Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 1

No Need for Panic: Planned and Unplanned Releases of Convicted Extremists in Indonesia ©2013 IPAC 1

TERRORISM, RECIDIVISM AND

PLANNED RELEASES IN INDONESIA

4 September 2020
IPAC Report No. 66
 CONTENTS

I. INTRODUCTION..................................................................................................1
II. THE DATA...............................................................................................................1
III. WHY FORMER PRISONERS RETURN TO EXTREMISM............................3
A. Personal Factors…………...............................................................................3
B. The Case of Asrak from Bima..........................................................................4
C. The Revealing Nature of the Second Offence………….…..........................6
IV. THE THREE CLUSTERS: POSO, ACEH AND ISIS..................................... ....6
A. Poso.....................................................................................................................6
B. Lessons from the Aceh Camp..........................................................................7
C. ISIS and the Syria Conflict..............................................................................8
V. THE CRIMINAL CASES…………........................................................................8
VI. PLANNED RELEASES IN 2020 AND 2021.........................................................9
VII. CONCLUSIONS: PREVENTING RECIDIVISM..……...................................11

APPENDIX 1 :
Figure 1: RECIDIVISTS BY AREA OF ORIGIN (2002-2020)............................13
Figure 2: LENGTH OF FIRST SENTENCE FOR RECIDIVISTS.......................13
Figure 3: AGE AT RELEASE FROM PRISON AFTER FIRST OFFENCE .......13
Figure 4: NATURE OF 1ST OFFENCE..................................................................14
Figure 5: NATURE OF 2ND OFFENCE................................................................14
Figure 6: YEAR 2ND OFFENCE COMMITTED.................................................14
Figure 7: LENGTH OF TIME BETWEEN RELEASE AND COMMISSION
OF 2ND OFFENCE..................................................................................15
Figure 8: LEVEL OF MILITANCY IN PRISON...................................................15
 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 1

I. INTRODUCTION

Indonesia needs to look more closely at the patterns of recidivism among terrorism offenders,
because many are due for release in the next 18 months. Since it passed a strengthened count-
er-terrorism law after the May 2018 Surabaya bombings, Indonesia has allowed the police to
make “preventive strikes” against men and women suspected of being members of extremist or-
ganisations. Hundreds have been arrested since, including for non-violent roles – among them
hiding fugitives, withholding information, purchasing food supplies, buying airplane tickets
and attending meetings. Many have received short sentences of two to three years. This means
the prisons have become even more of a revolving door for convicted terrorists than they have
been in the past, with individuals released after minimal in-prison counselling programs and
with almost no capacity on the part of the government for sustained post-release monitoring. In
2021, more than 150 of these prisoners will be freed. Most will not commit a second terrorism
offence or related crime after release, but the challenge is to understand the factors that could
tempt individuals to re-engage with extremist organisations and the program interventions that
might dissuade them.
This report examines the cases of 94 repeat offenders among a total of 825 men and women
convicted of terrorism and released between 2002 and May 2020. Most were re-arrested after
committing a second terrorist crime, but we have also included those who joined ISIS in Syria
after their release and a few others who might not fit a strict definition of recidivism but who
clearly re-engaged with violent extremism.
The report suggests that several factors are associated with recidivism: a high level of rad-
icalism in prison, a militant spouse or other close family member who is a close contact after
release; and the availability of a powerful ideological concept that carries with it the possibility of
collective physical action. Three of these concepts have been particularly potent for recidivists:
the idea of jihad tamkin, or jihad in the service of Islamic governance, that was linked to a train-
ing camp in Aceh in 2010; the transfer of this idea to Poso, under the banner of the Mujahidin
of Eastern Indonesia (Mujahidin Indonesian Timur, MIT) from 2012 to 2016; and the call to
emigrate (berhijra) to the new Islamic State from 2014 to 2017. Without the opportunity for
action, ideological commitment and radicalised family members alone might not be enough to
convince a released prisoner to return to violence. These are not the only factors, to be sure, but
they bear particular scrutiny.
The report looks at different definitions of recidivism and why they matter. It analyses avail-
able data on repeat offenders based on one of these definitions and looks at the patterns of
re-engagement that these data reveal; and then applies these lessons to the planned releases for
the remainder of 2020 and 2021.

II. THE DATA

A widely read study published in April 2020 suggested that recidivism rates among terrorists are
actually much lower than is often assumed, and that a few high-profile cases distort the actual
situation.1 This is also true in Indonesia, where every new incident involving a former prisoner
brings anguished hand-wringing about the country’s failure to manage released prisoners effec-
tively. The one agency that consistently downplays the problem is the one with the most respon-
sibility for addressing it: the National Counter-Terrorism Agency (Badan Nasional Penanggu-
langan Terorisme, BNPT). In 2018, the then head of BNPT claimed that Indonesia had only had

