Village head elections in Java: Money politics and brokerage in the remaking of Indonesia's rural elite

Abstract

To explore how democratisation is transforming Indonesia's rural elite, we examine two village head elections in Central Java. Despite the competitiveness of these elections, the campaigning modes employed by candidates, especially vote buying, points to elite continuity, because only wealthy villagers can compete for office. Moreover, links with higher state officials remain important for village elites, allowing them to win political support by obtaining projects from local government. However, rather than being incorporated as subordinates in a bureaucratic hierarchy as during authoritarian rule, village elites are now true rural brokers, exercising considerable leverage in their relations with the state.

Since the fall of the Suharto regime in 1998, government in Indonesia has been reshaped by the twin processes of democratisation and decentralisation. In place of the highly manipulated elections and entrenched bureaucratic and army power that characterised the Suharto era (1966–98), free elections are now used to fill key government positions at national and local levels. Centralised authority has made way for devolution, with extensive political and budgetary power passing from national ministries to about 500 district governments around the country. While scholars have devoted much attention to understanding democratisation and devolution at the national, provincial and district levels, ¹ relatively little literature focuses on how these changes have transformed power relationships at the lowest level of government in the rural areas: in Indonesia's approximately 80,000 villages.

As with other levels of government, the administration of Indonesian villages has been transformed. Already a feature of rural politics under Suharto, elections of village heads have since become more democratic. New modes of participatory budgeting and planning have been introduced. Village heads and other village officials have become important in Indonesia's emerging welfare regime, helping to identify recipients and distribute benefits. A new Village Law passed in 2014 (Law No. 6 of 2014) extends decentralisation by expanding village autonomy. The law provides greatly increased village budgets, injecting an additional 21 trillion rupiah (about US$1.5 billion) into village budgets in 2015 alone, ² potentially significantly reducing their dependence on higher levels of government.

Despite the importance of these changes, the literature on village-level political and social relations does not match the richness of that on decentralisation and democratisation at higher levels. In particular, while a wealth of case study research throws light on elections of regional government heads and legislators, ³ relatively little has been written on village elections. Accordingly, in this article, our research has been guided by two lines of enquiry. First, we seek to extend the study of Indonesia's political transformation by investigating the impact of democratisation on village head elections, and to identify the modes of campaigning that candidates use. By mapping this empirical terrain we hope, second, to examine how democracy is transforming village elites and their interactions with both ordinary villagers and supravillage structures of power. To answer these questions, we chose to examine two village head elections in neighbouring villages in a district in the north of Central Java province.

Our analysis proceeds as follows. After a brief survey of village political structures in the Suharto and reformasi eras, we present our case studies, analysing what they tell us about the pursuit and distribution of village power. We find a microcosm of the political dynamics higher up. The breakdown of centralised mechanisms of control has opened space for sometimes unruly political contestation in the villages. However, as has been extensively argued with regard to national, provincial and district politics, the forces best positioned to take advantage of this new openness are local-level elites, which in the villages typically means members of established families who have long dominated administrative positions and were the wealthiest landowners — in some cases for several generations. However, these established elites have lost their former monopoly on village power.

Moreover, in order to maintain their hold on village politics, elites have to invest major financial and political resources in electoral competition. In particular, we devote much attention to what in Indonesia is known as ‘money politics’ (politik uang) — the distribution of cash and other gifts to voters. At higher levels, money politics has been seen as a mode by which elites nurtured under authoritarianism have entrenched themselves within the new democratic order. ⁴ Anecdotal evidence suggests that vote buying is more intensive in village head elections than elsewhere, and our case studies confirm this picture. In both our villages, all candidates engaged in mass vote buying and other forms of clientelistic politics. This money politics was one obvious sign of elite entrenchment in the village, given that only wealthy villagers, typically well connected to supravillage bureaucratic and business networks, could afford to compete.

However, we argue that money politics does not signify closure of democratic space. On the contrary, it points to competiveness, albeit within a restricted scope. The tools of coercion and threat that were available to village elites in the New Order did not feature in the elections we studied. Instead, contestants felt under great pressure from popular expectations to distribute handouts and favours. Moreover, as with elections for higher positions, village head elections are not only about money. Rather than facing a populace which is uniformly deferential to their social betters and in thrall to clientelistic bonds, candidates must deal with villagers who are often highly cynical about village elites. Though cash and other gifts are important, voters also evaluate candidates by looking at their other qualities, including approachability, probity, development achievements and mediation skills. In short, even where vote buying is deeply entrenched in village political culture, as in our case study locations, there is space for more populist modes and for electoral give and take between ordinary villages and elites. Overall, village elections are part of a wider shift from coercive to economic power as the main ordering principle in political relations.

Finally, we address the question of what this reordering of village democracy tells us about the changing nature of rural elites and their integration into supravillage structures of power. During the New Order period, studies of rural power relations showed that village notables were consolidating positions of local privilege partly by integrating themselves into state-centred bureaucratic structures and patronage networks. Our findings suggest that this integration of rural elites into external structures of power is continuing, even accelerating. Though most (but not all) of the candidates in both elections came from established elite families in their villages, they also all drew on sources of wealth and authority that came from links forged outside them. Ties to higher state officials, and access to patronage, remain critical.

However, the mode of incorporation has changed. Under the New Order, village elites were being integrated from the top down and, though they could gain material advantage from interactions with higher authorities, they had little bargaining power in these relationships. Now, relations with higher officials are fluid, and more evenly balanced. Rather than being simply a small cog in the bureaucratic machine, village heads are now more akin to true rural brokers. They benefit from the patronage of higher officials, but those officials also need the access which village heads provide to the rural masses, especially during elections. Thus, although patronage relations between state officials and village elites remain important, there has been a shift of bargaining power within those relationships in favour of the rural elites-cum-brokers.

