3 Indonesia’s decentralization:
the rise of local identities and the
survival of the nation-state
Marcus Mietzner
3 INTRODUCTION
Nation-states launch decentralization programs in highly differenthistor-
ical contexts and for a wide variety of reasons (Eaton, Kaiser and Smoke
2010: 8-9). China and Vietnam, for example, have pursued moderate
decentralization policies in order to respond to economic modernization
pressures. Brazil, South Africa and Mexico introduced decentralization
as part of their post-authoritarian transitions, while Cambodia, Rwanda
and Uganda gave mote power to the local level to heal post-conflict trau-
‘mas. Most importantly, many developing countries use decentraliza-
tion as a tool to improve government effectiveness, responsiveness and
accountability (Asante and Ayee 2004: 3). Few states, however, imple-
‘ment decentralization as an urgent measure to prevent the disintegration
of the state in the middle of a severe political crisis. Arguably, indonesia
in 1999 was such a state: many observers were concerned that, having
‘emerged from 40 years of centralist authoritarianism, post-Suharto Indo-
nesia could become the next Yugoslavia or Soviet Union, Fuelled by the
secession of East Timor in 1999, anxiety spread within Indonesia that
other parts of the archipelago could also demand independence. While
‘most serious foreign scholars believed that a territorial break-up was
unlikely (Cribb 1999; Emmerson 2000; Aspinall and Berger 2001), they
nevertheless devoted much time to discussing such a probability. This,
in turn, did little to reassure Indonesian policy makers.
Faced with the threat of territorial disintegration, the Indonesian
elite turned to decentralization. First designed in 1999 and implemented
4546 Morcus Metzner
since 2001, decentralization transformed Indonesia from one of the most
centralist nations in the world into a polity with near-federal structures.
Asa result, the share of regional spending in total government expendi-
ture increased from 17 per cent in 2000 to 40 per cent in 2009 (World
Bank 2002, 2009: 65). This reform was implemented despite a number of
well-documented risks inherent in decentralization. First, the empirical
record of decentralization around the globe is ‘mixed at best in terms
of realizing many of the stated objectives of reform’ (Katon, Kaiser and
Smoke 2010: xi), providing no ‘unambiguous proof of its desirability’
(Smoke 2003: 7). Second, ‘decentralization is likely to accentuate the
already precarious imbalance within the state’ because it often favours
already rich districts over poorer ones (Asante and Ayee 2004: 2). Third,
decentralization frequently leads to wasteful spending on bureaucratic
apparatuses (Mawhood 1999). Fourth, political decentralization does not
automatically produce improved democratic quality, as ‘the devolution
of power may help to augment the dominance of those who, because of
‘their wealth orstatus,are already powerful atthe local level’ (Asante and
‘Aye 2004: 2), Finally, decentralization can increase rather than reduce
the levels of corruption (Prud'homme 1995).
‘Most studies of Indonesia's decentralization process have found
that the country has fallen victim to most of the phenomena outlined
above.! There is little evidence of significantly improved public services;
resource-rich areas have seen higher growth rates than poorer districts;
government spending on salaries and buildings has ballooned; elites
have become the main players in local elections; and corruption is ram-
pant. Yet, despite these deficiencies, the Indonesian nation-state not only
has survived, but is probably stronger today than at any other pointin its
history. As Aspinall (2013: 132) has pointed out, the ‘execution of decen-
tralization policies played a major role in heading off severe state crisis
in Indonesia’, mostly through its immediate taming effects’ on separatist
sentiment in some regions.
While Aspinall focuses on the elites and their endorsement of poli-
cies that gave them more resources, there are at least four additional
factors behind the consolidation of centre-periphery relations in Indo-
nesia, These are the strong public satisfaction with decentralization; the
non-violent flourishing of local identities; the increased level of state
penetration; and the effective design of local elections as a channel for
cross-constituency cooperation and the rise of alternative leaders, This
chapter explores the role of each of these factors in turn (sections 3.2-3.5),
1 See, for example, Hofman and Kaiser (2004), Aspinall and Fealy (2003), Mal-
ley (2008), Schulte Northolt and van Klinken (2007), Bante (2008) andl Hadiz
e010)‘The rise of lcal identities and the surviva of the nation-stale 47
‘Through a discussion of the ways in which decentralization has reduced
tensions and stabilized the state, the chapter confirms that effectively
designed policies can help heterogeneous states to decentralize without
stisking disintegration (Falleti 2005; Miodownik and Cartrite 2009).