1 Thomas Renard, “Overblown: Exploring the Gap between the Fear of Terrorist Recidivism and the Evidence,” CTC Sen-
tinel, April 2020.
2 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC

three terrorist recidivists since 2002 and that this was due to the 100 per cent success rate of its
deradicalisation program.2 A good article by a former prisoner on his friends who had returned
to extremism suggested the figure was closer to 80 between 2009 and 2019.3 IPAC calculates that
since the Bali bombing in 2002, at least 94 convicted terrorists, possibly more, have committed
a second terrorist offence.
To arrive at this figure, we started with information from the Corrections Directorate of the
Ministry of Law and Human Rights that between 2002 and the arrests following the Bali bomb-
ing, through the end of May 2020, 825 individuals had been arrested on terrorism charges,
convicted, and released after serving their sentences.4 That figure includes a few dozen cases
not linked to Islamist extremism, such as miscellaneous bomb threats and other crimes where
the perpetrators were charged with terrorism.5 IPAC independently maintains a database on
Islamists arrested on terrorism charges, so in most cases we have information both on the back-
grounds as well as on the various judicial procedures that the prisoners went through.
The Indonesian criminal code defines recidivism as the commission of a repeat offence, the
same as or similar to the first crime, within a specified time after release (the time varies depend-
ing on the offence, but generally five years).6
This definition would include all released terrorists who were subsequently re-arrested on
terrorism charges or killed in police anti-terrorism operations. These are the first two categories
in Table 1, below. It would include released terrorists who tried to go abroad to join ISIS-linked
groups but were caught en route, deported and arrested on arrival in Indonesia. It would not
necessarily include former prisoners arrested overseas for terrorism activities who were then
imprisoned in the country where they were caught. An example is Syarif Tarabubun, released in
2016 after ten and a half years in prison for terrorist actions in Ambon in 2005. He was arrested
in Sabah, Malaysia in 2017 for trying to get Mindanao to obtain guns and three years later re-
mained in a Malaysian prison. The above definition would also not include released prisoners
known to have later joined ISIS in Syria but who were then killed or simply disappeared from
view. Abdul Rauf, released in 2011 after serving just under ten years of a 16-year sentence for his
role in the Bali bombings, is one such case. In 2014, he became one of the first Indonesians to
be killed fighting for ISIS in Ramadi, Iraq. IPAC, however, considers both Tarabubun and Abdul
Rauf to be recidivists and includes them and similar cases in its count.7
We have also included another category that falls outside Indonesia’s definition. It consists of
the convicted terrorists who committed a second terrorism offence while in prison, leading to
a new trial and increased sentence without the prisoners ever having been released. Individuals
in this category include the sentenced prisoners (as opposed to the remand detainees) who took

2 “BNPT Klaim Program Deradikalisasi Berhasil 100 persen”, nasional.kompas.com, 22 May 2018. The three recidivists,
according to Suhardi, were Sunakim, perpetrator of the so-called Thamrin attack in Jakarta in January 2016; Juhanda, who
threw a grenade at a church in Samarinda, East Kalimantan in November 2016, and Yayat Cahdiyat, involved in a bombing
in Bandung in February 2017.
3 Arif Budi Setyawan, “Cerita Residivisme Pelaku Terorisme,” ruangobrol.id, 24 April 2020.
4 Information from Corrections Directorate on convicted terrorists released after completing their sentences, 2002 through
May 2020 as conveyed in interview, Jakarta, June 2020.
5 One well-known non-Islamist example is that of a man named Leopard Wisnu Kamala who between July and October
2015 placed four bombs around Tangerang, a Jakarta suburb, including at the Alam Sutera Mall, in an effort to extort
money.
6 Prianter Jaya Hairi, “Konsep dan Pembaruan Residivisme dalam Hukum Pidana di Indonesia,” Pusat Penelitian Badan
Keahlian DPR RI, 1 November 2018. Recidivism generally entails increased punishment.
7 IPAC does not include in its tally of recidivists those Indonesians who were first arrested outside Indonesia for terrorist
offences and who then were arrested or killed after their return to Indonesia in connection with terrorist activities. One
example would be Ahmad Sayid Maulana, first arrested in Sabah, Malaysia under the Internal Security Act in 2003 while
returning to Indonesia from training in Mindanao. He was released and returned to Indonesia in 2008, only to get involved
in the Aceh camp. He was killed in 2010 in a police operation to track down camp fugitives.
 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 3

part in a May 2018 uprising in the detention centre of the Mobile Police Brigade (Brimob) and
the inmates who in 2016 arranged to steal guns from a weapons depot inside their prison with
the intention of sending them to extremists in Poso, Central Sulawesi.
Together the individuals in these categories total 94, which then as a percentage of the 825
tried and released produces a recidivism rate of 11.39 per cent.

Table 1: Categories of Recidivists/Re-engaged 2002-2020

No Category N=
1. Tried for a second terrorist offence after being convicted and released for an earlier terror 68
offence.
2 Re-engaged in terrorism after release and then killed in police CT operations 6
3 Went to Syria or joined another jihad after conviction and release for a terrorist offence 7
4 Tried for a second terrorist offence committed in prison while serving a sentence for a 13
first terrorist offence
TOTAL 94

It is important to note that this figure undercounts the true numbers of people who pass
through Indonesians prisons only to join extremist organisations after their release. For exam-
ple, the pool of 825 only involves prisoners whose first offence was terrorism. It does not include
ordinary criminals recruited by extremists while in prison who may have been first convicted
for theft or drugs but who then joined their new friends for a jihad operation after their release
and were then arrested on terrorism charges. Between 2002 and 2020, 44 men were arrested in
this way. They deserve a separate study, but details on their backgrounds and first offence are
often too sketchy to do a proper analysis. While they are not included among our 94, we have
nevertheless included a brief section on them in this report, because the phenomenon of crimi-
nals radicalised in prison so frequently comes up in discussions of recidivism.
Of the group of 94 examined here, all are men. Women extremists only began being arrested
in significant numbers in 2018, and none of the handful arrested and released before that have
committed a second offence. (See forthcoming IPAC paper “Extremist Women Behind Bars in
Indonesia”.) In terms of geographic origin, Java and Sulawesi together account for more than 80
per cent of the cases, with Sumatra, Sumbawa and Maluku making up the rest (see Appendix 1,
Figure 1). Most were in their peak years (mid-20s and 30s) when they were released and com-
mitted their second offence within three years.

