The East-West Center is an education and research
organization established by the U.S. Congress in 1960
to strengthen relations and understanding among
the peoples and nations of Asia, the Pacific, and the
United States. The Center contributes to a peaceful,
prosperous, and just Asia Pacific community by
serving as a vigorous hub for cooperative research,
education, and dialogue on critical issues of common
concern to the Asia Pacific region and the United
States. Funding for the Center comes from the U.S.
government, with additional support provided by
private agencies, individuals, foundations, corporations,
and the governments of the region.
East-West Center Working Papers are circulated for
comment and to inform interested colleagues about
work in progress at the Center.
For more information about the Center or to order
publications, contact:
Publication Sales Office
East-West Center
1601 East-West Road
Honolulu, Hawai‘i 96848-1601
Telephone: 808.944.7145
Facsimile: 808.944.7376
Email: ewcbooks@EastWestCenter.org
Website: www.EastWestCenter.org
E A S T - W E S T C E N T E R W O R K I N G P A P E R SE A S T - W E S T C E N T E R W O R K I N G P A P E R SE A S T - W E S T C E N T E R W O R K I N G P A P E R SE A S T - W E S T C E N T E R W O R K I N G P A P E R SE A S T - W E S T C E N T E R W O R K I N G P A P E R S
Politics, Governance, and Security SeriesPolitics, Governance, and Security SeriesPolitics, Governance, and Security SeriesPolitics, Governance, and Security SeriesPolitics, Governance, and Security Series
No. 20, October 2009
Stephan Haggard is the Lawrence and Sallye Krause
Professor at the University of California, San Diego,
Graduate School of International Relations and Pacific
Studies.
Marcus Noland is Deputy Director and Senior Fellow at the
Peterson Institute of International Economics and Senior
Fellow at the East-West Center.
East-West Center Working Papers: Politics, Governance,
and Security Series is an unreviewed and unedited
prepublication series reporting on research in progress. The
views expressed are those of the author and not necessarily
those of the Center. Please direct orders and requests to the
East-West Center's Publication Sales Office. The price for
Working Papers is $3.00 each plus shipping and handling.
Repression and Punishment in
North Korea: Survey Evidence of
Prison Camp Experiences
Stephan Haggard and Marcus Noland
Repression and Punishment in North Korea:
Survey Evidence of Prison Camp Experiences
Stephan Haggard
Graduate School of International Relations and Pacific Studies,
University of California, San Diego
Marcus Noland
East-West Center and
Peterson Institute for International Economics
October 5, 2009
Abstract
The penal system has played a central role in the North Korean government’s response to
the country’s profound economic and social changes. Two refugee surveys—one
conducted in China, one in South Korea—document its changing role. The regime
disproportionately targets politically suspect groups, particularly those involved in
market-oriented economic activities. Levels of violence and deprivation do not appear to
differ substantially between the infamous political prison camps, penitentiaries for felons,
and labor camps used to incarcerate individuals for misdemeanors, including economic
crimes. Substantial numbers of those incarcerated report experiencing deprivation with
respect to food as well as public executions and other forms of violence. This repression
appears to work; despite substantial cynicism about the North Korean system, refugees
do not report signs of collective action aimed at confronting the regime.
Such a system may also reflect ulterior motives. High levels of discretion with
respect to arrest and sentencing and very high costs of detention, arrest and incarceration
encourage bribery; the more arbitrary and painful the experience with the penal system,
the easier it is for officials to extort money for avoiding it. These characteristics not only
promote regime maintenance through intimidation, but may facilitate predatory
corruption as well.
Keywords: Prison camps, political repression, North Korea, refugees
This research was supported by the Smith Richardson, MacArthur, and Korea
Foundations. We thank Nick Eberstadt and David Hawk for extensive comments on an
earlier draft, Jennifer Lee for diligent research assistance and Dan Pinkston and Chung
Tae-un for their assistance in conducting the South Korean refugee survey.
1
In the last decade, a variety of sources have allowed analysts to piece together an
increasingly clear picture of North Korea’s penal system. Technology has played a
surprising role in this process; satellite imagery, including images secured through
providers such as DigitalGlobe and Google Earth, have permitted a precise mapping of
the country’s gulag. Yet at the core of our understanding is information provided by
refugees, including both prisoners and guards, who have managed to flee North Korea.
This information has appeared in the form of memoirs (Kang 2001), unstructured
interviews (Hawk 2003; Muico 2007; KINU 2009), and databases of individual cases of
human rights violations (Database Center for North Korean Human Rights 2008). The
portrait that emerges is of a Soviet-style gulag characterized by an arbitrary judicial
system, an expansive conception of crime, and horrific abuses. These abuses include
extreme deprivation, particularly with respect to food and medical treatment, torture and
public executions.
The penal system is by no means limited to the political prison camps, however.
The evolution of the prison system also cannot be understood without reference to
profound economic and social changes that have occurred in North Korean over the last
decade and the government’s repressive response to them. During the 1990s, famine
killed between 600,000 and 1,000,000 people, 3-5 percent of the population (Haggard
and Noland 2007). While authorities blamed the collapse of the food economy on
weather, the famine was a classic case of state failure. As the state proved unable to
provide food through socialist distribution networks, the economy underwent a process of
marketization from below. Small-scale social units—households, factories and
2
cooperatives, local government and party offices, even military units—began engaging in
entrepreneurial behavior—much of it technically illegal—in order to survive.
This unplanned and unwanted marketization eroded state control of the economy
and therefore over pathways to wealth, prestige, and ultimately power. Not surprisingly,
the regime’s response to this process has been ambivalent. At times, the government has
acquiesced to the facts on the ground through reforms that decriminalized market activity.
At other times it has sought to turn back the clock and reconstitute the socialist system
through a revival of the state sector and the imposition of controls on private activity.
The penal system has played a central role in the government’s repressive
response to economic and social change. During the famine, the state established an
extensive system of low-level labor training facilities (ro-dong-dan-ryeon-dae) to
manage the unprecedented movement and market activity that sprung up as heavilyaffected
segments of the population literally wandered the countryside in search of food
(Noland 2000). A 2004 legal reform regularized these facilities and specified “labor
training” for up to two years as punishment for a wide variety of crimes, including a
growing number of economic and social crimes (Han 2006). Crossing the border into
China was always a very serious offense. As the number of refugees fleeing to China
increased, these facilities also played an important role in managing those captured in
route or repatriated by Chinese authorities.
As the state has attempted to reassert control over society in the decade since the
end of the famine, the penal system has evolved accordingly. Two refugee surveys, one
conducted in China, the other in South Korea, suggest a system characterized by high
3
rates of arbitrary detention and release. Horrific abuses are characteristic not only of the
camps for political prisoners, but at all levels of the penal system.