Village politics in authoritarian and democratic Indonesia

Under the New Order, the central government organised village administration in ways that strengthened its ability to monitor and control the rural population. Law No. 5 of 1979 on Village Governance not only homogenised village structures around the country, destroying or sidelining diverse traditional institutions, but also brought them ‘firmly under the supervision and control of higher authorities’, positioning the villages as the lowest level of the Indonesian government structure, with village heads as an ‘instrument of the central government’. ⁵ The position of village head was elective, and village head elections continued to be held. However, candidates underwent ‘screening’ to ensure that they did not pose a security threat to the regime, and had to be approved by higher-level authorities before they could stand. The village head was assisted by a village secretary and several other officials (the pamong desa), a consultative body, neighbourhood and subneighbourhood heads and a range of corporatist associations. ⁶ These structures played a dual role: they were channels by which the government delivered rural infrastructure, birth control, basic health, agricultural extension and other programmes, and they were mechanisms to supervise and discipline the population. Overall, this authoritarian pattern of governance, especially the vertical chain of accountability connecting villages to higher state organs, produced a gap between village government and governed. A study based on research in 48 villages in three provinces late in the New Order concluded: ‘Villagers don't expect accountability and responsiveness and the system allows leaders to ignore it. In the most serious cases there is total alienation between villagers and formal village leadership.’ ⁷

In part, what made village heads and other officeholders effective was that they were generally drawn from local elite families who drew on sources of authority that were deeply rooted in their communities. In some places, this authority was sanctioned by traditional structures, such as those based on hereditary rule and customary law, or adat. ⁸ More generally, members of the governing rural elite were distinguished by their relative wealth: they were the biggest landowners in the village, the teachers or other government officials, the rice wholesalers and rural entrepreneurs. Thus, according to Frans Hüsken, in the part of Central Java where he conducted his research, local political history was

closely linked to the competition between several well-to-do families whose members had dominated the village administrations for decades. In some cases present-day village heads were the fifth or sixth generation in a line dating back to the latter part of the nineteenth century. ⁹

The resilience of the New Order regime was in no small part due to its ability to co-opt such locally rooted elites into its own apparatus of control. A critical transformation was that this rural elite became thoroughly integrated into structures of power and privilege external to the village community, as demonstrated by a series of ethnographic studies conducted during the period. As Gillian Hart explained on the basis of research in Central Java in the 1970s:

The key to understanding the distinctive features of agrarian differentiation in Java lies in recognizing that the rural elite is not simply a capitalist class that has emerged in response to technologically determined commercialization. They are in essence a class of favored clients of the state whose opportunities to accumulate hinge in critically important ways on their links with the state apparatus. The general tendency over the course of the New Order regime has been for the state to cater directly to the interests of the rural elite and simultaneously reinforce control over them. ¹⁰

Despite the importance of village elites for the New Order regime's survival, individually they had relatively little bargaining power in their relationships with higher officials. To remain in office, village heads and other rural officials had to demonstrate loyalty to the state and utility to their bureaucratic superiors. Analysts thus characterised the state's relationship with village elites as fundamentally hierarchical and top-down. Hans Antlöv, writing on the basis of his observations of a village in West Java a decade after Hart, concurred with her analysis, designating key village officials and elites as ‘state clients’:

The headman is the key to village development (kunci pembangunan) and a gatekeeper (perantara) between state and village. Thus, seen from the state, the crucial status of village officials does not derive from their administrative leadership but from their ability to shape village life to conform to norms set by the state, and to supervise the population. It is their patronage that is their power. ¹¹

Of course, these village elites gained much from their subservience. Higher officials not only turned a blind eye to their petty corruption, but also actively involved them in the rent distribution that was the oil that kept Suharto's ‘franchise system’ running. ¹² Village elites became closely connected by a mesh of bureaucratic and economic ties with higher officials, and used those relationships to skim funds from village budgets, gain protection for land or business deals and access the subsidies, programmes, licences and contracts that the state provided. As Hart explained, ‘the rural elite's relationship to supravillage authorities provides them with preferential access to agricultural inputs and credit and, perhaps even more important, to a range of highly remunerative non-agricultural activities, such as rice hullers, transportation, and large-scale trade.’ ¹³ Indeed, we can see such connections as a part of a deeper process of social change taking place during the New Order, one that has been called the ‘urbanisation of the rural’, ¹⁴ by which improved transportation and communication infrastructure lessened the social gap between villages and small towns, non-agricultural economic activities became more important in rural areas, and villagers began to adopt increasingly urban lifestyles and be more connected by migration and by socioeconomic ties to towns and cities. Just as the New Order's reordering of rural political power transformed village elites into agents of the authoritarian state, social changes were making them increasingly dependent upon connections forged outside their home communities.

During the post-Suharto reformasi from 1998, Indonesia's leaders reorganised village governance, echoing changes taking place at higher levels. Law No. 22 on Regional Government in 1999 generated changes that Antlöv characterised as constituting ‘a quiet revolution in the countryside’. ¹⁵ Villages became, in theory at least, self-governing units rather than the lowest level in the government's machinery of control. Village heads were no longer vetted and approved by higher authorities and became accountable, not to the district head (bupati), but to a new elected council, the Badan Pemusyawaratan Desa (BPD; Village Consultative Body). Some of these changes were later revised (the BPD was significantly weakened and is now ‘appointed through consensus’, rather than elected). ¹⁶ New participatory village development approaches were also rolled out, with the central government running a large programme, initially with World Bank funding and support: village communities would decide through consultative processes what projects they wanted to implement, manage and monitor. ¹⁷

Such changes have certainly affected rural politics. There is plenty of evidence that village elections take place in an environment of general openness, that villagers are more assertive when dealing with rural officials, ¹⁸ and that corrupt village officials are more vulnerable to exposure and punishment than during the New Order. Yet questions remain about the extent to which such changes have fundamentally reordered rural power relations. Perhaps the gloomiest assessment offered so far has been that by Takeshi Ito. ¹⁹ On the basis of ethnographic fieldwork in West Java early in the reformasi period, Ito's conclusions about village politics echo those of proponents of the ‘oligarchy thesis’ who argue that at higher levels of government the primary beneficiaries of political reform have been actors who became entrenched in power under the New Order, and who benefited from the enmeshment of private privilege and public office that characterised that regime. ²⁰ Ito argues that those who have benefited most from reformasi in the countryside have been village elites whose influence was formerly consolidated under state patronage. He suggests that ‘power relations within the village seem to be unaffected’ by democratic change. ²¹ Not only does patronage remain the main glue connecting village and higher-level elites but

transferring power and resources to local institutions has reinforced the dominant power of village officials who serve as interlocutors between state and countryside, and has prevented local people from using their rights to participate in decision making as guaranteed by democratic institutions. ²²

Elections in two villages

To study village elections, we chose two villages in a single subdistrict in one district (kabupaten) in the northern part of Central Java province, an area where we had for some time been conducting research on other elections, and where the district organised over 200 simultaneous village head elections in early 2015. In choosing our research sites, we were driven partly by practical considerations: we chose villages that were about 10 kilometres apart in order to facilitate travel between them, though we took care to choose villages with different physical, demographic and political features. During the course of our research over about three months (Noor Rohman was longer in the field), we interviewed several dozen informants, including all the candidates, many of their leading brokers and advisers, as well as village officials, religious and other informal leaders, and ordinary citizens. We also held informal conversations with many residents of both villages, and witnessed key events in the campaigning and voting process. In order to avoid identifying our informants — who spoke to us on condition of anonymity — we do not name the district, subdistrict or villages where our research took place in this article.