III. WHY FORMER PRISONERS RETURN TO EXTREMISM

What prompts these men to risk all and return to terrorism after serving time in prison? The fac-
tors can be both personal and ideological, but some interesting patterns emerge from the group
of 94. Family, status and reputation and the availability of an ideologically attractive movement
to join can combine to exert a powerful pull.

A. Personal Factors
The personal factors appear to be often linked to status and desire for acceptance, though the
evidence is mostly anecdotal. In several cases, the released prisoner found himself suspected by
his old comrades of “singing” to police during his incarceration and wanted to prove he was still
clean. The easiest way to do this was through taking part in a jihadi action. This was reportedly
the motivation of Bagus Budi Pranoto (Urwah) for re-engaging with Noordin Top in 2009, two
4 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC

years after his release. It is also a factor mentioned by a former prisoner in an article describing
seven of his friends from prison who returned to extremism.8
A second factor may be the presence at home after release of radical family members who
can put pressure on the newly released son, brother, or husband (and perhaps in the future, wife
or daughter) to return to violent extremism. Here the data are spotty but there are several cases.
Arman Galaxi, for example, spent more than ten years in prison for his role in the 2002 Makas-
sar bombing. He went home in 2013 to a wife who was deeply engaged in pro-ISIS activities
with former members of Laskar Jundullah. Within two years, he had left for Syria. In the case of
another convicted Makassar bomber, Anton Labbase, who had not been particularly hardline in
prison, it was his father who pressed him back into service for MIT once he got out. Darwin Go-
bel, a Poso recidivist, was also under pressure from his father.9 Several other released prisoners
had wives who were as hardline as they were, if not more so.
Status can also be also an issue, especially for those who are built up in prison as leaders of the
hardline “rejectionists” and either want to maintain that status after their release or are expect-
ed by their followers to do so.10 Ustadz Yasin, arrested for the third time on 2 September 2020,
would fall in this category.11 This observation corresponds with IPAC’s findings that 38 men out
of the 94, or 40 per cent, had been ranked as “highly ideological” by prison authorities, using
a rough index of high, medium and low to assess ideological commitment of terrorist inmates
(See Appendix 1, Figure 8).12 Those ranked “high” refuse to take part in any prison activities,
have made a concerted effort to radicalise other prisoners and refuse to request remissions or
conditional releases on the grounds that to do so would be acknowledging the legitimacy of a
non-Islamic legal system. Those ranked “medium” refuse some activities but not all and main-
tain close ties to the ideologues; ten men among the 94 fell in this category. Those ranked “low”
generally take part in all activities from flag-raising to Friday prayers in the prison mosque. A
high level of ideological commitment may lead former prisoners back to extremism especially
when another factor is present: the availability of a movement that exerts a strong ideological
pull.

B. The Case of Asrak from Bima

One striking case that combines several of these factors is that of Asrak, from Dompu, Bima.
Born in 1986, he was first arrested at the age of 25 in 2011 for his involvement in a siege at a
radical boarding school, Pesantren Umar bin Khattab (UBK) where a teacher accidentally killed
himself making a bomb. Asrak had been a member of the municipal police (pamong praja) but
left after he began to attend study sessions of Jamaah Ansharul Tauhid (JAT) where police were
portrayed as anti-Islamic oppressors. He became more radical after marrying a woman from a

8 Arif Budi Setyawan, op.cit.

9 “Darwin Gobel dan Residivisme Mantan Napiter,” PAKAR, 23 April 2020, https://www.radicalismstudies.org/wp-con-
tent/uploads/2020/04/Darwin-dan-Residivisme-Mantan-Napiter.pdf.
10 Arif Budi Setyawan, op.cit
11 Sutomo bin Sudarto alias Ustadz (teacher) Yasin was born in Semarang on 5 July 1965. He studied both accounting and
architecture after high school but never got his university degree. He was one of the first JI teachers to be sent to Poso in
late 2000 as part of the “Uhud Project” under the direction of senior JI leader Abu Tholut, designed to build a mass base
in the Poso-Palu area through religious outreach (dakwah). While teaching at the al-Amanah pesantren there, he worked
as a religious broadcaster and sold clothes for a living. He was wounded in a clash in Poso between JI supporters and
police on 22 January 2007 and turned himself in. He served most of his subsequent five-year sentence in Palu, where he
systematically recruited and radicalised ordinary criminal offenders. He was out in a few years and established two pesant-
rens in the Poso area, both called Darul Anshar, that became centres of extremist teaching. In 2012 he was arrested again
and sentenced to four years for support of Santoso’s armed group in Poso. He was released in 2016, by then a committed
ISIS supporter, and continued to use his schools as centres for recruitment and radicalisation. He was arrested again on 2
September 2020.
12 IPAC interviews, staff of Corrections Directorate, Ministry of Law and Human Rights, May 2020, Jakarta.
 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 5