In a survey of more than 1,300 refugees conducted in China between August 2004
and September 2005, roughly one-quarter reported having been arrested in China and
repatriated to North Korea and nearly 10 percent of the respondents reported
incarceration in correctional and political detention facilities. Among this latter group, 90
percent reported witnessing forced starvation, 60 percent deaths due to beating or torture,
and 27 percent executions. A second survey of 300 refugees conducted in South Korea in
November 2008 confirmed these results. Just over one-third had been detained in the
penal system and similar numbers reported witnessing extreme forms of abuse.1
Such surveys are susceptible to self-selection bias: refugees may leave precisely
because of the intensity of their ill-treatment and disaffection. Those who undertake the
risks of trying to leave North Korea probably have some otherwise unobserved individual
characteristics that differentiate them from the rest of the population. These may include
more adverse life experiences and more severely truncated opportunities, both of which
could give rise to behaviors and attitudes that are quite different from the population as a
whole. The survey presented below may thus accurately capture the experiences of the
refugee community in South Korea, but may provide a limited perspective on North
1
The more secure legal environment in South Korea enabled us to administer a longer
questionnaire asking more detailed questions about initial arrest and detention, the precise
facilities in which they were held, and the conditions they witnessed while incarcerated.
The experiences of this second sample largely confirm that of our first survey, but also
provide more detailed information on both constant and changing features of the North
Korean police state.
4
Korea.2
However, there are some reasons to believe that the sources of bias are somewhat
less pronounced than might be thought. Refugees are asked questions not only about their
own experience but their observation of others’ experiences as well. Moreover, there is
strong evidence that the punishment of border-crossing is now being treated in ways that
resemble a widening array of other economic and social crimes that are associated with
the process we describe as “marketization from below.” The penal system appears to
process large numbers of people engaged in illicit activities for relatively short periods,
exposing inmates to terrible abuses. This pattern not only serves to intimidate; other
research we have conducted on the pervasiveness of corruption suggests that abusive
treatment may also benefit corrupt officials extracting bribes from those seeking to avoid
entanglement with the penal system.
The repressive apparatus also appears to work. Our surveys uncover deep
dissatisfaction with the North Korean regime. Respondents are also highly cynical about
the regime’s arguments that controls are necessitated by the country’s adverse security
environment. Yet the surveys also reveal a highly atomized society in which barriers to
collective action are profound.
We begin with a brief overview of the North Korean penal system, the variety of
different facilities and their administration and purposes. We then turn to a descriptive
overview of respondents’ experiences with the penal system. A striking finding is that the
2
A second, more tractable issue is that the population of refugees may not be
demographically representative of the resident, non-refugee population, over-representing
particular segments of the population such as women, or people from particular
occupational categories. The problem can in principle be addressed ex post using
multivariate techniques, as done in Chang, Haggard, and Noland (2009) and Haggard and
Noland (2009a) where this source of potential bias in these surveys was shown to be
negligible.
5
conditions that are frequently seen as characteristic of the country’s infamous gulag of
political penal-labor colonies--such as extreme deprivation and exposure to violence--in
fact pertain across the penal system, from the penitentiaries designed to house felons to
lower-level jails and the new labor training facilities.
We then explore some of the determinants of incarceration. There is some
evidence that this repressive apparatus disproportionately targets politically contestable
groups, the “wavering” class in North Korean parlance, and particularly those involved in
economic activities beyond direct state control. Sadly, as discussed in the penultimate
section, this repressive apparatus appears to pose an effective deterrent to collective
action although certainly not to disaffection.
The North Korean Penal System
As would be expected of a highly repressive regime, the North Korean legal and penal
system is differentiated and complex.3
In the first instance, political crimes are treated
very differently from ordinary crimes, with very much more severe punishments. Yet
since the onset of the famine and increasing marketization of the economy, the legal and
penal system has also had to cope with an explosion of economic and social crimes that
are peculiar to a state-socialist system. These crimes ultimately reflect the inability of the
official state sector to provide employment and basic necessities, including food. Some of
these crimes involve theft or diversion of state property and assets; others simply reflect
3
The Korean Institute for National Unification (KINU)’s White Papers on Human Rights
in North Korea provides the most consistent and detailed information on changes in the
legal and penal system.
6
the efforts on the part of households, and even work and military units, to engage in
various income-earning activities.
From the perspective of the regime, however, even these latter activities can pose
political as well as economic challenges. For example, people engaged in unauthorized
private enterprise and trading do not show up at their work units. Their activities are also
difficult to tax. As a result, authorities have incentives to punish both failure to appear at
work and the associated private activity. Illegal movement is also an important feature of
the new economy, including leaving one’s home without appropriate travel permits,
overstaying travel permits or leaving the country without authorization. This last offense
is of obvious significance to our consideration of refugees, as many of them seek to leave
North Korean and are either caught in transit or are forcibly repatriated by Chinese
authorities.
More serious economic crimes include diversion of state output to private use,
including food grown on cooperatives, and the illegal use, profiteering from, or even sale
of state assets. Some of these crimes are managed through increasingly institutionalized
administrative punishments, but others are handled through the criminal justice system.
In the 1999 criminal code, the “Chapter on Offenses against the Management of the
Socialist Economy” included eight articles. In 2004, it was re-titled “the Chapter on
Offenses against the Management of the Economy” and included seventy-four. It is
noteworthy that this revision of the criminal code came only two years after the initiation
of the most significant reform effort in the post-1990 period.
The most notorious component of the North Korean prison system is the massive
kwan-li-so, variously translated as political prison camps, labor colonies, or concentration
7
camps; we will refer to them as political penal-labor camps. The number of these camps
has recently been consolidated from 14 to about five large sites; one camp (Camp 22 near
the Chinese border) is estimated to be 31 miles long and 25 miles wide and to hold
50,000 inmates (Harden 2009). With one exception, these camps are administered by the
National Security Agency (NSA), an agency with wide-ranging external and internal
security functions that include border and immigration control.4
The political penal-labor camps are reserved for those deemed to pose a direct
political threat to the regime. This group initially included counter-revolutionary social
forces such as landlords, the religiously active and members of purged political factions.
Over time, it came to encompass anyone guilty or suspected of political or ideological
crimes. Those with extensive knowledge of life outside Korea have been particularly
vulnerable to incarceration in these facilities, including repatriated Japanese-Koreans,
those who have studied abroad and those accused of “trafficking” people out of North
Korea. When asked if they were aware of the kwan-li-so, 77 percent of the respondents in
the South Korean survey answered affirmatively. When asked if they thought that those
sent there were incarcerated justly, 93 percent responded “no.”