Alit Village (not its real name) is located at the point at which the coastal plain that stretches along the north coast of Java gives way to the foothills of the volcanic mountains that dominate the interior. The village, which is made up of three hamlets (dusun), is relatively small, with a population of around 1,200 persons, of whom 920 were registered voters. Set back a few kilometres from the main coastal road, it is considered relatively poor by residents of neighbouring villages. When we conducted our research, about a fifth of the 500 households were not directly connected to the electricity grid, a high figure for this part of Java. Most villagers were engaged in agriculture, with their efforts equally divided between the cultivation of rice in sawah (wet fields) and dryland crops such as cassava on tegalan or dry fields. Sweet yellow watermelons, which someone had discovered a few years earlier grew well in the village, were an important crop. As with nearby villages, part of the land consisted of borgan — land owned by Perhutani, the state forestry company, and occupied by villagers as part of the political upheaval accompanying the fall of Suharto in 1998. The villagers had felled the teak trees and negotiated ongoing use rights, replanting the land with teak or cash crops. Though there were a few prosperous villagers with relatively large landholdings, better houses, cars and other assets, Alit had few truly rich persons. There were no major rice wholesalers who lived here, no big retailers, nor even any significant home industry. Alit, like most villages in this part of Java, was overwhelmingly Muslim, with only about 12 Christians. Most of the Muslims were of the nominal, abangan variety, however. ²³

Ageng Village, located on the main road connecting the subdistrict to the district capital, had a population of 6,000, of whom 4,400 were registered voters. Most of the population were wet rice farmers, though some farmed on tegalan in the higher parts of the village, or operated fish ponds closer to the coast. Unlike Alit, there were a number of very well-off people in the village, notably several merchants who had done so well out of buying and hulling rice that they had been able to perform the haj. There were also several shopkeepers and others operating small businesses along the main road, as well as the usual array of school teachers, petty traders and more prosperous farmers. What made Ageng unusual was that about 70 per cent of the population were Christians, making it one of very few majority Christian villages in this part of Java. Most of the Christians lived in one of the three hamlets that made up the village.

In Alit, there were three candidates for village head. The first, whom we will call Sukardi, was the incumbent. The younger son of a large and relatively important family in the village, Sukardi was not particularly prosperous. After finishing junior high school, he had spent much of his youth working in Jakarta. Now, he rented land from others in the village and used it to grow crops (in this election year, his harvest of watermelon failed, causing him financial distress precisely when he most needed cash). Though respected by many elders, he had a reputation for cultivating support among some of the wilder village youths. A corruption scandal came to light in the final year of his term, when he used funding from a government village infrastructure development programme and from the village's own budget to build a poor-quality road. As one village notable explained, ‘people could see for themselves from the length and depth of the asphalt that it was worth only a fraction of what it was supposed to cost’. ²⁴ Around the same time, Sukardi also built a large new house for himself, despite being warned by village officials that this would look bad. This scandal was hushed up by members of the BPD and the rest of the village elite, who held a meeting and collectively decided that they would not report it to the police, allowing Sukardi to save face. Nevertheless, news of the scandal inevitably leaked through the village, to Sukardi's considerable disadvantage.

Sukardi had two rivals. The first, who was ultimately victorious, with about 360 votes of the 800 cast (about 43 per cent) to about 260 votes (32 per cent) for Sukardi, we will call Farid. Farid was the youngest son of one of the richest men in the village. Trained as an architectural engineer in the province's main public university, he was one of the few people in the village with a university education. Aged 40, he lacked charisma and was rather inarticulate, struggling in interviews to explain the duties of a village head, or his plans for village improvement. However, these shortcomings ultimately mattered little, seeing that it was his father, Supomo, who was the driving force in Farid's campaign (indeed, many villagers talked about their support for, or opposition to, ‘Supomo’ rather than Farid). Supomo was a retired civil servant who had worked in the subdistrict office of the Ministry of Education and Culture. During the New Order years he had been a leader of Golkar, the state's electoral machine, in the village, as well as being the son-in-law of a long-serving and respected hamlet head. He now owned about 5 hectares of land there, making him one of the bigger landowners. Several village notables were obligated to him: the village secretary, for example, was a distant relative who owed his position to Supomo's endorsement of him to Golkar leaders about twenty years previously. As a result, most perangkat desa (village officials) and other leading families supported Farid.

Importantly, Supomo had leveraged his bureaucratic authority to advance his children by ensuring that each received a good education outside the village: his eldest son was a lecturer in a reputable university in the provincial capital, and his two daughters had married, respectively, an official in Bulog, the state logistics agency, in East Java, and a Supreme Court judge in Jakarta. All three were able to help fund their brother Farid's campaign. In short, Supomo was a classic ‘state client’ who owed his wealth and status to his career in the bureaucracy and Golkar. His resources came from outside the village, but now he was reinvesting part of them back home, in land and in his son's campaign. As Supomo explained, getting Farid elected as village head was part of his own retirement plan: ‘I have four children; three of them have done well for themselves and moved away. So I thought I needed one here — I'm old and if I get sick there's nobody to take care of me.’ ²⁵

The final candidate was Rusdi, who won about 190 votes, or about 23 per cent. Aged 50, in his youth he had been a foreman for a local sugar factory, before working for a couple of years in a hypodermic needle factory in Korea. Returning from Korea, he used his savings to invest in sugar production, renting land and using it to grow cane. By the year of the election, he was renting about 20 hectares annually, making him one of the most successful agricultural entrepreneurs in this and neighbouring villages, and one of its richest men. He also worked as a village official, as head of general administration (Kaur Administrasi Umum). Rusdi was known as a devout Muslim, and respected for his honesty. He lived in a modest house and dressed simply. His main supporters were his immediate neighbours and the more pious villagers, especially the people he prayed with at regular pengajian (Islamic prayer meetings) — a minority in this mostly abangan village. As we shall see, his simplicity of personal style led him to adopt a unique vote-buying strategy.