radical family who had graduated from a JI-affiliated boarding school for girls, Maratus Sholi-
hah, in Bekasi, outside Jakarta.13 Several of her siblings taught at the UBK school, and the man
who blew himself up was her brother-in-law. Another brother-in-law joined MIT in Poso and
was killed there.
For his role in the UBK siege, Asrak was sentenced to three years and six months. While
he was in prison, in 2014, yet another of his wife’s brothers-in-law was killed by police at the
family home under disputed circumstances.14 Asrak was initially placed in Tangerang prison
where several other UBK prisoners were held, including the leader of group, Abrory, who was
the head of the UBK pesantren. Abrory, an extremist ideologue, was eventually transferred to a
maximum security prison on Nusakambangan after he began radicalising other inmates. Asrak
himself was transferred to a different prison in the Nusakambangan complex where he came
under the influence of a radical inmate from Cirebon named Yadi al-Hasan. Yadi persuaded him
to take the oath of allegiance to Abubakr al-Baghdadi, the ISIS leader, and by the time Asrak was
released in November 2014, he was a committed ISIS supporter.
Asrak returned to Bima where he was treated as a hero by other JAT members – now trans-
formed into the pro-ISIS organisation JAD – for having demonstrated his commitment to jihad
by never cooperating with authorities or requesting early release. At the same time, he showed
little inclination to follow in the footsteps of JAD members who had gone to fight in Poso or
attacked local police. Instead he focused on religious study through JAD meetings at the Is-
tiqomah Mosque, a known gathering place for extremists in Bima’s Penato’i neighbourhood.
He had clearly re-engaged, because in addition to the meetings, he took part in outdoor fitness
training and taught martial arts to JAD recruits, but initially, he went no further.
Then in April 2018, the spokesman for ISIS Central exhorted supporters to disrupt Iraq’s
election and warned that voting centres and voters would be the “target of our swords.”15 ISIS
supporters in Bima decided this meant they should disrupt Indonesia’s April 2019 election, too.16
Asrak and his friends stepped up their training and found a firearm that had belonged to one of
his wife’s late brother-in-laws, with the aim of carrying out an anti-election plot.
A month before the election, however, the widow of MIT’s leader Santoso, a young Bima
woman named Jumiatun, was released from prison after serving a short sentence for having
been one of Indonesia’s first female combatants in Poso. Asrak offered to marry her as his sec-
ond wife. She accepted, and they were married in April as the national election was taking place.
Plans for an attack were put on hold, briefly, while the newlyweds had a honeymoon and Asrak
went to Poso to retrieve Jumiatun’s young daughter. Before they could proceed further with
plans for an attack, Detachment 88 arrested Asrak for the second time. He was sentenced to five
years in prison in May 2020.
Many factors were likely involved in Asrak’s re-engagement with extremism, although IPAC
could not interview him directly to ask. His family, and particularly his first wife, would likely
have been a stimulus, not a deterrent, to rejoining the group, especially with the deaths of three
relatives to avenge. It is unlikely that Santoso’s widow, Jumiatun, would have accept the offer of
marriage from a known extremist if she herself did not want to continue her first husband’s fight.
The status question was also likely a factor, given Asrak’s warm reception by his former com-
rades as someone whose legitimacy was enhanced by his time in prison. And he was released

13 For more on the school, see its Facebook page, https://www.facebook.com/amsbekasi/. The founder, Yusuf Irianto, an
alumnus of the al-Mukmin pesantren in Ngruki class of 1980-81, left for Syria to join ISIS in September 2015.
14 Police, who were conducting a larger operation to round up terrorist suspects in Bima, said Nurdin was getting ready to
throw a bomb; his family said he was praying. “Sempat Lempar Bom Nurdin Tewas Saat Disergap Densus 88, tribunnews.
com, 21 September 2014.
15 Borzou Daragahi, “Iraq’s Looming Election Has ISIS Spooked,” Atlantic Monthly, 10 May 2018.
16 IPAC telephone interview with police official, Jakarta, 15 May 2020.
6 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC

from prison just as ISIS was exerting its strongest pull on supporters around the world to join
the new caliphate. Under the circumstances, it would have been hard for Asrak to pull back,
even if his initial reluctance to join in violence immediately after his release might have been a
tiny opening for a determined mentor and disengagement counsellor to build on, had one been
available.

C. The Revealing Nature of the Second Offence

The first crimes of the Indonesian recidivists do not fall into any obvious pattern. They cover
most of the major terrorist acts that Indonesia has experienced, including the first Bali bombing,
the 2002 Makassar bombing, the 2004 explosion in Cimanggis, and the 2004 Australian embassy
bombing as well as a host of failed and foiled terrorist plots, fund-raising robberies and attacks
in Ambon and Poso.
The second offences, however, fall into three large clusters: support for the Mujahidin of
Eastern Indonesia (Mujahidin Indonesia Timur, MIT) in Poso; involvement in the 2010 Aceh
training camp; and pro-ISIS activities. These three clusters account for 85 per cent of the repeat
offences (Appendix 1, Figures 4 and 5).
The clusters become clearer if we look at when the second offences were committed (Appendix 1,
Figure 6). The spikes are in 2010, the year the Aceh camp was started, broken up and produced
a wave of actions in revenge; 2014, as MIT in Poso, founded wo years earlier, proclaimed itself
the armed force of Islamic State in Indonesia; and 2016 to 2018, the peak period for hijra to Syria
and for waging war at home as travel to Syria through Turkey became more difficult. The fact
that there were only four cases of recidivism before 2010 can be explained at least in part by the
fact that there was no strong pull factor between the Bali bombs and the Aceh camp. Jemaah
Islamiyah did not produce many recidivists, except for local recruits in Poso, and many of those
getting released before 2010 were JI. Noordin Top, head of a JI splinter group, was active but the
people he attracted were either fugitives or new recruits – they were mostly not people who had
spent time in prison. Urwah, mentioned above, was an exception.

