Testing for Incumbency Advantages in a Developing Democracy: Elections for Local Government Leaders in Indonesia.

• Author: Schraufnagel, Testriono

One of the most elementary facts of political life in mature or consolidated democracies is that incumbent politicians--in legislative and executive elections at all levels of government--have an advantage when they run for re-election. (1) By convention, scholars use the term "incumbency advantage" to refer to the electoral edge enjoyed by politicians who are in-office. (2) Scott Ashworth, Ethan Bueno de Mesquita and Amanda Friedenberg, for instance, note that "information-based" incumbency advantage occurs as voters often choose incumbents over challengers as a risk aversion strategy. (3) In such cases, the voter presumably imagines "better the politician I know than the one I don't know". Incumbents may also have a track record of policy-making they can highlight in the election campaign to prove their worthiness. Moreover, they often engage in tactics to scare off potential serious challengers by amassing significant re-election resources. (4)

Notably, incumbency advantages can compromise the legitimacy of elections when taken to an extreme. For instance, if an incumbent lawmaker makes use of election law "reform" that entails elements of ballot rigging, democratic accountability is certainly compromised. Yet, it is also the case that elections can hold incumbents, and the governments they serve, accountable. In legitimate elections, the connection between government performance and support for incumbents should be strong. (5) According to this retrospective framework, elections can or should be referendums on the performance and competence of the party in power and its incumbent politicians. (6) Citizens can either incentivize politicians by sanctioning poor performance or by rewarding good performance. (7) This ex post facto scrutiny, when it is working, will often be based on the contemporaneous performance of the economy, with macro-economic clues providing the greatest explanatory power. (8)

Much of the work described above is quite well developed and regarded. What we know less about is whether incumbency advantages, and elections as accountability mechanisms, is occurring in newer democracies. Yet, attention to this matter has been changing, and Indonesia has been receiving some of this recent scrutiny. (9) At issue in these works is whether elections in the world's third largest democracy are legitimate. The concern is that electoral processes are not sufficiently institutionalized in Indonesia and that elections, in some manner or another, may be compromised by "personality politics", (10) patronage networks, (11) "money politics", (12) or the phenomenon of "dynastic politics". (13) The genuine apprehension is that these other considerations, more than incumbent government performance, are the primary determinants of election outcomes.

To investigate questions about elections as accountability mechanisms in newer democracies, we also turn to Indonesia. Specifically, we study elections for district heads and city mayors in the 2015 elections. Indonesian legislative elections have been previously studied, and researchers either offer qualified support for an incumbency advantage or find it necessary to simply conclude with a whole host of new questions. (14) We believe that executive elections hold more promise as an avenue for studying incumbent accountability because, arguably, there is a more direct public knowledge or awareness of the performance of executive leaders. While these local executives in Indonesia clearly do not have perfect sovereignty over policy adoption and implementation, one might expect that their higher profile and status will make the voting public more likely to hold them accountable for socio-political and economic outcomes which occur while they are in office.

Importantly, by 2015, it had been sixteen years since considerable devolution of political and economic power had taken place in Indonesia. Laws No. 22 and 25, passed in 1999, transferred functions, personnel and assets from the central government to provincial and district governments. These laws were expressly intended to increase local economic capabilities and capacities. (15) This increased power of local governments and their elected leaders arguably provides a suitable opportunity to test for democratic accountability, as manifest in the results of local elections. (16)

To be clear, the central question addressed in this article is: which factors contribute to the probability that incumbent city mayors and district heads in Indonesia get re-elected? In other words, it seeks to test whether the well-known incumbency advantage, which occurs in developed democracies, is a significant factor in a newer democracy. We are also interested to learn more about democratic consolidation in Indonesia specifically. Indonesia, a country with a large and growing economy, is already an important player in regional geopolitics. Gaining a better understanding of democratic consolidation and electoral legitimacy in Indonesia is thus an important undertaking.

This article will join a growing body of literature on subnational elections, which tests for relationships between local government expenditures and incumbent success while controlling for political factors and the individual-level characteristics of candidates. (17) Our specific tests will focus on whether city and district leaders perform better in their re-election contests when their time in office is characterized by more government spending in the constituencies they serve. We examine both re-election rates and the vote percentages obtained by the incumbent leaders. Our results show that subnational government leaders' electoral success is empirically tied to local spending. And even though much education spending is controlled by the national government, (18) we find that spending in this policy arena is especially closely linked to incumbent leader vote margins, after controlling for a host of other possible explanations.

Local Elections in Indonesia

Prior to 1998, Indonesia held national party-based elections and subnational government heads were handpicked by local legislative leaders based on their loyalty to the government of President Suharto, whose New Order regime ruled Indonesia from 1967 to 1998. Elections in this earlier era were largely illegitimate, and civilian leadership positions all over the country, including mayors and governors, were mostly filled by members of the military. (19) While concerns over corruption and state-sponsored illegalities have persisted in the post-New Order era, (20) Indonesia has undeniably undergone considerable political reform since 1998, with the country's electoral climate in 2015 being regarded by Freedom House as "partly free"--an indication that elections that year might be reasonably scrutinized using conventional modelling developed in studies of Western democracies. (21)

The People's Consultative Assembly, the national legislature of Indonesia, has amended the 1945 Constitution several times in the contemporary period to strengthen vertical accountability mechanisms by further devolving political power to the provinces, districts and cities. Based on the amended Constitution, and as a response to growing public pressure, Law No. 32 was passed in 2004 to establish the direct election of district heads and city mayors. (22) Under the new electoral arrangement, there have been direct elections for these local leaders every five years since 2005. (23)

The first round of direct elections occurred only in the constituencies where the tenure of the indirectly elected local executive was ending. (24) In all, 182 locales conducted direct elections for their subnational government leader in 2005. One important development to note--which unfolded at the same time as the move towards direct local elections--was the creation and proliferation of new districts in Indonesia. Beginning in 2001, district splits took place outside of Java, in larger areas, as well as districts that had significant natural resource endowments. From just over 300 local constituencies in 2001, the number increased to 434 between 2001 and 2007, and by 2016 had reached 514, where it stands today. Notably, the six cities in Jakarta province do not have direct elections for local leaders. This proliferation of constituencies has resulted in considerable staggering in direct elections of district heads and city mayors.

Figure 1 displays the number of local elections since 2005, and also the number of incumbents who sought and achieved re-election. We must concede that there are some discrepancies reported in government, scholarly and media accounts of these elections. However, we proceed using information from all available resources to produce what we believe is the most accurate accounting possible. It should be noted that not all incumbent local leaders sought re-election: some were completing their constitutionally limited second five-year term, while others sought higher office or retired. Consistent with research on electoral politics in consolidated democracies, (25) some local leaders strategically retired amid fears of electoral retribution and, in some cases, ongoing criminal investigations. (26)

Figure 1 The Election of District Leaders in Indonesia: 2005-15 The Number of Direct Direct Local Elections Incumbent Local Elections with Incumbent Winning Involvement 2005 182 137 91 2006-2008 219 164 120 2010 215 154 106 2011-2013 235 161 100 2015 243 188 126 Note: Table made from bar graph. The data presented in Figure 1 reveals incumbent re-election frequencies in five recent election periods. During the entire period, incumbents won 67.5 per cent of the 804 elections that had an incumbent running. In the first round of direct elections in 2005, for example, incumbents won 66 per cent of the 137 races that had a leader running for re-election. Meanwhile, in the 2015 elections, incumbents ran in 188 out of the 243 contests, and won...