A distinctive feature of the management of political crimes is that there is little
pretense of due process. Political crimes appear to fall outside of criminal statute
altogether and are managed with a high level of discretion by the NSA. The NSA either
apprehends those accused of political crimes directly or they are remanded to NSA
4
A useful summary of the history and functions of the NSA can be found in Namgung
Min, “What Kind of Organization is North Korea’s National Security Agency?”
DailyNK, September 9, 2007
at http://www.dailynk.com/english/read.php?cataId=nk00400&num=2645.
8
custody; the NSA operates its own interim detention centers, including several dedicated
to those caught attempting to cross the border or those caught and repatriated by Chinese
authorities (Muico 2007). Prisoners accused of political crimes are detained in these
interim facilities and tortured to extract confessions.
Once a case is deemed political, the National Security Agency also assumes
control of the prosecutorial process. A prosecutor from the NSA will hand down
sentences in a closed-door local court session in the name of the Central Court in
Pyongyang, including the decision of whether to imprison the criminal for life and
whether the criminal’s family will also be sent with him; group punishment in the form of
incarceration of extended family and confiscation of property is a distinctive feature of
the management of political crimes and incarceration in the political penal-labor camps.
Inmates of the kwan-il-so are typically incarcerated under prolonged or lifetime
sentences at hard labor in mining, logging, and farming enterprises in the highly
inhospitable north and north-central part of the country.5
Prisoners are also kept on
starvation rations and many die of malnutrition and disease. Not surprisingly, the number
of escapees from these camps is small; of the 300 refugees interviewed in the 2008 South
Korea-based survey, only three report internment in a kwan-il-so.6
5
. Only in two kwan-il-so is there reported evidence of efforts at political re-education.
6
. Unfortunately, we have no information on the conditions under which these three were
either released or escaped, but curiously the length of their incarceration does not differ
significantly from that of the other respondents: one reported incarceration between one
and five years, one of less than a year, but one reported being in a kwan-il-so for less than
a week. It is clearly difficult to draw inferences from three respondents, although the
survivor testimony from these camps on these issues is now fairly extensive. See
particularly Hawk 2003.
9
A second component of the penal system is the kyo-hwa-so--literally, a “place to
make someone better through education”--and sometimes translated as correctional or reeducation
centers. In fact, there is little evidence from other refugee testimony that these
facilities perform correctional or re-education functions. Superficially, they resemble
prisons for housing felons and we will refer to them as penitentiaries. Prisoners in the
kyo-hwa-so penitentiaries are presumably arrested for violations of the DPRK criminal
code,7
subjected to a judicial process and given fixed-term sentences, which are
described as “limited” if they are under 15 years and “unlimited” if longer than 15 years.
The kyo-hwa-so are administered by the People’s Safety Agency, the national police
rce.
al
the
-so
fo
However, there are predictable differences with penitentiaries in other pen
systems. First, the definition of felony crimes in North Korea includes a range of
activities which appear political rather than criminal: “anti-state, anti-people crimes,”
“crimes injurious to socialist culture,” and so on (Table 1). Hawk (2003, 46) describes
case of a woman imprisoned in a kyo-hwa-so penitentiary who had been convicted of
disturbing the “socialist order” for singing a South Korean pop song in a private home.
As in the political penal-labor camps (kwan-li-so), prisoners in the kyo-hwa
penitentiaries are compelled to perform hard labor. Satellite imagery and refugee
testimony reveal that they are typically maximum-security compounds that combine
buildings housing prisoners and administration with work units; some are located near
mine faces. Refugees with experience in them report that they are subjected to brutal
7
In addition to the police, there are standing “anti-socialist inspection groups” consisting
of party and government officials and the prosecutorial office that are involved in
monitoring of illicit activities, including border crossing.
10
treatment and torture and deprived of adequate food and medical care. Many inmates do
not live to serve out their sentences and escape may even be more difficult than from the
sprawling political penal-labor camps (Hawk 2003, Muico 2007, KINU 2009, 97-101).
the 2008 survey,
In
9 percent of those incarcerated report spending time in a kyo-hwa-so
eniten
of
for
orcibly repatriated from China are also transferred to the jip-kyulo
colle
re not
and
ctivity,
p tiary.
The third and fourth components of the North Korean penal system that we
consider manage lower-level crimes and misdemeanors. The jip-kyul-so or “collection
centers” house low- or misdemeanor-level criminals for periods of up to six months
hard labor. As KINU (2009, 95) describes this level of the penal system, “the cases
handled by ‘collection centers’ include those whose crimes are not serious enough
[kyo-hwa-so penitentiaries] but too serious to send off to ‘labor training camps.’”
Examples would include violating a designated or restricted area or overstaying travel
permits, but the KINU report also lists absence from work or group training sessions.
Some North Koreans f
s ction centers.
Finally, in addition to the collection centers for incarceration for lower-level
offenses there has been an explosive growth of ro-dong-dan-ryeon-dae or labor-training
centers, a network of facilities that dates to the 1990s. The labor-training centers we
initially a statutory feature of the penal system. Rather, they emerged as an ad hoc
response on the part of authorities to the fraying of socialist control during the famine
in its immediate aftermath, including unauthorized movement, black market a
border-crossing and the other economic crimes listed above (Noland 2000).
11
The ro-dong-dan-ryeon-dae labor-training centers are operated at the municipal
level, and in effect constitute mobile labor brigades of relatively small numbers of
prisoners--30 to 60—typically held for less than six months in small, minimally-gua
and fenceless compounds. If not apprehended locally, these prisoners have already be
through interrogations and been shipped back to their hometowns for final release.
Sometimes the detainees in the labor-training centers are even allowed to go to their
homes for food or to recover from illness. Detainees resem
rded
en
ble corvee labor: they do road
n
g,
icitly introduced as a new form of punishment and the
lized
).
ains
a crime in the 2004 penal code, which provides ample discretion for penalties up to, and
repair, construction, and substitute for the lack of other forms of energy and transport i
the face of shortages, for example, by pushing train cars.
From 2001, this sort of labor training emerged more formally as the preferred
sentence for dealing not only with petty crimes, including hoodlumism and racketeerin
but for the growing range of economic crimes as well. In the 2004 revision of the penal
code, “labor training” was expl
existence of the ro-dong-dan-ryeon-dae labor training centers therefore institutiona
(Han 2006; KINU 2009, 90).