IV. THE THREE CLUSTERS: POSO, ACEH AND ISISI

The three clusters of Poso, Aceh and ISIS all represent a powerful triple whammy: a powerful
concept; the possibility, indeed the obligation for physical training or battle; and backing by
influential clerics. In an ideal world, officials would always be one step ahead of the next big
ideological idea to come along. In practice, they have their hands full taking care of the last one.

A. Poso
Poso clearly stands out as a problem area. Of the 21 cases where the first offence was Poso-relat-
ed, 16 men or 76 per cent committed their second offence also in relation to Poso,(for example,
helping with supplies for MIT) even though only half were actually from the area (10 of 21).
There are several possible explanations for why the temptation to return to fight in Poso is
so high. Poso is the only place in Indonesia where a territorially-linked jihad has taken place
continuously since 2000.17 It thus has taken on a symbolic significance out of all proportion to

17 Poso was the site of conflict between Muslim and Christian communities from 1998 to 2001, with the most intense fighting
taking place in 2000-2001, which is when extremist organisations arrived there: Jemaah Islamiyah, KOMPAK, Laskar
Jundullah and factions of Darul Islam from West Java/Banten and Makassar. A peace agreement known as the Malino
Accords ended the communal fighting in 2001 but the local affiliates of JI and KOMPAK continued terrorist attacks on
Christians.
 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 7

its size. From 2000 to 2007, when JI and its local affiliate in the neighbourhood of Tanah Run-
tuh were the paramount extremist groups, Poso was also seen as a “secure base” for the broader
struggle for an Islamic state, a place that could serve as the nucleus of an Islamic community
and which had the potential, in the form of abandoned plantations, to replace Malaysia as JI’s
economic base. At the same time, many in the local Muslim community saw Poso as the site of
unfinished business, where the peace agreement’s promised justice and prosperity was not deliv-
ered and where the desire for vengeance for lives and property lost was still high. A major clash
with the police in the streets of Poso city in January 2007 brought an uneasy peace for two years,
and JI itself decided from that point on to cease conducting any form of violent jihad on Indo-
nesian soil, a decision it has stood by to this day. Abu Bakar Ba’asyir’s Jamaah Anshorul Tauhid
(JAT), founded in 20 08, opened a branch with a military wing in Poso in late 2009, drawing in
many ex-JI, and after the collapse of the Aceh camp in 2010, JAT-Poso stood ready to further the
aim of a regional training centre that would again be a secure base for an Islamic state. It was the
military wing of JAT-Poso, led by Santoso, that developed into MIT in late 2012 or early 2013.18
The Poso recidivists among the group of 94 represent three groups: pre-2007, post-Aceh to
the emergence of ISIS (2010-2013), and pro-ISIS (2014-2020). For each, Poso represented an
attractive ideological cause, but in many cases there also would have been strong personal rea-
sons to re-engage, including peer pressure.19 On 15 April 2020, two former prisoners arrested
for involvement in MIT were killed by police after rejoining MIT almost immediately upon
their release. Amirudin alias Aco Gula Merah, had been released in 2018 after serving a six-year
sentence; Darwin Gobel was released in 2019 after serving a two-year sentence. He was 19 at the
time of his death.
Under the circumstances, the surprising fact is not how many became repeat offenders but
how many more decided to disengage.20

B. Lessons from the Aceh Camp

The constellation of recidivists around the 2010 Aceh training camp, like Poso, looms large both
as a first and second offence among recidivists. Many people who had never been arrested before
were caught and prosecuted for having taken part in this training camp that organisers hoped
would bring all extremist groups in the country together and serve as the nucleus of an Islamic
community. More than 100 people were arrested after the camp was discovered and broken up
by police in early 2010. It also, however, exerted such an attraction that 15 ex-prisoners (12 ar-
rested for terrorism and three former criminal offenders), including those who had been model
inmates and fully cooperative with prison authorities, were willing to take part. Participation in
the Aceh camp thus became the second offence that turned them into recidivists.
In the Aceh case, the powerful idea was jihad tamkin. As conceived by Jordanian scholar
Abu Muhammed al-Maqdisi, then translated and promoted by Aman Abdurrahman, it was a
rejection of indiscriminate bombing against the enemy in support of a more strategic approach
that saw jihad in the service of Islamic governance. Noordin Top and his bombing of luxury
hotels in Jakarta achieved little, because he only aimed at creating fear, whereas Indonesian

18 For more on this period see International Crisis Group, “How Indonesian Extremists Regroup”, Asia Report No.228, 16
July 2012, pp.6-8.
19 Seven of the first-offence Poso cases were carried out by members of Jemaah Islamiyah or its local affiliate in Poso’s
Tanah Runtuh neighborhood in the period up to 2007, including several arrested in connection with the January 2007
clash. Only three of the men were Poso natives, but all committed their second offence there, several almost as soon as
they got out.
20 Julie Chernov Hwang, Rizal Panggabean and Ihsan Ali Fauzi, “The Disengagement of Jihadis in Poso, Indonesia,” Asian
Survey, Vol.53, No.4, pp.754-777, 2013 and Julie Chernov Hwang, Why Terrorists Quit: The Disengagement of Indonesia
Jihadists, Ithaca,NY, 2018.
8 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC

militants, Aman argued, should be working to establish a community that could enforce Islamic
law. Where better than in Aceh, which since 2001 was the only province in Indonesia allowed to
apply some aspects of shari’ah? 21 The idea came at a critical time, as the Islamic State in Iraq (ISI)
of Abu Mus’ab Zarqawi was moving in just this direction. It had the support of Indonesia’s best-
known ideologues – Abu Bakar Ba’asyir, who was willing to use his network for fund-raising,
and Aman himself – and one man whose militant credentials were second to none: Dulmatin,
one of the Bali bombers, who had secretly returned from Mindanao in 2007. It was a collective
enterprise, involving virtually every salafi jihadi organisation in Indonesia except for JI.