3.2 MESSY BUT POPULAR: DECENTRALIZATION IN OPINION
POLLS
While the popularity of decentralization among local elites is impor-
tant for its success (Aspinall 2013), sufficient levels of popular support
are equally critical. The degree of this support, in turn, can be assessed
through public opinion polls. Until the early 2000s, most potitial sci
entists were reluctant to use opinion surveys to gauige the views of citi-
zencies in new democracies. This scepticism was only partly related to
questions of methodology; more importantly, it was based on the domi-
nant view among theorists that most polities in Asia, Africa and Latin
America were inevitably elite driven (Linz and Stepan 1996). It was only
with the emergence of multi-country opinion survey projects such as
Asian Barometer (which started in 2001) that senior political scientists
began to take a closer look at public opinion as a determinant of political
processes (Diamond 2010).
In Indonesia, for instance, the study of public opinion has been crucial
in explaining the surprising stability of post-Suhario democracy despite
the endemic corruption, weak rule of law and complaints about govern-
ment effectiveness (Aspinall and Mictzner 2010), While some analysts
have traced this stability to concessions made to potential spoilers such
as the military (Aspinall 2010), a more compelling answer is hidden in
opinion polls: since 1998, between 71 and 90 per cent of Indonesians have
consistently supported democracy. Regardless of whether this support
is deserved or not, the fact that it exists is politically relevant in itself
In essence, it means that there is no public pressure to adopt a different
system ~ much in contrast to Russia, for example, where the authoritar-
ian leadership can operate in the knowledge that only 29 per cent of its
citizens support democracy (Pew Research Center 2009: 26)
A similar pattern marks the study of Indonesia's decentralization.
A first glance it seems peculiar that amidst questionable performance
indicators and continued corruption, decentralization should have sta-
bilized the relationship between the capital and the periphery. But as in
the case of democracy, opinion polls can shed a different light on the way
decentralization has played out in post Suharto Indonesia. In a2007 poll,
73 per cent of respondents supported decentralization and only 27 per
cent rejected it (LST 2007). More specifically, 61 per cent of Indonesians48 Marcus Mictaner
Table 3.1 Views on education and health services after decentralization,
20078
Better Samo Worse Don't know
Eckucation 61 25 9 5
Health 54 29 2
1 Respondents wre asked whether the quality of education and health services hl im-
proved, stayed the same or declined afer decentralization
Source: LS1(2007).
believed that education services had become better under decentr
zation then they had been under centralist rule, while only 9 per cent
thought they had become worse (Table 3.1). In the area of health, 54 per
cent of respondents saw improvements, while 12 per cent thought that
the quality of services had declined. Indonesians were less enthusias-
tic, however, about the impact of decentralization on the fight against
poverty and unemployment: 34 per cent believed that poverty had
become worse under decentralization (27 per cent thought that it had
been reduced); and 40 per cent felt that unemployment had increased,
‘while 23 per cent thought it had declined. The high level of overall sup-
port for decentralization has been confirmed by polls conducted by the
World Bank, in which around 70 per cent of respondents testified to
improved quality of health and education services and about 56 per cent
to improved administrative services (World Bank 2009: 66).
Other opinion polls show consistently high levels of satisfaction with
local government officials. In fact, with the exception of the presidency
and the military, the popularity of government institutions increases
inversely to administrative level. In other words, satisfaction levels are
highest for the lowest level of state administration (village leaders), sec-
ond highest for subdistrict offices and so forth (Figure 3.1). In a 2008 poll
(Democracy International 2008: 19), village administrations received a
net satisfaction rating of 61 percentage points, followed by subdistrict
offices (67 per cent), district heads and mayors (38 per cent) and gover-
nors (31 per cent). This pattern is even reflected in the hierarchy of the
least popular political institution in Indonesia, the legislature. While the
national parliament received a negative net satisfaction rating, of minus
16 per cent, its provincial pendant enjoyed a positive rating of 6 per-
centage points, and the district-level legislature of 11 points. In the same
poll, 70 per cent of respondents described local government leaders as
‘accountable’, 67 per cent thought that local government services were
of ‘good quality’ and 56 per cent said they had more confidence in local‘The rise of loco identities and the survival ofthe nation-state 49
Figure 3.1. Net satisfaction with political administrations and institutions,
2008 (%)*
National parliament
Provincial parliament
District parliament
Province
District
Subdistrict, 37
Village er
40 202 0 CBD
isfaction is caleulated by subtracting the percentage of respondents dissatisfied with
' parcular aniston or institution from the percentage of saised respondents,
Source: Democracy International (2008: 19).
administrations than in national government (Democracy International
2008: 29).
