which gradually moved toward democracy without significant resistance from the elites; Argentina, where the elites strongly resisted democratic demands, resulting in a long period of instability moving from dictatorship to democracy and vice-versa; and Singapore, which ended up with a stable and happy anocracy.

14. The economic crisis served as both a catalyst and a trigger, in line with Haggard and Kaufman's (1995, p. 26) observation that ‘the probability of a democratic transition increases during periods of economic distress'. In the case of Indonesia, this tendency was noted by McBeth (1999, p. 22), who wrote, ‘without the collapse of the economy … there would not have been the opportunity for political change'. It would have been difficult to predict the fall of Suharto as the first step to transition while the nation was experiencing high, stable economic growth.

15. Suharto rose to power in the mode as, and roughly at the same time as, Zaire's Mobutu (1965) and Chilli's Pinochet (1973). They were military generals, anti-communist and autocratically ruled their countries with strong supports from the West. Since the collapse of Communism in late 1980s, these dictators were simply no longer needed.

16. Again, as footnote 2 has explained, it does not mean that the national level elections were totally free of violence.

17. We have to admit that this is a rather optimistic view of Indonesia's democratisation, which is in accordance with Diamond's (2010) assessment on a decade of Indonesia's democracy in a comparative perspective at global level and in line with MacIntyre and Ramage's (2008) view that sees Indonesia now as a normal country. A less optimistic view, for example, argues that the democratic change in Indonesia has been superficial by which the oligarchic elites who controlled the Suharto's New Order have survived the 1998/99 regime change and continue to use the state for rent-seeking objectives (Boudreau 2009, Robison and Hadiz 2004). For comprehensive analysis of a decade of political reform in Indonesia, see Aspinall and Mietzner (2010), Crouch (2010) and Mietzner (2009a).

18. This is not to reduce the failure of the democratic experiment in the 1950s simply because of Indonesia's lack of socio-economic development at that time. Democracy functioned reasonably well until 1956, when non-democratic forces (President Sukarno, the Indonesian Communist Party – PKI and the military) started to undermine parliamentarism. It could be argued that democracy could have survived without the organized and coordinated pressures from these non-democratic actors (Mietzner 2009a). In line with the modernisation theory, Bhakti (2004) argued that the failure was the result of a lack of adequate institutional backup for democracy: a lack of education, a lack of democratic culture and an insufficient economic base.

19. By providing legitimacy, development and diversification reduced political risks. By contrast, Mobutu of Congo did not opt for development and diversification since he faced a high degree of societal opposition, which led him to believe that investments in infrastructure and other public goods would pose a threat to his grip on political power.

20. During Suharto's rule (1966–1998), provincial and district heads were nominally elected by the local (provincial and district) parliaments, but they were essentially Suharto's choices. After the fall of Suharto in 1998, Jakarta virtually lost control over local elections and regional heads were purely elected by local parliaments. And since 2005 they have been elected through popular votes of direct local elections.

21. Erb and Sulistyanto (2009) is the best collection of case studies on the direct local elections covering wide range of issues such as political parties, local elites, corruption and money politics, media, ethnic politics, women, conflict, etc. Other case studies, among others, include Buehler (2007), Buehler and Tan (2007), Choi (2007), ICG (2010, 2009, 2008, 2006), Mietzner (2007, 2009b), Nurhasim (2009), Palmer (2010).

22. The 1999 free and fair multiparty elections marked Indonesia's second entrance to democracy.

23. This is very similar to Brancati's (2006) approach in constructing an ordinal scale for determining the severity of ethnic conflict in cross-country settings, based on the Minority at Risk (MAR) dataset.

24. They are KOMPAS, Koran TEMPO, Media Indonesia, Republika, Sinar Harapa, Suara Karya, Suara Pembaruan and The Jakarta Post. I thank Teguh Yudo Wicaksono, who facilitated my data collection at CSIS.