C. ISIS and the Syria Conflict

The establishment of a new caliphate in Syria and Iraq was biggest idea of all, one that constitut-
ed a huge draw for Indonesians, including for former prisoners. It was backed not only by lead-
ing clerics like Ba’ashir and Aman Abdurrahman but also by Islamic prophecies. The conflict in
Syria heralded the Islamic Armageddon, the final battle at the end of time, al malhamah qubra,
in which Islam would be victorious, and Islamic State was the caliphate of the Prophet, khilafah
ala minhajul nubuwah, that would then appear. At such, it exerted a powerful pull, including on
prisoners just out or due to be released, as well as on prisoners released long before.
There was also the pull of what ISIS propaganda portrayed as a functioning Islamic state
where Islamic law was fully and perfectly applied. This was also a powerful attraction, particu-
larly for families with children.
Of the 94, 43 were linked to pro-ISIS activities, including the cases of the 13 men arrested for
committing crimes in prison. Twelve tried to join jihads overseas. Of these, three were arrested
trying to join pro-ISIS groups in the Philippines (one arrested in central Mindanao, two in Ma-
laysia trying to leave for Mindanao from Sabah). One, Hari Kuncoro alias Uceng, was arrested in
Indonesia after being stopped at the airport in 2019 trying to leave for Afghanistan. The others
tried to leave for Syria and five successfully did so. Of those four were killed and the fate of one,
Dr Syarif Usman, remains unknown.22
Those who elected to stay in Indonesia and wage war at home engaged in a wide range of
crimes as their second offence, from attacks on police to bomb plots. Of the 38 men with “high”
ideological commitment according to the correction officials’ index, 26 were involved in pro-
ISIS crimes as their second offence.

V. THE CRIMINAL CASES

As noted above, the Corrections Directorate’s list of 825 only includes those tried for terrorism
offences and released. It does not include those whose first offence was an ordinary crime, and
they are therefore not included in the group of 94. But the 44 men who were arrested first for a
criminal offence, often drugs or petty theft, who then went on to become radicalised in prison,
are still worth noting. The data are incomplete, mostly because in many cases we lack detailed
information about the nature of their first crimes. Arrests for minor drug offences, motorcycle
thefts, or purse-snatching are not newsworthy, and it is often only possible to even know about
first offences in the context of material that comes out at trials for the second crimes. Generally,
however, they become drawn to the extremist side because of appeals to repent; availability of
more resources among the terrorists; or appreciation on the part of the latter for criminal skills,

21 For more on the background to Aceh, see International Crisis Group, “Jihadi Surprise in Aceh,” Asia Report No.189,
Jakarta/Brussels, 20 April 2010.
22 Those killed were Abdul Rauf, Syaiful Bahri alias Apuy; Arman Galaksi, and Sholeh.
 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 9

access to weapons and familiarity with violence.

Until communications and movement of convicted terrorist inmates was substantially tight-
ened in 2017-8, intermixing of terrorists and criminals was common, with radicalisation of
some of the latter a frequent result. Association with the terrorists could provide protection
and status, particularly for a group of ordinary criminal offenders known as “lost children” –
those never visited by family members, with no access to extra food, cash or cigarettes that
other prisoners receive and that are vital to survival inside poorly resourced prisons.23 When the
hardcore extremists in the maximum security prisons on Nusakambangan island were moved to
single-occupancy cells in 2018, they left behind many ordinary criminals who had joined them
and who were now in a position to recruit others. This means, among other things, that radical-
isation in prison takes place not just from terrorists to criminals but from radicalised criminals
to other criminal inmates.
Suherman alias Herman is one case of a “lost child.” A Cirebon man born in 1987, he was
given a sixteen-year sentence in 2007 for killing a mosque watchman in a fit of rage when the
watchman tried to stop him cuddling his girlfriend. He was released in 2014 but re-arrested in
2018 for several pro-ISIS activities. In prison, Herman had been taken under the wing of the
extremists arrested for their role in the 2011 bombing of the Cirebon police mosque. Herman
had begun to study religion anyway, believing it could help him control his emotions. With the
arrival of the Cirebon group, he not only had new teachers, he had something approaching a
new family. They introduced him to writings and audio lectures of Aman Abdurrahman and he
became a committed follower. After he was released, he joined JAD Cirebon and in 2016 swore
an oath of allegiance to Abubakr al-Baghdadi. The same year he married a radical widow who
was the elder sister of a JAD Cirebon member from Brebes named Rajendra. In May 2018 after
the uprising at Brimob headquarters, Herman wanted to heed the call from ISIS Central to assist
the detainees in the prison takeover. He joined his brother-in-law and another JAD member,
then travelled to Tasikmalaya where they joined a JAD group there in preparing Molotov bombs
to aid the rioting inmates. Police got wind of the plan and arrested several of the plotters but
Herman and Rajendra got away. As fugitives, they stepped up attacks, hacking a policeman in
Brebes with a machete on 11 July 2018, then attacking two traffic police on 22 August 2018,
killing one. Rajendra and another friend were tracked down and killed by police. Herman was
eventually captured, tried and sentenced to death on 13 October 2019.24
Others like Herman who are released after being radicalised have a high potential for recid-
ivism, but they will likely not be monitored closely – or be included in the government’s recid-
ivism statistics.