As indicated above, the message of these opinion poll numbers
stands in stark contrast to the often unflattering accounts of decentrali-
zation by scholars. In one of the most striking examples of the large
gap between scholarship and opinion surveys, Kristiansen and Santoso
(2006: 247) have described health services under decentralization as non-
transparent, profit oriented and disadvantageous for the poor. However,
their own survey data showed that 75 per cent of respondents thought
the quality of health services was better today than before decentrali-
zation. Puzzled by this result, the authors dismissed their own survey
as ‘weak’, ‘remarkable’ and having ‘limited reliability’ (Kristiansen and
Santoso 2006: 255),
What, then, accounts for this discrepancy between scholarly assess-
ments of performance indicators and the ‘remarkable’ levels of satisfac-
tion among Indonesians with decentralization? To begin with, there are
many statistics that would confirm the general sentiments of the citi-
zeny. For example, Indonesia's score on the Human Development Index
(HDI, a composite of life expectancy, literacy, education and living stand-
ards) has consistently increased since decentralization, According, to the50 Marcus Mietaner
Table 3.2 Human Development Index, selected Indonesian provinces,
1996-2010"
1996 1999 2002 2004 2005 2006 2007 2008 2009 2010
West Nusa Tenggara 56.7 542 578 606 62.4 630 63,7 64.1 646 652
West Kalimantan 63.6 60.6 629 65.4 66.2 67.1 67.5 682 688 69.2
West Sulawesi == B84 65.2 67.1 677 686 692 696
Gorontalo — = 681 55.4 67.5 68.0 688 693 69.8 703
Jakarta 761 72.5 756 758 761 763 766 77.0 774 776
National 67.7 643 65.8 68.7 696 701 70.6 71.2 718 72.3
1 Tre purnbes inthe table cfr slightly from those in the Human Development Index cl
culated by the Unites Nations,
Source: BPS 2012)
calculations of Indonesia‘s central statistics agency, Badan Pusat Statistik
(BPS), Indonesia’ HDI score was 67.7 in 1996, declining to 64.3 in 1999 as
aresuilt of the Asian financial crisis. By 2010, however, ithad risen to 72.3,
Even more importantly, the provinces with the highest HDI growth rates
have all been areas in the Outer Islands: West Nusa Tenggara, West Kali
mantan, West Sulawesi and Gorontalo, The capital, Jakarta, by contrast,
has recorded below-average growth in its HDI score (Table 3.2),
Bat the discrepancy between scholarly assessments and public views
of decentralization has its roots not only in different readings of statistics;
italso points to the fact that Indonesians base their evaluations of decen-
tralization on factors other than just government services. Significantly,
‘Thomas Carothers has applied a similar argument to refute Larry Dia-
mond’s notion of an inseparable link between government effectiveness
and the level of support for democracy. For Diamond, dissatisfaction
with the performance of government is the main factor causing support
for democracy to decline and democratic reversals to occur (Diamond
2008). While acknowledging the ‘powerful intuitive logic’ of Diamond's
approach, Carothers nevertheless believes that it misses a large part of
citizens’ motivations to support a particular polity. For Carothers (2009:
12), people's attachment to ideals such as democracy or decentralized
governance ‘does not necessarily rest solely on the socioeconomic perfor-
mance of their government’.
The survey data certainly endorse Carother’s point. Most Indo-
nesians are disappointed with the lack of improvements in poverty alle-
viation and unemployment, and they believe that local governments fail
to inform the citizenry about their activities (Democracy InternationalThe rise of local identities ane the survival of the nation-siate St
2008: 29). Yet, three-quarters of the Indonesian population still supports
decentralization. This indicates that public backing for decentralization
~ and the stability of centre-periphery relations that results from it - are
die to factors that go beyond bureaucratic performance. ‘The following
sections look at these factors in detail.