Labor-training centers have played a particularly important role in the
management of those caught crossing the border or repatriated from China (Muico 2007
Leaving the country without permission was initially considered equivalent to treason. Of
necessity, the government has been forced to soften this stance. The 1999 criminal law
revision first distinguished defectors leaving for a subversive purpose from migrants who
illegally leave the country largely for economic reasons; those in the first category were
subject to very much harsher punishments. Traveling abroad without permission rem
12
including, death if such activity is deemed to have a national security or anti-regime
dimension (Article 62). However, Article 223 of the revised penal code of 2004 permits
h
enalcollection
centers (jip-kyul)
or l
g
rcent of the entire 300 person sample,
inal
tivities that are subject to labor training under the
vised criminal code (Table 1).
sentences of up to two years in a ro-dong-dan-ryeon-dae labor training center.
Initial screening of repatriated North Koreans typically includes extensive
questioning at special National Security Agency detention facilities about contact wit
South Koreans while in China or exposure to South Korean propaganda, broadcasts,
movies or music; those deemed to be involved in these more serious political offenses are
liable to incarceration in kyo-hwa-so penitentiaries or even the kwan-li-so political p
labor camps. The NSA retains discretion to either release those involved in border
crossing after initial detention, which can last up to several months, or release them to the
People’s Safety Agency for incarceration in locally-managed
so abor training centers (ro-dong-dan-ryeon-dae).
Incarceration in these two types of lower-level facilities was by far the most
common form of contact with the penal system among our respondents. Of the 102 South
Korea-based survey respondents who reported some incarceration 49 reported spendin
time in a labor-training center and 68, or 23 pe
reported being detained in collection centers.
Table 2 summarizes the nature of the four main penal institutions including their
administration, the nature of the offenses, the prosecutorial process and sentencing, and
the number of respondents in our 2008 survey of South Korea-based refugees falling into
each institution. Particularly noteworthy is the porous line between political and crim
activities and the wide range of ac
re
13
Who Gets Arrested and Imprisoned?
al
h
-jeonrefugees
to have been detained for
illicit b
d
a
deterrent from trying to escape again on release. The risks of repatriation in China are
The first point of contact with the legal and penal system in North Korea is typically
either with the National Security Agency (NSA) or the People’s Security Agency (PSA),
although ad hoc “anti-socialist inspection units” have also recently been deployed to de
with border crossing and trafficking as well as economic crimes. The NSA deals wit
political offenses and conducts the first screening process of those apprehended for
border crossing or repatriated by Chinese authorities. Just under 30 percent of the 2008
survey respondents report being detained and questioned by the NSA (Bo-wi-bu).8
The
exact same share reports being detained and questioned by the criminal police (An
bu) or PSA, with the remaining 40 percent reporting that they were detained and
questioned by both. It is possible that the high share detained by the NSA or both the
NSA and the PSA reflects the greater propensity for
order crossing efforts at some point in time.
This suspicion gets some support from the results of the earlier Chinese survey.
Roughly one-quarter of the respondents in our Chinese survey had been repatriated, an
of those repatriated, 26 percent had been repatriated twice and another 15 percent had
been repatriated three or more times; for these individuals, even imprisonment was not
8
. The People’s Security Agency replaced the Social Safety Agency in 1998, and the
name of the police was changed from an-jeon-bu to in-min bo-an-sung. However, the
local police criminal continue to be known colloquially as an-jeon-bu, and we thus used
this term in the survey.
14
great (Kurlantzik and Mason 2006).9
However, some of those repatriated may have also
been engaged with cross-border trafficking, business or simply survival efforts, and thus
further increased their risk of capture and repatriation as a result. About one-fifth of the
China survey had returned to North Korea voluntarily, with the overwhelming reason
cited to take money or food back (79 percent and 11 percent of those returning,
respectively). Quite naturally, those who were repatriated were incarcerated at a
significantly higher rate than those who returned voluntarily.
This analysis is extended in Table 3, which reports a multivariate probit analysis
of the likelihood of being arrested among respondents in the second, South Korea-based
survey. The probability of being arrested is highly correlated with involvement in private
market activities, and to a lesser extent participation in an August 3rd
unit, a form of
entrepreneurial activity operated through existing state-owned enterprises and other
officially-sanctioned entities.10
Among this sample of refugees, the likelihood of being arrested is also positively
associated with having an advanced, post-college education, even when controlling for
occupation; being a professional was negatively correlated with probability of arrest, but
with a smaller estimated impact. One possibility is that those with higher levels of
education are better positioned than others to pursue illicit activities. Another possibility
is that the regime is more sensitive to the activities of the intelligentsia than other social
9
It should also be noted that there is credible evidence that refugees detained by Chinese
authorities are also subject to abuse and even torture prior to repatriation (Amnesty
International 2000, 2001, 2004; Lee 2006:53).
10
See Haggard and Noland (2009a) for more details on the August 3rd
movement.
15
groups. Intriguingly, being assigned to a military unit is associated with a higher
probability of being arrested in this sample.
Despite the ubiquity of “basic” illicit activities such as market trading, it is
striking that the prospects of arrest among this group of respondents was significantly
lower for both laborers and housewives, even though the involvement of housewives in
the market is widespread. This fact may reflect at least some forbearance where market
activities are seen as serving primarily survival purposes.
The North Korean regime has conducted a succession of classification exercises,
dividing the population into a class of reliable supporters, the basic masses, and the
“impure class”; these are commonly called the “core” (haek-sim-gun-jung), “wavering”
(gi-bon-gye-cheung) and “hostile” (gyo-yang-dae-sang) classes. Family class background
is a key determinant of life in North Korea (Hunter 1999).11
There is modest evidence
that being a member of the “wavering” class was positively correlated with likelihood of
arrest relative to both the “core” and “hostile” classes.
To what extent have patterns of arrest changed over time? Regressions 3.2-3.4
include dummy variables marking the period that refugees left North Korea. These
periods were defined by major turning points in North Korea’s post-famine history: the
2002 economic reform, which also coincided with the onset of the second nuclear crisis;
and the 2005 economic retrenchment, when previous reforms were partly reversed. Using
11
“Core” supporters of the government, including party members, enjoy educational and
employment preferences, are allowed to live in better-off areas, and have greater access
to food and other material goods. Those with a “hostile” or disloyal profile, such as
relatives of people who collaborated with the Japanese during the Japanese occupation,
landowners, or those who went south during the Korean War, are subjected to a number
of disadvantages, assigned to the worst schools, jobs and localities, and sometimes wind
up in labor camps.