In our second village, the incumbent village head, whom we will call Anto, was also not from a particularly wealthy or prominent family. A Christian, after graduating from senior high school in the early 1990s and working for some time as a conductor on public transport, he went to Lampung province in south Sumatra, where he worked for a while harvesting sugar cane. Several village youths moved to Lampung around this time, and a few became wealthy there. Anto called in some of these connections during his election campaign in 2015: one of his strongest backers had made it big in prawn farming in Lampung, and paid for three or four mini-buses of migrant workers, around 50 or 60 in total, to return home for the election. Not only were they a handy bloc of voters, but they also worked as enthusiastic brokers, being assigned special tasks such as paying off the local security forces. As we shall see, some of the more dramatic stories of vote buying focused on this Lampung connection.

On returning home in the mid-1990s, Anto worked as an agricultural labourer and sharecropper in a neighbouring village, and married a woman from one of the village's better-off Christian families. He won his first election as village head in 1998, when he was advantaged by the fact that he was the sole candidate from the majority Christian hamlet, while the votes of Muslims and other hamlets were split between three competitors. Anto was re-elected in 2008, and in 2015 was standing for his third term. By the time of the 2015 election, his wife and he jointly owned four hectares of fish farms and half a hectare of rice fields, making them a prosperous household, especially after adding Anto's annual salary as village head of about 12 million rupiah. Many villagers admired Anto for his down-to-earth style, and for his reputed honesty. They referred approvingly to his readiness to ‘go down personally’ (turun langsung) to resolve disputes, such as brawls between drunken youths. Admirers spoke of him in terms we have often heard used to describe successful politicians at higher levels of government, noting, for instance, his ability to merakyat (be at one with the people), and the fact that he was not sombong (arrogant). As we shall see, he was also appreciated for his ability to improve village infrastructure.

The second candidate, Najib, a Muslim, was not a native to the village, but he was the preferred candidate of most of the village elite, especially the perangkat desa. Born into a leading family in a nearby village, Najib had married into one of the village's wealthiest families. His wife, Yuningsih, was related by marriage to a wealthy haji who had been the village head for about 20 years from the start of the New Order in 1965. Yuningsih's brother-in-law, Haji Zul, was a rice wholesaler and miller famous for his wealth throughout the subdistrict. Haji Zul was also the village's largest landowner, with about 10 hectares. Yuningsih was related to many other leading families — the village secretary, for example, was the grandson of the former village head, and thus distantly related to both Yuningsih and Najib. Yuningsih was an energetic entrepreneur, who managed to get elected to the district legislature in 2009, representing one of the smaller parties, and had won back her seat, after changing parties, in 2014. As Yuningsih and Haji Zul's fortunes had improved, so had those of Najib. In around 2009 he had become a perangkat desa, taking a position as Kasi Pemerintahan (head of government business), a job that made him responsible among other things for issuing the all-important KTP (Kartu Tanda Penduduk, Citizen Identity Cards) and KK (Kartu Keluarga, Household Identity Cards) which citizens need to access government services. This gave him the opportunity to build up debts of obligation among his neighbours. Najib thus had several advantages, despite his minority Muslim faith. He was backed by two of the wealthiest residents of the village, his wife and brother-in-law, who funded a lavish campaign effort. He could draw on his wife's political resources, notably her network of brokers and the government projects she could deliver. Najib was also supported by other village notables, especially his fellow perangkat desa.

Despite these advantages, Najib did not win the election. He gained about 1,450 votes (42 per cent) to Anto's 1,960 (58 per cent). Najib's loss illustrates the power not only of religious identity, as we explore below, but also of Anto's common touch. One of Najib's key backers, the village secretary, reflected bitterly on the loss, saying that the result showed that family line was no longer important in village elections; what counted now was the candidate's ability to forge good relations with the community (bermasyarakat). As he said of Anto: ‘He lacks a network up above, but his network down below is good.’ ²⁶

Vote buying and other strategies

In both villages, the candidates all shared a basic repertoire of campaigning techniques that are standard in village head elections, some of which had been observed in the Suharto period. ²⁷ Above all, candidates aimed to project to voters a combination of generosity and humility, demonstrating simultaneously that they had the common touch and that they could provide direct benefits and gifts — notably, cash — to villagers.

There was, however, little that could be called explicitly programmatic in the campaigning. To be sure, election committees in both villages launched the campaign periods with formal events in which candidates delivered speeches outlining their ‘vision and mission’. Nobody thought that the content of these speeches had much influence, however. Most voters felt that they had already got the measure of the candidates’ character and abilities and, though they might differ slightly in what they promised — repaving a village road here, dredging a canal there — most believed that it was not their plans that counted but their access to funding, their administrative abilities, probity, and similar attributes.

Candidates used campaign techniques that showed they could interact warmly with ordinary villagers. All made personal visits, accompanied by their wives, to all or most households in their villages. The main purpose of these visits was to strike a posture of informality, friendliness and humility. Candidates knew they had to show that they did not put on airs in their interactions with ordinary villagers. Candidates also demonstrated their openness and generosity by organising open houses (mele'an), throwing open their homes during the evenings in the weeks leading to the election, so supporters could come along and discuss the campaign or just chat in a convivial atmosphere, with plenty of free food, drinks and cigarettes.

Each candidate established a team of vote brokers. These brokers were known as success team (tim sukses) members — a generic Indonesian term used all around the country — or as sabet (a local Javanese term that means ‘whip’). Such teams were organised in the pyramidal structures also used in higher-level elections. ²⁸ The better-organised candidates had hamlet coordinators to oversee the grassroots brokers. The grassroots brokers, in turn, were supposed to recruit a number — usually between 5 and 20 — of their household members, relatives and neighbours to vote for the candidate. The largest and best organised team was Najib's in Ageng: at the apex it consisted of a chair, a treasurer, a secretary, and then 3 hamlet coordinators, under whom were 29 neighbourhood coordinators and then a total of about 260 sabet, each charged with recruiting a maximum of 12 voters. ²⁹ Anto had a similarly large success team, though we were not told the precise numbers. In Alit, success teams were smaller and less rigidly organised. Farid's was the most elaborate here, with 5 core members and a total of about 20 brokers.