3.3. THE NON-VIOLENT RENAISSANCE OF LOCAL IDENTITIES
Arguably, one of the reasons that the majority of Indonesians support
decentralization despite its many deficiencies concerns the renaissance
‘of local identities after 2001. To grasp the dimensions of this revival, one
needs to recall the tight management of nation building under Suharto's
centralist regime. Despite the New Order’s slogan of ‘unity in diversity’,
at the core of its policies towards the regions was an aggressive push for
uniformization. Politically, Law 5/1974 on Basic Principles of Regional
Covernment led to the de facto appointment of district heads, mayors
and governors by the central government, and Law 5/1979 on Village
Government standardized previously heterogeneous forms of village
administration. Culturally, the regime unashamedly promoted its Java-
nese values as the nucleus of Indonesian nationhood. As Flatley (1993:
50) has observed, the regime celebrated
. conservative, hierarchical values. The central state ~ Java-based and Java-
nese-dominated - supports Javanese culture of a particular type that of court
ttadition. Images of noble grandeur and hierarchical socal order serve to dis-
play ane! confirm the authority of the contemporary state,
In this regime, the government tolerated alternative expressions of local
identity as long as they came in the form of folklore, but cracked down
‘on them if they turned political
The regime party, Golkar, and the military also drove this stand-
ardization, spreading their symbols, colours and thetoric throughout
the archipelago. The archetypical small town in New Order Indonesia
‘would have as its power centres the local military command, the Golkar
building (usually painted in yellow) and the district head's office (run
by either a military officer or a bureaucrat). Importantly, most military
officers and senior bureaucrats stationed in the Outer Islands were from
Java, Indeed, it was a declared policy by army headquarters to keep non-
Javanese officers out of their home areas ~ a response to the involvement
of local military officers in the regional rebellions of the late 1950s. Simi-
larly, between 1965 and 1998, 73 per cent of all provinces outside Java
‘were ruled by Javanese or Sundanese governors for at least one term.
Half the provinces experienced more than 10 years of government by52 Marcus Mictzner
Javanese and Sundanese officials, and 23 per cent witnessed more than
15 years of external rule. In the province of West Nusa Tenggara, Java-
nese leaders ruled continuously between 1958 and 1998.
But decentralization turned Indonesia into a very different place.
The 1999 decentralization laws abolished the right of the central govern-
menteffectively to appoint local government heads, which put an almost
immediate end to the dominance of Javanese and Sundanese bureau-
crats in the Outer Islands. The 2005 introduction of direct local elections
strengthened that trend, and by 2013 only one province outside Java
(North Sumatra) was ruled by a Javanese or Sundanese, Political liberali-
zation and decentralization also terminated Golkar’s uniform hegemony
across the archipelago, In today’s polity, the political affiliations of areas
differ not only from district to district, but often from village to village
(ones 2010), indeed, the electoral rights of Indonesians have been mas-
sively expanded: since 2005, citizens have been able to elect personnel for
11 political institutions in up to seven different polls within a five-year
cycle. This in turn has encouraged democracy indexes such as Freedom
in the World to grant Indonesia the highest score in political rights in
Southeast Asia (Freedom House 2013).
‘This electoral liberalization and decentralization has resulted in a
renaissance of long-suppressed local identities. In fact, appealing to spe-
cific local identities is now a prerequisite for political leaders to com-
pete in local elections. In her ethnographic study of post-decentralization
politics in West Sumba in East Nusa Tenggara, Jacqueline Veil (2008)
described how local politicians had to command extensive knowledge of
local customs and participate regularly in rituals in order to get elected,
Incontrast, outsiders who concentrated their campaign efforts on distrib-
uuting money in a short period of time generally failed. Most importantly,
the post-2001 localization of polities has not only strengthened existing
identities, but also created new ones. In many cases, citizens located far
from the main administrative centres of their localities began to organ-
ise around their collective feelings of marginalization, and to ask for the
division (pemekaran) of the original district or province. Local elites in the
western part of West Sumba, for instance, pushed for the creation of a
new district of Central Sumba, arguing that they no longer wanted to be
dominated by the West Sumbanese capital, Waikabukak. The split was
formalized in 2007, defining a new Central Sumbanese identity fora pop-
ulation that now viewed itself as a distinct politico-cultural community.
This pattern of diversification is also mirrored in the ‘new pluri-
formity of village governance’ (Henley and Davidson 2008: 852), which
removed the institutional standardization imposed by Law 5/1979 on
Village Government. The new Law 22/1999 on Regional Government
explicitly stated that ‘Law 5/1979 ... which made the name, shape,