16
the year that respondents left North Korea is at best an indirect means of assessing
whether arrest patterns have changed in response to these events, since incarceration is
only imperfectly related to time of departure. Nonetheless, the coefficients on these time
period dummies are statistically insignificant; there is no variation in the likelihood of
incarceration over time.12
This could be because the respondent’s date of exit is simply
too imprecise a measure to get at changes in penal practices over time. However, the
finding could reflect the fact that there is in fact no time trend in the government’s overall
propensity to incarcerate, even if certain forms of punishment such as labor training have
become more institutionalized over time.
In short, there is some suggestive evidence that the authorities disproportionately
incarcerate politically-suspect populations: those among the wavering classes, those
involved in economic activities beyond direct state control, and those with higher
education. However, the strategy of intimidation is not simply related to detention and
incarceration, but what happens to inmates once imprisoned.
The Nature of Punishment
North Korean statute has become both more legalistic and complex over time. Perhaps
due to a desire to conform—at least superficially--with international norms,13
revisions of
12
For the remainder of the paper, when referring to dates of exit, the pre-1999 period will
be referred to as the famine era; 1999-2002 as post-famine; 2003-2005, and 2006-present
as post-retrenchment.
13
. North Korean behavior in this regard is complex. On the one hand, the regime has
vehemently rejected the actions of the UN Council on Human Rights, a political body
subsidiary to the General Assembly that since 2003 has passed annual resolutions on
17
the legal code have included a number of standard legal protections. Habeas corpus was
introduced in the 1998 revision to the constitution. The 2004 criminal procedure law
stipulates that “no one shall be arrested or detained in a manner not provided for in the
law or without following the procedures prescribed in the law (Art. 177).” As obvious as
this might appear to anyone living in a liberal democracy, this article in fact overturned
the use of analogy in the application of criminal law, which granted prosecutors and
courts wide discretion to charge and sentence defendants for crimes that resembled, but
did not conform exactly, to existing statute (Han 2006, 3). The law also now stipulates
that no arrest shall be made without a warrant, that only investigators and “pretrial
agents” can make an arrest (Art. 180), and that a pre-trial agent making an arrest must
apply for, and receive, pre-approval from a prosecutor (Art. 181). A number of provisions
in the 2004 penal code revision even outline harsh penalties for those violating rules
governing arrest, detention, search and seizure.
A similar set of provisions appear to pertain with respect to the criminal trial
process. The National Security Agency is a gatekeeper and retains significant discretion
North Korea’s human rights record. The DPRK has also refused to meet with special
rapporteurs or the High Commissioner for Human Rights. On the other hand, North
Korea appears to have taken a somewhat different stance toward the UN Human Rights
Committee, a “treaty body” or technical committee of individuals—mostly constitutional
and international lawyers “treaty body”--that reviews the required implementation reports
under the International Covenant on Civil and Political Rights. The DPRK cooperates
with UN treaty bodies by submitting implementation reports and by sending
representatives to Geneva to appear before the review sessions. These review sessions are
followed by the issue of “Concluding Observations and Recommendations” on how the
States Parties can improve the implementation of their legal obligations. Some recent
changes described below appear to have been undertaken to bring North Korean law into
conformity with standards and recommendations associated with the treaty body process.
18
with respect to all political crimes. But revisions of the Criminal Procedure Law in 2004
and 2005 stipulate that “all criminal cases shall follow the principles, procedures and
methods stipulated in the Criminal Procedure Law” and that “trials be conducted at
appropriate levels of court, and the punishment levels shall be determined by court
decisions.”
These procedural changes do not seem to matter: of the 102 respondents in the
2008 survey who had been incarcerated, only 13 reported even receiving a trial at all.
Although the numbers are small, this share does not change significantly among those
who left after 2005 following the revision of code; of 25 leaving after that date, 3 (12
percent) report receiving a trial but 22 (88 percent) did not. Moreover, as the low share of
positive responses suggests, the absence of a trial and conviction was by no means
limited to those cases that ended up with detention in the political penal-labor camps
(kwan-li-so) and penitentiaries for more serious crimes (kyo-hwa-so). To the contrary, the
share of those reporting that they did not receive trials and convictions was even higher
in the lower level penal institutions: 86 percent of those incarcerated in the labor training
centers (ro-dong-dan-ryeon-dae) and 91 percent of those who served time in the
collection centers (jip-kyul-so). The North Korean legal and penal system clearly retains
an extraordinary level of discretion not only with respect to political crimes, but with
respect to lower-level infractions as well.
We do have some evidence that discretion is not only exercised with respect to
detention, but may be exercised with respect to release as well. Given the duration of
statutory sentences we expected that those incarcerated would have spent a long time in
prison. This did not prove to be the case. Table 4 provides information on how long
19
respondents were imprisoned, broken down by type of facility. Average time in prison is
certainly longer for the political penal-labor camps and the kyo-hwa-so penitentiaries.14
But the information on collection centers (jip-kyul-so) and labor training centers (rodong-dan-ryeon-dae)
is suggestive. Virtually all of those incarcerated in the labor
training centers are held for less than a year, even though sentencing guidelines suggest
the ability to hold prisoners up to two years for border crossing. Equally if not more
interesting are the local level facilities for criminal activities, including economic crimes.
Sixty-three percent of those incarcerated in these collection centers (jip-kyul-so) were
released within a month.
There is much about this system that we do not understand. It is possible that
inmates are escaping or bribing their way out of detention. However, this information is
consistent with a model of a police state in which authorities have a high level of
discretion in detaining, arresting and prosecuting people, but also a high level of
discretion in their ability to release them. One reason that such a model might be effective
is precisely because the conditions in the facilities are designed to have a powerful
deterrent and even psychological impact, in effect terrorizing those who are detained.
Nearly one-quarter of the sample in the initial, China-based, survey reported
having been arrested in China and repatriated to North Korea. Nearly 10 percent of the
respondents reported having been incarcerated in a political detention facility or
14
Although the number of those incarcerated in the political penal-labor camps is small in
our sample, this conclusion is certainly warranted from an abundance of other evidence;
again, Hawk 2003 is exemplary.
20
penitentiary.15
Ninety percent of this group reported witnessing forced starvation, 60
percent reported witnessing deaths due to beating or torture, 27 percent reported
witnessing executions. It has been alleged that pregnant women thought to be carrying
children of possible Chinese paternity have been subject to forced abortions or
infanticide; 5 percent of the respondents indicated that they had witnessed these practices.
This pattern of a high rate of affirmative response to general phenomena such as hunger
in the prison system and a much lower response on the highly specific practice of
infanticide suggests respondents were not simply providing the answers they believed
interviewers wanted to hear. This reassurance makes the response to a final question all
the more chilling: when asked if they believed that prisoners were used in medical
experimentation, a practice alleged by Demick (2004) and Cooper (2005) among others,
55 percent of the respondents believed (but did not necessarily witness) that this had
occurred at the facilities in which they were incarcerated.