Patronage distribution was central to the campaigns. It came in two varieties. First were ‘club goods’ ³⁰ — gifts that are targeted at groups rather than individuals. Only two candidates had the resources to be generous in this regard. In Alit, Farid, or rather his father Supomo, provided several such gifts, erecting 41 streets lights, paying for the entire village's Land and Building tax allocation for the year, purchasing sports equipment and uniforms for all village youths who wanted them, and providing a free trip for all the women in the village to see the sights in the provincial capital, Semarang. In Ageng, meanwhile, Najib presented several infrastructure projects as part of his campaign. Rather than drawing on private money, however, his wife directed a stream of projects to the village as part of her allocation of ‘aspiration funds’ (discretionary funds allocated to individual legislators for projects in their constituencies). ³¹ Najib gave nets to some fishers’ cooperatives, and promised ambulances for each hamlet and excavation of canals. When Najib lost the election, his wife retaliated by pulling all these projects and redirecting the funds to neighbouring villages.

The second, and more important, form of patronage was vote buying — the distribution of individual cash payments to voters. All five candidates engaged in this practice. They all used the same basic method to distribute their money, placing it in plain envelopes to take house to house one or two days before the election. Sometimes the candidates distributed the money themselves; usually they delegated the job to brokers. The distribution of sangu (pocket money), as these payments are known locally, is deeply entrenched in local political culture. Several studies indicate that the distribution of cash to voters already featured in village head elections in rural Java during the New Order. ³² Informants in our villages agreed; some claimed distribution of sangu had occurred in village head elections since time immemorial. Vote buying is also endemic in higher-level elections in this part of Java. ³³ Indeed, legislative candidates whom we interviewed in the area in a related research project explained that in designing their vote-buying strategies they mostly adapted techniques that had long been used in village head elections. ³⁴ We should thus be careful not to read the vote buying we describe below as a post-Suharto seepage of unsavoury national practices into village politics. If anything, the process has been the exact reverse, with practices of vote buying long experienced at the village level being upscaled into national elections after democratisation.

Though all candidates distributed cash, the amounts they distributed, the number of recipients, and the precise methods of delivery varied. In Alit, Rusdi paid 100,000 rupiah (US$7.25) per voter, Sukardi paid 150,000 rupiah (US$10.90) while Farid — or, rather, his father, Supomo — paid 200,000 rupiah (US$14.50). There were 920 voters in the village; while Rusdi made payments to all of them, the two more serious candidates targeted their payments only to potential supporters, or about 600 voters each. This meant that, while Rusdi and Sukardi each spent around 90 million rupiah (US$6,500) on vote buying, Farid/Supomo spent at least 120 million (US$8,700) — in fact, they spent more because, as Supomo explained, he made multiple payments to numerous households. In Ageng, where there were about five times as many voters, payments were lower, at 50,000 rupiah (US$3.65) per head. Najib's team distributed around 3,500 such payments, while Anto's distributed 2,700, meaning total expenditures of approximately 175 million rupiah (US$12,700) and 135 million rupiah (US$9,800) respectively. In fact, Anto had planned to pay only 30,000 rupiah per voter but, on learning that Najib was paying more, he made a last-minute decision to match him. This was inconvenient, because the envelopes had already been stuffed and now had to be reopened and refilled. Back in Alit, Sukardi wished he could match the 200,000 rupiah paid by Farid/Supomo, but could not afford to do so. ³⁵

Most candidates distributed this money through their brokers. They also used those brokers, weeks in advance of the poll, to identify potential supporters, and so target their payments. Though the brokers were supposed to talk to voters and lock in their support, in fact this targeting was typically rather loose. Anto explained his strategy as follows:

Those we targeted for ‘execution’ [eksekusi, another local euphemism for vote buying] were, first, those who were not members of the success team [of my rival] and, second, those who did not have kinship ties, work ties or other relations [with Najib or his core supporters]. We wouldn't want to give them money — that's what we would call in Javanese nguyahi segoro [salting the sea]. ³⁶

Accordingly, when distributing cash, each candidate estimated that they would experience wastage, and distributed to a proportion of voters sufficiently large that they would still win if they secured the votes of only 60 or 70 per cent of recipients (some of these calculations failed: in Ageng, Najib, with about 1,450 votes, received a ‘return’ on his vote-buying efforts of only 41 per cent, while the victorious Anto achieved 70 per cent). The only exception to this rule was Rusdi, the more religiously devout candidate in Alit. Not only did he deliver his payments personally, rather than relying on brokers, but he also paid all 920 voters in the village, even those he knew would support other candidates. He did so, he said, so he would have ‘no burden’ in the future, and because he thought of these contributions as religious alms (sedekah) rather than bribes. ³⁷

Another variation on the general pattern came about as the result of the integration of a gambling scheme with vote buying in Ageng. Gambling on the outcome of village and other elections is a long-established tradition in this part of Java, and big-time gamblers (known as botoh) sometimes wager huge sums in the process, with ordinary villagers making smaller bets. ³⁸ In Ageng, Najib's team offered voters a deal: they could keep the 100,000 rupiah they were being given, or they could use it to bet on Najib's victory, at 3:1 odds. If Najib won, they would thus receive 300,000 rupiah, a considerable sum. About 800 voters accepted the offer, so Najib's team set aside 80 million rupiah of the money they had allocated for vote buying for this scheme. To cover their financial risk should Najib win, they then bet this 80 million rupiah against another botoh — one who was backing Anto (stories varied but, in one version, this was Anto's friend from Lampung who had made it big in prawn farming; we were not able to verify this part of the story). If Najib won, his backers would thus not be out of pocket. The plan failed, with Anto trouncing Najib on voting day. This failure made Najib's supporters suspicious: they did not believe that so many voters would have wagered their own money on Najib and then voted for Anto. As we shall see, this made them suspect foul play. In Alit, meanwhile, though there was small-scale betting, no big-time botoh came in from outside — simultaneous elections elsewhere in the district meant that there were larger villages where they could run their schemes.