VI. PLANNED RELEASES IN 2020 AND 2021

By the end of 2020, more than 120 convicted terrorists will have been released. Of these, most
will be released from the prison system under the direction of the Corrections Directorate, but
this does not include many convicted prisoners in police custody at Jakarta Metropolitan Police
Headquarters (Polda Metro Jaya) or Cikeas, the remand centre south of Jakarta which houses
some recently convicted prisoners. In 2021, over150 are due for release from the prison system,

23 See IPAC, “Update on Indonesian Pro-ISIS Prisoners and Deradicalisation Efforts,” Report No.34, 14 December 2016,
pp.9-11.
24 Death sentences for terrorism are relatively rare. Three of the Bali bombers were judicially executed in 2008, and three
other prisoners remain on death row: Ahmad Hasan and Iwan Dharmawan alias Rois, involved in the second Bali bombing
in 2005, and Aman Abdurrahman, accused of having masterminded the 2016 Jakarta attack from his prison cell. IPAC
is opposed to the death sentence as a matter of principle and believes all these sentences should be commuted to life in
prison.
10 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC

again not counting those in police custody. (There are some slight discrepancies between IPAC
and government figures due to differences in definition or calculation in some cases, but they
are not significant.)
If we look at the number of convicted terrorists due for release in 2020-2021, 83 were arrest-
ed in 2018 or later, with most getting two- or three-year sentences. When asked why they even
bothered with expensive prosecution and trial procedures if many of the suspects were going to
be released within two years or less, one police officer said, “This way we have control over them,
even for a short time, so we have a possibility to change them.”25 There is no system in place,
however, for an evaluation of whether any change has taken place.
The short sentences reflect the increasing tendency, especially since the new anti-terrorism
law went into force, of police making what they call “preventive strikes”, arresting individuals
involved in extremist organisations who have not committed violent acts. They may have taken
part in activities such as outdoor fitness training, disseminating ISIS propaganda, hiding a fugi-
tive, or playing a minor role in helping arrange travel to Syria. In one case, a newly released but
high-risk prisoner – already a recidivist – was re-arrested, tried and given a one-year sentence,
simply for having been in contact with former JAD friends. The re-arrest hardly seems worth it.
BNPT programs are largely ineffective, partly because they mostly draw in prisoners who
have already decided on their own to disengage (if indeed they ever used violence in the first
place). In one study of released prisoners, mostly from JI, personal factors such as the birth of
a first child, disillusionment with leaders or tactics, and individual cost-benefit analyses were
found to be more important than any government interventions.26 For some of the more ideo-
logically-inclined pro-ISIS prisoners, exposure in prison to arguments developed by al-Qaeda
clerics against ISIS have been persuasive, if delivered through appropriate channels. Some astute
police officers have also exploited divisions within extremist groups to woo one set of inmates
away from the rest.
The BNPT programs are voluntary, and many prisoners are not interested. They also focus
more on instilling loyalty to the Indonesian state than on giving prisoners, while still in deten-
tion or after release, new goals, marketable skills or access to new networks. In 2018, of 630
released prisoners eligible for post-release programs, only 325 elected to take part. 27 These were
the men and women who had signed an oath of loyalty to the NKRI. The problem is that just as
BNPT and some local initiatives – often better focused, with better knowledge of the economic
and political context – try sporadically to help, extremist organisations target the same individ-
uals, offering assistance and trying to ensure their ideological commitment stays high. These
organisations include the 1,000 A Day Movement (Gerakan Sehari Seribu, GASHIBU), the Alif
Centre, the Anfiqu Centre and other organisations that raise money online or in mosque meet-
ings to help detained or just-released extremists, thereby creating a moral debt.
Anfiqu Center is a case in point. It was founded by ISIS supporters from Solo and managed
by Alfandy Suko Andrenanto, son of a convicted terrorist serving his sentence in Pasir Putih,
one of Nusakambangan’s maximum security prisons. Anfiqu ran many programs, including a
house in Cilacap (the nearest city on the mainland of Java to the prison complex) for relatives
of detained extremists, where they could stay overnight before going to visit family members,
and a van service, Samudera Travel, to assist them with transport to and from their homes. In
addition, Anfiqu provided families with small amounts of cash assistance, about Rp.500,000
(US$34)a month; helped pay for medicines if family members became ill; helped with expenses
for celebrations at the end of the fasting month; and occasionally helped with small amounts of

25 IPAC interview with police officer, Jakarta, 19 May 2020.

26 Julie Chernov Hwang, Why Terrorists Quit: The Disengagement of Indonesian Jihadists, Ithaca, 2017, pp.50-78.
27 “Dari 630 eks napiter, baru 325 mengikuti program deradikalisasi,” merdeka.com, 26 May 2018.
 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 11

capital for micro-business projects. Given the attention paid to the families by these pro-ISIS hu-
manitarian organisations, it is small wonder that in many cases, BNPT and local NGOs looked
ineffective by comparison.
The question for Indonesian authorities is whether changing procedures – more draconian
detention regimes and shorter sentences – will have any impact on the recidivism rate. It would
be worth setting up a study now within the Corrections Directorate to collect data.