The psychological impact of these experiences is profound. Incarceration is
highly correlated with psychological distress akin to post-traumatic stress disorder
syndrome (Chang, Haggard, and Noland 2008).
The China survey did not differentiate these experiences by the precise type of
penal institution (see note 15), but this was a focus of the 2008 South Korea-based survey.
Table 5 shows the share of respondents by level of penal institution that witnessed
15
Specifically, we asked whether they had been detained in either a penitentiary (kyohwa-so)
or other detention facility for political prisoners (jung-chi-bum su-yong-so); the
objective was to also capture the NSA’s detention facilities. This wording would leave
out the lower-level facilities referenced above, the co-called collection centers and labor
training centers, but the wording is admittedly vulnerable to the interpretation of
respondent.
21
executions, forced starvation, deaths from beatings or torture, or the killing of newborns.
The pattern of responses is quite similar to that obtained in the earlier survey: a high
response rate with respect to generalized forms of abuse, a much lower response rate on
the highly specific question on infanticide, again lending credence to the responses on the
other questions.
In the South Korean survey, the question concerning medical experimentation
was posed like the other questions as a direct interrogative about what they had witnessed,
not merely heard. Had the respondent seen with their own eyes medical experimentation
on prisoners who had received capital punishment? In contrast to the China-based survey,
none said that they had. This suggests two possible interpretations. The first is that such
experimentation does not occur. The second is that it occurs, but no one with direct
knowledge lives to bear witness to it. Respondents in the Chinese survey may have heard
about such experimentation, but not have seen it, or been adequately traumatized to
believe it was possible. By contrast, respondents in the South Korean survey answered
negatively simply because the bar was set higher by the question: that they had actually
witnessed such experimentation, something that they were unlikely to do.
What is striking about these findings is the ubiquity of violence and deprivation
across the various levels of the prison system and different initial points of contact with
authorities. The small number of respondents with experience in the political penal-labor
camps—and the short-time one respondent was incarcerated in one—prohibit any firm
conclusions about them from our survey, although the record with respect to these
institutions has now been thoroughly documented by a number of accounts from
survivors (Kang 2001 in particular).
22
But the findings with respect to lower levels of the prison system are arguably
even more striking. In both the lower level criminal facilities (the jip-kyul-so collection
centers) and the labor training centers, nearly half of respondents report seeing executions,
roughly three-quarters report forced starvation, and nearly a third report witnessing
deaths from beatings and torture—despite the generally shorter periods of incarceration
in these lower level facilities. The mean period of incarceration in both types of facility
was in the range of one month to one year. Prisoners experiencing this typical length of
incarceration in a jip-kyul-so collection center witnessed abuses at the following rates:
executions (75 percent), forced starvation (100 percent), and death by torture and
beatings (50 percent). For the labor training centers incarceration for the typical period of
time was associated with observing abuses at slightly lower rates: execution (60 percent),
forced starvation (90 percent), and death by torture or beating (20 percent). Nonetheless,
the conclusion is clear: even at these lower level facilities, inmates are exposed to
extreme levels of abuse.
Repression as a Barrier to Collective Action
Not surprisingly, the refugees in both surveys hold overwhelming negative attitudes
toward the incumbent regime. In the more recent survey of refugees in South Korea
nearly 87 percent of the respondents disagreed or strongly disagreed with the statement
that the Kim Jong-il regime was getting better.16
More than 90 percent of the
16
It is possible that the respondents’ perceptions were shaped by information that was
contemporaneous with the administration of the survey. However, it is a plausible assumption
that their impressions were shaped predominantly by conditions at the time they left; these were
23
respondents disagreed or strongly disagreed with the statement that the economy was
improving, citing rising materialism (92 percent), corruption (87 percent) and inequality
(84 percent) as problems (Haggard and Noland 2009a).
If anything, the state’s attempts to reassert control over the economy that become
visible from 2005 exacerbated adverse perceptions of the regime. A striking feature of the
survey is the very high share of respondents (71 percent) engaged in private trading.
Among those who left after the state began to reverse earlier reforms, including through
restrictions on market trading, 85 percent of the respondents reported a need to pay bribes
to engage in such activities. Official position was seen as valuable by respondents not
because merit or diligence is rewarded, but because it enables the pursuit of business and
corrupt or criminal rent extraction.17
The respondents increasingly hold the North Korean government accountable for
their plight, with the share placing primary responsibility on the North Korean
government at more than 95 percent among those who left in the post-retrenchment
their last first-hand experiences with the country, and their responses do not appear to correlate
either with time spent outside of North Korea or time spent in South Korea.
17
When asked the best way to get ahead in North Korea, officialdom (including both
government and party) trumped either the military or engaging in business, flagged by 80
percent of the respondents in the post-retrenchment subsample. The share citing
“engaging in business” more than doubled from 8 percent among respondents departing
in the famine era to 16 percent for those leaving in the post-2005, with this shift coming
almost completely at the expense of joining the military. (While the military as an
institution may be of rising influence in North Korea, low level conscripts appear to be
treated badly.) When asked “what is the easiest way to make money in North Korea:
work hard at assigned job; engage in market activities, engage in corrupt or criminal
activities, none of the above,” the most frequent response was that engaging in market
activities was the easiest way to make money, but a steadily increasing share—more than
one-quarter in the post-2005 cohort—saw corruption and criminality as the most lucrative
career path. There is no sense that fidelity is rewarded; only a small—and falling—share
reported that working hard at your assigned job yielded fruit.
24
period (i.e. after 2005). The share citing the policies of foreign governments as
responsible for their predicament—a core claim of the regime--falls steadily from 18
percent among the famine era leavers, to 4 percent in the post-retrenchment group
(Haggard and Noland 2009b).
This growing tendency to hold their government accountable naturally raises the
question of how these opinions have been formed and the extent that they are
communicated to others. A striking feature of the marketization process –and one that is
no doubt seen as dangerous to the regime—is a declining ability to control the flow of
information. A rising share of respondents and a majority of the final, post-retrenchment
era subsample report watching or listening to foreign media. Even more striking is that
efforts to curtail the flow of information do not seem to be working. A falling share (nil
in the post-retrenchment period) report that they have access to foreign media but decline
to watch or listen. Not only is foreign media becoming more widely available, inhibitions
on its consumption are declining as well (Lankov 2007).
That the refugee population is disaffected and holds the government accountable
may not be surprising. However, the survey casts important light on the effectiveness of
repression and the possibilities for collective action as well. The share of respondents
agreeing or strongly agreeing that the government is increasing restrictions on the
citizenry remained relatively constant at 55-65 percent across all four subsample periods.