When asked to describe the meaning or effect of the payments of sangu, candidates, brokers and recipients alike often had trouble articulating a clear response, beyond stating that they were a local custom, and that people would not turn out to vote without them. The typical conceit used to justify the payments was that people needed sangu to compensate them for the day's income they would lose by going to vote, though of course money went to everyone regardless of their regular income. Almost everyone agreed that a candidate who did not pay would not be considered a serious contender — in fact, such a thing was quite unheard of. Members of the village election committees accepted the practice. Indeed, in Alit the committee even offered to regularise it by collecting money from each candidate, pooling it and then distributing it equally to all villagers. The candidates failed to agree on this plan, however, so it did not go ahead.

In fact, some practices are considered illegitimate in the villages, pointing to the existence of strong local moral norms concerning vote buying. While payment of sangu is normalised, giving additional payments — usually known as tembakan or ‘shots’ — to target voters at the last moment is widely viewed as underhanded and dishonest. Each side in these elections suspected their competitors of targeting small groups of voters with additional payments on the night before the election, in order to induce them to switch sides. Najib's supporters in Ageng said they only lost because Anto's team had targeted some voters with last-minute payments of 300,000 or 400,000 rupiah (US$22 or US$29). The story was that they had specifically targeted participants in the gambling scheme mentioned above and, in one version, the man making the payments was the botoh from Lampung who backed Anto, who was simultaneously trying to destroy Najib's chances and secure his own bet. Though we spent much time trying to track down recipients of these alleged payments, we could not verify that they were anything but apocryphal. Even so, fear of last-minute ‘attacks’ meant that in both villages the night before the election was tense, with groups of supporters of the candidates guarding entrances of laneways they considered to be their base areas, to prevent entry by rivals.

What are we to make of these payments in these elections? Most obviously, they present sobering confirmation of continuity of elite power in these villages, in that each candidate had to invest what were, in local terms, huge sums in their campaigns. Only the wealthiest villagers could compete. To provide context, in Alit the income of the village head included the rental income from tanah bengkok (village lands allocated to village officials) equivalent to about 50 million rupiah per year plus a formal subsidy from the district of 1.1 million rupiah per month. These sums totalled about 380 million rupiah (US$27,500) over the six years that the winner would serve. The winner's father spent 120 million on vote buying alone and, though he claimed not to have kept count of his other expenditure, supporters agreed it could easily have been double this sum, meaning that Supomo spent well over half and perhaps close to the entire formal income that his son would earn over his full term. In Ageng, the situation was more dramatic, given that the BPD had recently voted to take the income from its bengkok lands away from village officials and allocate it instead to the village budget. This meant that the village head's formal income was only the monthly stipend, equivalent to about 79 million rupiah over six years — much less than both candidates paid on vote buying and only about one-quarter of the 300 million rupiah which (according to one supporter) Najib and Yuningsih spent on the campaign (we could not get the total from Anto's camp). In other words, candidates could expect to recoup their campaign investments barely or not at all from their formal incomes as village heads. As at higher levels of electoral competition, in order to fund their campaigns candidates had to draw on private income, money from wealthy backers, or from illicit rents they could obtain from the village budget or project funds — points we return to below.

An obvious sign of the power of money in these elections was the victory of Farid, which everyone in Alit agreed reflected the superior financial resources which Supomo mobilised on his son's behalf. Even Supomo shared this view. When we asked him to explain his son's victory, Supomo was blunt: ‘I won't say it myself, because it will sound conceited. But people here say I have a lot of money.’ Indeed, he thought there was a wider lesson to be learned: ‘Everyone praises Indonesian democracy. But in fact it's money that's number one here.’ From the start, Supomo had made it clear that he would not allow himself to be outbid. As he put it, ‘there was only one word I used: “Yes”. I never said “no”. Whatever people asked for, I gave it to them.’ ³⁹

Findings from neighbouring villages confirm the power of money. Though we did not conduct systematic research outside the two villages, we did collect information on nine other village head elections held on the same day in the same and one neighbouring subdistrict. In each case, the pattern of vote buying was similar, with all candidates making payments to voters, ranging from 50,000 rupiah to 250,000 rupiah per head. In general, victory went to the candidate making the largest payment, or to one of two or more candidates whose payments were equal. The candidate making a smaller payment won in only two of these villages, and in one he was a Muslim facing a Christian in a majority Muslim village. We know from research carried out elsewhere in Java (a limited Indonesian language literature — much of it consisting of undergraduate theses — is available on vote buying in village head elections ⁴⁰ ) that this pattern of vote buying is endemic to village head elections throughout the island (less is known about patterns elsewhere), and that the most cashed-up candidates often win. Overall, it is hard to avoid the conclusion that ‘money politics’ is at least as integral to village-level politics as it is to politics at the district, provincial and national levels.

However, it would be wrong to conclude that these elections were only about money. Other factors also shaped voting behaviour. In Ageng, the more cashed-up candidate, Najib, was defeated. This outcome was partly attributable to the unusual religious element in this election. Despite generally harmonious interreligious relations in the village, including much intermarriage, the priests and lay workers spread the word that the more numerous Christians should vote for a co-religionist, and most apparently complied. A similar campaign by some of Najib's Muslim supporters backfired, when rumours spread that some Muslims were saying it was forbidden to vote for a kafir. Candidates from the majority group typically win in such polarised elections at higher levels; ⁴¹ it is not surprising that the same should happen in a village. At the same time, Anto gained support from both Muslim and Christian residents for his approachable style, his record of problem solving, and, as we shall see, his achievements in village development. By contrast, in Alit, voters were swayed by the blitzkrieg of payments made by Supomo, at least partly because no alternative candidate was particularly attractive: Sukardi, the incumbent, was discredited by his recent graft, while Rusdi was considered by many to be an impractical idealist.

In fact, it is possible to argue that cash payments to voters in Indonesia rarely determine election results, but are more akin to an ‘entry ticket’ ⁴² — candidates have to make payments to show that they are serious contestants and to signal that they are genuinely concerned about their constituents, and willing to help them. Cash is the material equivalent of the house-to-house visits which the candidates make, and only once it is paid do voters give a candidate serious consideration. This does not mean that money politics is unimportant, but rather that it provides the framework within which other factors — a candidate's record, personal style and so on — are evaluated by voters.