25. Jawa Pos is the biggest newspaper company in Indonesia and owns local newspapers in almost every province.

26. The provinces are Central Java, East Java, Yogyakarta, Aceh, Riau, West Sumatra and East Kalimantan. The three Javanese provinces were deliberately visited to read the PILKADA-related archives of Suara Merdeka (Central Java), Jawa Pos (East Java) and Kedaulatan Rakyat (Yogyakarta). A similar thing was done in East Kalimantan (Kaltim Post), Riau (Riau Pos) and West Sumatra (Singgalang). The cross-check on local newspaper reports for Aceh was made possible by the availability of the UNDP's Aceh Peace and Development Monitoring database and The World Bank's Aceh Conflict Monitoring Update.

27. In the follow up study, I plan to collect such data directly from provincial newspaper archives in the capital cities of Papua and West Papua provinces.

28. See Appendix 1 for the detail list of district PILKADAs belong to the categories of medium, high and very high hostilities.

29. It should be noted that this is not to deny the instrumental role of ethnic and religious factors for political mobilisation during local elections; however they do not serve as the cleavage by which the hostility takes place. It can be seen that the cleavage and the main reason for the hostility are mainly due to fierce electoral competition and dissatisfaction towards electoral

management.

30. Although two district PILKADAs in North Maluku province are categorised as very high hostility PILKADAs and four district PILKADAs in Maluku province fall within the category of high hostility PILKADAs, they seem to be not driven by Christian–Muslim religious divide that fuelled inter-communal violence in the region during the early phase of democratic transition. In fact, a mixed pair of Christian and Muslim for local leadership (governor and its deputy; or regent and its deputy) has become a newly adopted local norm in local direct elections in Maluku; see Tadjoeddin (2010).

31. Five district PILKADAs in West and Central Kalimantan are categorised as high hostility PILKADAs; however the hostilities were not related to the anti-ethnic Madurese sentiment that fuelled inter-communal violence in the two provinces’ several localities during the early phase of democratisation.

32. In Poso, which experienced a series of bloody Christian–Muslim clashes between 1998 and 2004, there were several cases of bomb explosions directed at a political candidate, approaching the voting day of the district PILKADA in June 2005. However, the attacks were not related to religious divide. They can be seen as the utilisation of a specific skill acquired in the period of ethnic violence.

33. Author's calculation based on short descriptions of violent incidents in the twenty district PILKADAs in ICG (2010, pp. 23–24). It should be noted that ICG (2010) is not a systematic large sample study, it provides case studies of four district PILKADAs in 2010 (Mojokerto, Tana Toraja, Toli Toli and Poso). A proper comparison should be based on the similar methodology applied to the second cycle of district PILKADAs starting in 2010.

34. It should be also noted that the relative size of district government revenue might not be a good proxy of elite competition. Another alternative proxy for elite competition is to look at electoral competition itself, which can be measured by the margin of victory between the winner and the runner up. Smaller margin means higher competition. We could not exercise this due to data limitation; however, this can be a direction for further research.

35. Before 2007, the district head or the provincial governor proposed candidates for region's respective KPUD commissioners, then the provincial KPUD or national KPU selected the definitive commissioners of the respective KPUD (Law no 12/2003). The mechanism was revised in 2007 (Law no 22/2007). National KPU and provincial KPUD form a selection committee in each of province and district, respectively. The committee members must be representative with a set of minimum qualifications. KPUD commissioners’ candidates are chosen through an open recruitment process with written exams, health checks and psychological tests. The selection committee propose candidates for respective KPUD commissioners and then the national KPU and provincial KPUD select definitive commissioners of respective KPUD through fit and proper tests. The latter mechanism is arguably better than the previous one.

36. As Sen (1999) points out, it would be wrong to see democracy as the end product of a largely economic process. He argues that it is wrong to ask if a country is ‘fit for democracy’: the correct way to look at the issue of economic and social development is to understand that a country becomes ‘fit through democracy.'