VII. CONCLUSIONS: PREVENTING RECIDIVISM

It is important to underscore that while recidivism is a bigger problem than BNPT has been
willing to admit, it is not as big as many in the public or even in the police assume. Almost 90
per cent of prisoners convicted of terrorism will not commit a second offence after their release.
This paper suggests that some of the factors involved for those who do is the radicalism of their
immediate family members; their ideological commitment and status in prison; and the avail-
ability of a movement once they are out to join that combines a powerful idea with the oppor-
tunity for action.
Of these factors, there is little that authorities can do about the last, but the first and second
have more possibilities. One suggestion from a detained terrorist is that if the family is known to
be radical, the individual in question should be detained in a prison as far away from the family’s
home as possible, but this could just encourage extremist charitable organisations to come to the
rescue with funding for visits. A better strategy, which is already being applied in some areas,
is for prison authorities to get to know family members who come for visits and see if there are
any opportunities for identifying needs or opportunities – such as the offer of medical assistance
to a sick child – that might prompt a change in attitude. If prison officials notice an opening
in the case of a hitherto hardline inmate, they can offer longer visits or more opportunities for
communication as an incentive to cooperate. These efforts tend to be ad hoc, however, and
depend on the skills, experience and interest of individual prison staff in their charges. At the
very least, the organisational affiliations of an inmate’s immediate family should be mapped
while the prisoner is still in custody so that scarce monitoring resources can be better deployed
once an inmate with known radical relatives is released.
On the issue of ideological commitment, there may be scope for a more systematic counsel-
ling program, using texts that directly challenge ISIS arguments with targeted discussions led by
clerics with religious credentials seen as legitimate by the inmates in question. Such initiatives
take place now, but more on an ad hoc basis, and often only after a prisoner has given some in-
dication of openness.
The limited number of geographic areas that produce recidivists suggests the need for a more
focused set of rehabilitation activities, targeted in particular areas. Particular resources need to
go to into building a different pattern of police-community relations in these areas, since dis-
trust of police runs particularly high in known radical neighbourhoods.
One of the most important tasks is to closely monitor the activities of the men and women
arrested in 2018 under the letter or spirit of the new anti-terrorism law whose sentences are now
coming to an end. It will be important to assess whether after such short sentences, the released
prisoners are coming back more ideologically committed, less committed or the same as when
they went in. If incarceration produces no positive change, then it might be worth considering
non-custodial sentences, but with reporting requirements, for relatively minor and non-violent
crimes.
12 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC

Finally, as professional capacity increases within Detachment 88 to monitor social media,

it will be important to look for signs of the next big attraction, ideologically or geographically,
although the appeal of ISIS, even if on the wane, is not going to disappear any time soon.
 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 13

APPENDIX 1
Figure 1: RECIDIVISTS BY AREA OF ORIGIN (2002-2020)

n=94
80
58
60
40
18
20 10
4 4
0
JAVA SUMATRA SULAWESI SUMBAWA, NTB MALUKU

n=94

Figure 2: LENGTH OF FIRST SENTENCE FOR RECIDIVISTS

n = 94
35 32
30
25 19
20 15
15 10
10 6 6
3 3
5
0
up to 3 yrs 3-5 yrs 5-7 yrs 7-9 yrs 9-11 yrs 11-13 yrs over 13 no data
(0-35 mos) (36-59 (60-83
mos) mos)

Figure 3: AGE AT RELEASE FROM PRISON AFTER FIRST OFFENCE

n = 94
50 42
26
12 8
4
0
16-25 26-35 36-45 46 and over no data

Most of those who committed a second offence were in their peak years when released from prison.

1
14 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC

Figure 4: NATURE OF 1ST OFFENCE

n = 94
25 21
20
15 13
11
10 8
6 5 4 3 3 4 3 4 3
5 2 1 1 1 1 1
0

Figure 5: NATURE OF 2ND OFFENCE

30
25
25
20 18
13 14
15 12 12
10
5
0
Poso support 2010 Aceh camp General pro-ISIS Hijra or Crimes inside Other
attempted hijra prison

Figure 6: YEAR 2ND OFFENCE COMMITTED

25
21
20
15
15 12
10
10 8 8
6
4 4 4
5 2 1
0
Before 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
2010

2
 Terrorism, Recidivism and Planned Releases in Indonesia ©2020 IPAC 15

Figure 7: LENGTH OF TIME BETWEEN RELEASE AND COMMISSION OF 2 ND OFFENCE

n = 94
40 32 35

20 9 9
4 5 N = 94
0
0-1 yrs 2-3 yrs 4-5 yrs 6-7 yrs 8+ yrs no data

Figure 8: LEVEL OF MILITANCY IN PRISON

n=94
38
40
28
30
18
20
10
10

0
high medium low no info

3
 INSTITUTE FOR POLICY ANALYSIS OF CONFLICT (IPAC)

The Institute for Policy Analysis of Conflict (IPAC) was founded in 2013 on the principle
that accurate analysis is a critical first step toward preventing violent conflict. Our mission
is to explain the dynamics of conflict—why it started, how it changed, what drives it, who
benefits—and get that information quickly to people who can use it to bring about positive
change.
In areas wracked by violence, accurate analysis of conflict is essential not only to peaceful
settlement but also to formulating effective policies on everything from good governance
to poverty alleviation. We look at six kinds of conflict: communal, land and resource, elec-
toral, vigilante, extremist and insurgent, understanding that one dispute can take several
forms or progress from one form to another. We send experienced analysts with long-es-
tablished contacts in the area to the site to meet with all parties, review primary written
documentation where available, check secondary sources and produce in-depth reports,
with policy recommendations or examples of best practices where appropriate.
We are registered with the Ministry of Social Affairs in Jakarta as the Foundation for Pre-
venting International Crises (Yayasan Penanggulangan Krisis Internasional); our website
is www.understandingconflict.org.