The role of the state and rural brokerage

As noted above, a significant body of literature argues that village elites during the New Order became increasingly dependent on connections outside their communities. In particular, they became reliant upon the authority and patronage of the state and, as a result, became subservient to it. Close observation of the two elections that are our focus shows that, despite democratisation and greater vertical accountability connecting the village head to ordinary villagers, the dependence of village elites on external sources of support, especially higher state officials, has continued and even deepened. However, this dependence no longer gives rise to a straightforward hierarchical relationship of subservience as it formerly did, but now offers considerable leverage — and pitfalls — to village elites.

One striking fact about these two elections was that all candidates drew upon supravillage connections and resources. We can show this by looking at the candidates’ life stories and the funds they used to finance their campaigns. All of them had spent significant periods outside the village, periods which contributed to their prominence once they returned. Their financial resources were also largely derived from beyond the village borders. In Alit, the money that Supomo lavished on his son's campaign was accumulated during his bureaucratic career, supplemented by contributions from his other children who had prospered far from home as a result of it. We did not obtain a clear picture of how Sukardi funded his campaign, though most villagers assumed he drew at least partly on government project funds that he had illicitly accessed. Even Rusdi, in some ways the least worldly of our candidates, owed his entrepreneurial success to his time overseas. In Ageng, Anto made his start in life in far-off Lampung province, and was able to call in his connections there. Najib, meanwhile, had married into the most entrenched family of notables in the village, but depended partly on resources that his wife could mobilise as a legislator in the district parliament.

More precisely, the outcomes of these elections were at least partly determined by the competitors’ relations with the local state, especially bureaucrats and elected politicians at the district (kabupaten) level — the locus of political and budgetary power in the new decentralised Indonesia. In particular, one criterion by which villagers evaluated the candidates was their record in, or potential for, promoting village development by attracting projects from higher levels of government. Everywhere we went, as in much of rural Indonesia at election time, much of the talk was about the state of the village roads, irrigation canals, public buildings, houses of worship and other facilities, and what had been, or could be, delivered to improve them. In this regard, the difference between our two incumbents was striking. In Ageng, Anto was judged by most villagers as having been successful at attracting projects to address residents’ development and welfare needs. For example, the village roads had first been asphalted under his tenure, and he was praised for attracting and successfully administering various small-business schemes from the district. Villagers spoke warmly about his good relations with higher officials and politicians and about his skills in lobbying (melobi) for projects. In Alit, Sukardi fumbled in this area. Not only did many people believe that his record of project delivery was poor and skewed toward his own hamlet, but he made a mess of the projects he did deliver by getting caught with his hand in the till. By doing so, he violated a norm that has been observed to bring down political leaders in higher-level elections — while a certain level of graft is considered acceptable, ‘excessive’ corruption, especially combined with poor policy performance, can fatally undermine a leader's legitimacy. ⁴³ Villagers also evaluated the other candidates in this light. In Ageng, Supomo's history of state links was seen as an asset, as was Farid's background as a contractor; in contrast, Rusdi was viewed by some as being too ‘straight’ (lurus) to deal successfully with government. In Ageng, Yuningsih was simultaneously Najib's great strength, because she could guarantee a stream of projects, and his Achilles heel, with some villagers fearing that the village would be ‘owned by one family’ ⁴⁴ if its village head was married to its only representative in the district parliament.

Such factors underline the continuing significance of links to overarching state structures in the internal dynamics of village politics. In order to succeed, village heads need, as under the New Order, the patronage of higher state functionaries and officials. However, their mode of interaction with such officials has changed dramatically. During the New Order period, village elites were integrated into a rigidly hierarchical, top-down structure. Higher state officials judged them on their ability to maintain order in the village, implement government programmes and deliver election victories to Golkar. Though village elites could sometimes frustrate government programmes that conflicted with their own interests, they had relatively little bargaining power in these relationships.

The new mode of integration occurs in a context, not only of decentralisation, but also where a significant proportion of state resources are dispensed through ‘projects’ — parcels of funds that are designed for specific purposes and which undergo at least a formal bidding, application and/or verification process. ⁴⁵ Projects proliferate in rural Indonesia, as in other parts of the country. As Li puts it, ‘villages are awash with small projects sponsored by dozens of different national and transnational agencies that distribute free goats, sewing machines, water systems, and micro-credit schemes’. ⁴⁶ Village elites value projects not just for the development or welfare benefits they deliver, but also for their utility as patronage resources that can be distributed among supporters.

In order to be proficient in gaining projects, village heads and other elites require two sets of skills. First, they need specialised literacy and numeracy skills — the ability to write convincing project proposals and accountability reports is now a highly valued resource throughout Indonesia (in a neighbouring subdistrict we met one businesswoman who proudly boasted that her young son — who was still in elementary school — was teaching himself to write good project proposals, tailored to the requirements of particular government bureaus). The second, even more important set of skills are those needed to negotiate the tangled political map that constitutes state authority at the local level, in order to identify, build connections with, and garner projects from the officials who can dispense them. In place of the old clear lines of bureaucratic command, village elites now confront competing elites in multiple and fragmented state institutions — they can extract projects, not only from the bupati or his or her deputy, but often also directly from bureau chiefs, subdistrict heads, or other civil servants, as well as from members of the district parliaments, to say nothing of officials or politicians from the provincial and central governments. Navigating these webs successfully demands a mixture of political judgment, subtle social skills and a finely tuned ability to assess political and legal risk.