The shares affirming the statement that people make jokes about the government, while
rising, never exceeds 45 percent in any of the subsamples, and the share agreeing with the
statement that people complain about the government never reaches 40 percent. Even
among an unusually disaffected sub-group of the population, refugees, and despite their
25
overwhelmingly negative assessment of the regime, less than half of the sample report
that their peers joked or complained about the government.18
To what extent might the market itself become the locus of overt political conflict
with the government? In response to the regime’s efforts to control markets, there have
been occasional reports of incidents in which market traders, mostly women, have
publicly protested such restraints (Martin and Takayama 2008). Moreover, as previously
discussed, involvement with the market is correlated with the likelihood of detention.
Could participation in market activities serve to overcome barriers to collective action?
Survey responses depict relatively low levels of collective action. When asked
whether traders cooperated with each other, the share of respondents agreeing or strongly
agreeing ranged from 32 percent to 42 percent across the four time periods with no
perceptible trend. Likewise, when asked whether traders in the market were beginning to
organize to protect their interests, the affirmative response rate was 28-29 percent in all
time periods—implicit evidence of the continuing atomization of North Korean society.
Were anti-regime organizing to be effective, the political preferences of the
respondents are very clear. The respondents were asked three questions concerning their
preferences regarding the political organization of the Korean peninsula. They were first
asked which alternative more accurately represented their views while in North Korea:
maintenance of the current North Korean government; installation of a new non-Kim Ilsungist
government in North Korea; unification with South Korea (presumably under
18
Kim Jong-il appears sacrosanct: although free discussion of Kim rises steadily among those
who left the country after 1998, even among those who fled during the post-retrenchment period,
only 8 percent of the respondents report that people spoke freely about Kim Jong-il, figures
almost precisely mirroring the shares reporting that people were organizing against the
government.
26
South Korean leadership given their negative perceptions of the regime in the North); or
don’t know/none of the above. In addition to their own views at the time of departure,
respondents were also asked what they believed now and what they believed the
preferences of other North Koreans were.
Unification is supported overwhelmingly (figure 1): not only is there little support
for the maintenance of the status quo (only a single respondent out of 300), there is little
support for “third way” solutions in which North Korea would remain independent under
an alternative political regime (Haggard and Noland 2009b). Exposure to South Korea
intensifies these preferences at the margin. But the respondents also indicate that their
own views mirror those of their peers remaining in North Korea, even though there is
obviously no way of judging the accuracy of this projection. There is a slight tendency
for these views to be held even more strongly among those who have recently exited
North Korea.
Conclusion: A Model of Repression in North Korea
The refugee literature provides a much more eloquent testimony to the abusive nature of
the North Korean system than anything we can add here. However, this brief review of
the development of the criminal and penal system and evidence from two surveys does
shed some additional light on the nature of repression in North Korea. First, the
development of the legal system exhibits at first glance what appear to be contradictory
trends. There is a marginal increase in legalization, in the sense of incorporating basic
legal protections into statute, and a relaxation in the treatment of some crimes related to
economic survival. The best documented example of this forbearance is the legal
27
treatment of border crossing, which has been demoted from the equivalent of treason to a
misdemeanor offense, at least for those showing no political motive and avoiding contact
with “depraved” foreign culture.
Yet the dominant tendency is the dramatic expansion in the range of economic
activities deemed criminal, evident in the finding that those engaged in market activities
were more likely to be incarcerated. Also apparent is the institutionalization of “labor
training” as a means of dealing with these and other lower-level crimes. These two trends
can be reconciled by noting that the government maintains a very high level of discretion;
whatever the law says, the security apparatus is capable of making adjustments in
detention and incarceration with few checks on its authority. One might expect an uptick
in detention and incarceration whenever the government is intent on checking market
activity, as it has been since 2005 in particular (Haggard and Noland 2009a).
In fact the statistical analysis of detention experiences suggests that the regime
disproportionately targets politically suspect groups, particularly those involved in
economic activities beyond direct state control. The penal system subjects them to terror
in an attempt to keep them atomized and quiescent. A major finding of our paper is that
conditions in lower level penal facilities approximates in several measurable ways
conditions in facilities designed to house felons and even the most dangerous political
prisoners. Of course, incarceration in political penal-labor camps and penitentiaries
carries much longer sentences, and many prisoners of these penal institutions end up
dying in them. Yet it is nonetheless surprising that among our respondents, there was
very little difference in the propensity to witness extreme forms of violence and
deprivation in the notorious kwan-li-so penal-labor camps and penitentiaries than there
28
was in the lowest-level detention or labor training facility—particularly when taking into
account the generally shorter periods of incarceration at the lower level facilities.
In combination, these findings provide insight into how to think about North
Korean politics, and the centrality of discretion and terror to the maintenance of the
regime’s power. Obviously, an authoritarian regime has an incentive to mete out
particularly harsh punishment for those posing political challenges to the regime.
However, in a fraying socialist system, individuals are of necessity thrown into a variety
of market-like activities for their sustenance, activities over which the government almost
by definition exercises less control than activities in the state sector. However, the regime
has maintained, and perhaps even expanded its discretion to arrest, detain and terrorize
those operating in this sphere and appears to treat them as harshly as they do either
common criminals or the most dangerous counter-revolutionaries.
Such a system obviously has the effect of sowing fear, and labor training has the
additional benefit of constituting a form of corvee labor or tax. But this pattern of
detention may also have a somewhat different economic motivation. Our surveys
provide evidence of an increase in corruption in North Korea. High levels of discretion
with respect to arrest and sentencing and very high costs of detention, arrest and
incarceration actually have the effect of increasing bribe costs. The more arbitrary and
painful the experience with the penal system, the easier it is for officials to extort money
for avoiding it. As a result, these features of the penal system not only serve the interests
of regime maintenance through intimidation, but may provide incentives and
opportunities for the corruption of the internal security apparatus as well.
29
30
References
Amnesty International. 2000. Persecuting the Starving: The Plight of North Koreans
Fleeing to China. London.
Amnesty International. 2001. Human Rights in China in 2001 – A New Step Backwards.
London.
Amnesty International. 2004. Starved of Rights: Human Rights and the Food Crisis of the
Democratic People’s Republic of Korea. London. Available at web.amnesty.org
Chang, Yoonok, Stephan Haggard, Marcus Noland. 2008. “Migration Experiences of
North Korean Refugees: Survey Evidence from China,” Working Paper 08-04.
Washington: Peterson Institute.