Accordingly, while some analysts see projects as functioning primarily to depoliticise rural life, ⁴⁷ we believe that they must be seen as central to the new rural politics. In particular, projects are a critical new currency in coalition-building, patronage politics and mediation between different levels of government. Crucially, district-level elites now need village elites almost as much as they are needed by them. At election times, aspirants to public office need village connections. Just as our candidates in these village elections worked through grassroots brokers, so too higher-level legislative and executive candidates mobilise voters by way of large ‘success teams’. Most village heads are vote brokers par excellence and, all over Indonesia, political candidates try hard to attract their support. Successful village heads are often able to leverage their campaigning skills and connections with district or provincial-level politicians by launching into higher-level political careers; it is especially common to see former village heads running for seats in district parliaments at election time, and they often do so with the backing of politicians whom they have assisted in the past. ⁴⁸

Village heads are particularly effective vote gatherers, partly because, as elected officials, they wield authority to speak on behalf of their communities, but also because villagers themselves depend on village heads for assistance in dealing with government agencies and accessing welfare programmes. ⁴⁹ In some places, village heads can enter into one-on-one clientelistic relationships with bupati that mimic the hierarchical relationships they developed with state officials under the New Order, so that a bupati becomes a patron who provides them with a stream of projects and other rewards, which they repay with political loyalty. ⁵⁰ But one-on-one relationships can leave village heads dangerously exposed should there be a change at the top, so many successful village heads spread their risks by diversifying their political connections and, hence, their sources of projects. In our two villages, the stand-out performer in this regard was once again Anto — in the last three elections he had supported a range of legislative and bupati candidates, from multiple parties. With only one of these candidates failing to be elected, he was apparently skilled at picking winners, providing him with a rich bank of favours to be called in from successive district governments and parliaments. In Alit, Sukardi was much less successful at building alliances, relying only on the support of a powerful local parliamentarian who had been locked in a bitter struggle with his own party and who had tried, but failed, to become bupati.

Another point of contact where village elites have considerable leverage in their dealings with higher officials is in the patronage system that surrounds projects. As in the past, a major source of the skim-offs and mark-ups that fuel Indonesia's patronage system come from local projects. Given that project delivery typically occurs at the village level, district officials need the involvement or at least acquiescence of village officials, especially village heads, in order to extract funds. This necessity increases the value of village heads for higher officials, but also exposes them to the risks involved in corruption (the figure of the corrupt village head has become something of a cliché in post-reformasi Indonesia, with regular media exposes of such individuals). The village secretary in Alit explained the system like this:

Usually it's [district legislative] council members who are active in lobbying for projects. They are the ones who know what programmes there are at the higher level, and if they know someone down at the grassroots, they will tell them about it, and tell them to make a proposal, and then promise them: ‘I'll be the one who will deliver it’. Then, there is typically a bribe of 10 or 20 per cent [of the project value]. Sometimes it can be 50 per cent, though 10–20 per cent is typical. Then, when it comes to making the accountability report [as the project recipient] there can be trouble if you are supposed to account for the full 100 million [rather than the 50–90 million the village actually received]. Often the village head will make his own report, and stamp it himself [i.e. keeping other village officials in the dark]. But if you don't get involved in ‘lobbying’, you don't get the projects. ⁵¹

On the one hand, therefore, the system provides leverage for village heads, who can use their support for higher officials’ plans not only to attract projects, but also to gain a share of the rents themselves. On the other hand, the system of patronage extraction and distribution surrounding project delivery can be a trap — one which, as we have seen, Sukardi fell into. We could not ascertain whether Sukardi was caught out in a double cross or whether he was just too greedy and lacked the skills to cover his tracks effectively. In Ageng, we should stress that Anto was not able to avoid the pay-offs that must be made to secure projects. In an interview, he reeled off going rates in ‘fees’ for various types of projects. But he made it clear that he would reject projects if the take was too high, ⁵² and he was apparently able to negotiate more effectively with higher-level officials and to know when sticking to the formal rules would confer political advantage. As an example of the latter course, Ageng was the only village in the subdistrict where allocations of subsidised rice were distributed only to the poor villagers identified by the responsible government agency as intended recipients. In other villages — as through much of Indonesia ⁵³ — village heads and RT (subneighbourhood) heads succumb to popular pressure and distribute the rice equally to all villagers, or direct it toward their own supporters. Anto made this programme a central plank of his campaign strategy, telling the more than 500 recipient households that if Najib won their allocations would be reduced.

Conclusion

Our analysis of these two elections suggests that, as with other parts of Indonesia's political life, there are elements of both continuity and change in rural power relations in the post-Suharto period. With regard to democratisation of rural politics, our study confirms that village head elections are truly competitive. Higher authorities play no overt role in them: subdistrict and district civilian officials, to say nothing of the security apparatus, neither decided who could run nor guided the villagers toward any candidate. Instead, both elections were genuine contests, in which the candidates had to devote — in village terms — massive network resources, intellectual creativity, physical effort and material wealth to attract the support of their fellow villagers, including the poorest among them. On the other hand, as has been exhaustively demonstrated and debated in the literature with regard to higher-level politics, there is considerable continuity in the underlying power relations. The primacy of vote buying shows that only the wealthiest villagers could compete, and, much as with Indonesian politics writ large, there was little policy debate in these contests.

But we should not exaggerate this continuity. Not all the contestants, nor indeed all the incumbents, were drawn from the narrow groups of elite families who had long dominated village affairs in both locations. They were wealthier than other villagers, but some were arrivistes — individuals who had prospered through the greater opportunities for mobility outside the village that have accompanied the remaking of the rural economy over recent decades. The village political elite is heterogeneous, and its boundaries are highly porous, as with higher elites. ⁵⁴ Moreover, while the ubiquity of vote buying was a sobering sign of the reach of money politics to the very base of the political system, it was decidedly not the only thing determining the results, and should not be read as signifying lack of competitiveness. On the contrary, contestants devoted such resources to buying support precisely because they could not take the votes of villagers for granted.

In much the same way, both continuity and change are evident in village elites’ relations with the local state. In some respects, village elites such as those competing in these elections are still ‘state clients’. Most of the wealthiest and successful villagers seek village-level state office, either positions as perangkat desa or seats on the BPD, and almost everyone has an eye out for opportunities to capture rents from the projects that flood rural life. The strong candidates in these elections were able to draw upon allies and resources they had built up through their interactions with higher-level state officials. Ability to snare projects from state agencies is a key ingredient in village political success. Thus, although higher state authorities played no direct role in these elections, their indirect role was considerable — by deciding which villages or individuals to favour with projects or other resources, they could greatly affect village electoral outcomes. Despite this underlying continuity, however, the nature of the relationship between village elites and state power has changed greatly. Rather than a rigid vertical hierarchy, we now find a messy zone of multiple connections where state meets village. Village officials confront multiple sites of potential contact in the local state as they seek projects and other resources to develop their villages and improve their own material and political circumstances. Moreover, not only do village elites need state officials, but those officials need partners in the villages to themselves attain power and patronage. In such a context, suppleness and subtlety in how they cultivate state connections is more important for village elites than simply demonstrating loyalty. Deal-making and brokerage are the keys to political success in rural Indonesia's patronage democracy.