Chang, Yoonok, Stephan Haggard, and Marcus Noland. 2009. “Exit Polls: Refugee
Assessment of North Korea’s Transition,” Journal of Comparative Economics. 37:2 144-
50.
Cooper, Abraham (2005), “Toxic Indifference to North Korea,” Washington Post, 26
March.
Database Center for North Korean Human Rights. 2008. White Paper on North Korean
Human Rights Statistics 2007. Seoul: Database Center for North Korean Human Rights.
Demick, Barbara (2004), “North Korea’s Use of Chemical Torture Alleged,” Los Angeles
Times, 3 March.
Haggard, Stephan and Marcus Noland. 2007. Famine in North Korea: Markets, Aid, and
Reform. New York: Columbia University Press.
Haggard, Stephan and Marcus Noland. 2009a. “Reform from Below: Institutional and
Behavioral Change in North Korea,” Journal of Economic Behavior and Organization,
forthcoming.
Haggard, Stephan and Marcus Noland. 2009b.”North Korean Political Attitudes:
Evidence from a Refugee Survey,”
Han, In Sup. 2006. “The 2004 Revision of Criminal Law in North Korea: A Take-Off?”
Santa Clara Journal of International Law 5, 1: 122-133.
Harden, Blaine. 2009. “N. Korea’s Hard-Labor Camps: On the Back Burner,”
Washington Post, July 20.
Hawk, David. 2003. The Hidden Gulag. Washington: US Committee for Human Rights
in North Korea.
31
Hunter, Helen-Louise. 1999. Kim Il-song’s North Korea. Santa Barbara: Greenwood.
Kang, Chol-hwan. 2002. The Aquariums of Pyongyang. New York: Basic Books.
Korea Institute for National Unification (KINU). 2009. White Paper on Human Rights in
North Korea 2009. Seoul: KINU.
Kurlantzick, Joshua, and Jana Mason (2006), ‘North Korean Refugees: The Chinese
Dimension’, in: The North Korean Refugee Crisis: Human Rights and International
Response, Washington: US Committee for Human Rights in North Korea, pp. 34-52
Lankov, Andrei. 2007. North of the DMZ : Essays on Daily Life in North Korea.
Jefferson, N.C.: McFarland & Company.
Lee, Keumsoon. 2006. The Border-Crossing North Koreans: Current Situations and
Future Prospects. Studies Series 06-05. Seoul: Korean Institute for National Unification.
Martin, Bradley and Hideko Takayama. 2008. “North Korean Women Fight Back as Kim
Orders Them Out of Markets,” Bloomberg
at http://www.bloomberg.com/apps/news?pid=20601109&refer=home&sid=aIL0fcgH66
G4# (May 28).
Muico, Norma Kang. 2007. Forced Labor in North Korean Prison Camps. London: AntiSlavery
International.
Noland, Marcus. 2000. Avoiding the Apocalypse: The Future of the Two Koreas.
Washington, DC: Institute for International Economics.
United Nations Population Fund, 2009. Preliminary Results of the 2008 census of
population of the Democratic People’s Republic of Korea conducted on 1-15 October,
2008. On file with authors.
32
Table 1. Types of Crimes and Corresponding Place of Detention
Category
Correctional Centers Designated location
Unlimited Term Limited Term Labor Training
Anti-state, anti-people
crimes
Conspiracy to overturn
the state
Conspiracy to overturn
the state
--
(14 types) (5 types) (14 types)
Crimes disruptive to
national defense systems
-- Neglecting preparedness
for wartime production
Neglecting preparedness
for wartime production
(16 types) (15 types) (10 types)
Crimes injurious to
socialist economy
Taking or robbing state
properties
Stealing or robbing state
properties
Stealing or robbing state
properties
(104 types) (6 types) (83 types) (76 types)
Crimes injurious to
socialist culture
Smuggling historical
relics and smuggling
and selling of narcotics
Importing and spreading
depraved culture
Importing and spreading
depraved culture
(26 types) (3 types) (25 types) (16 types)
Crimes injurious to
administrative systems
-- Collective disturbance;
Interfering with official
business
Interfering with official
business; Creation or
dissemination of false
information
(39 types) (30 types) (29 types)
Crimes harmful to socialist
collective life
-- Acts of hoodlumism or
racketeering
Acts of hoodlumism or
racketeering
(20 types) (15 types) (18 types)
Crimes injuring life or
damaging property of
citizens
Willful murder or
kidnapping Willful murder Excessive self-defense
(26 types) (3 types) (25 types) (13 types)
Source: Korean Institute for National Unification's White Papers on Human Rights in North Korea 1009, Table 2-4
33
Table 2. The North Korean Prison System: An Overview
Facility Supervising
institution
Offenses Prosecutorial
process and
sentencing
Number in 2008
sample
(n=300; number
incarcerated =
103). Numbers do
not sum to 100%
because of
multiple
incarcerations
Kwan-li-so
political penallabor
camps
National Security
Agency (Bureau 7)
Serious political
and ideological
crimes, but also
imprisonment of
suspect categories
High level of NSA
discretion; life
sentences,
including for
extended family;
confiscation of
property
3 (3.9% of those
incarcerated)
Kyo-hwa-so
penitentiaries
People’s Security
Agency
In addition to
criminal felonies,
serious crimes
disruptive of
national defense,
injurious to the
socialist economy,
injurious to
socialist culture,
injurious to
administration and
harmful to socialist
collective life
Trial and court
sentencing;
“limited” terms of
1-15 years and
“unlimited” terms
of more than 15
years of
correctional labor.
9 (11.3%)
Jip-kyul-so
collection centers
People’s Security
Agency
More serious
misdemeanors and
economic crimes,
including theft of
state property,
spreading
“depraved
culture,” some
border crossing
Trial and court
sentencing;
sentences of six
months to one
year.
68 (75.6%)
Ro-dong-danryeon-dae
labor
training centers
People’s Security
Agency, operated
at county or
municipal level
In addition to
lower-level crimes,
an expansive
number of
economic crimes,
violations of labor
administration and
rules governing
socialist culture
Initially ad hoc
rehabilitation
facilities.
Institutionalized
with 2004 revision
of the penal code
and expanded use
of “labor training”
as punishment.
Sentences of six
months to two
years.
49 (55.7%)
34
Table 3. Detainment by Bo‐wi‐bu or An‐jeon‐bu polices in North Korea (probit arrested=1) 
           
  
(1)  (2)  (3)  (4) 
Arrest: Detained by either Bo‐wi‐bu or An‐jeon‐bu police 
Class: Wavering  0.305*  0.307*  0.304*  0.304* 
   (0.166)  (0.167)  (0.166)  (0.166) 
Private Activity  0.561***  0.581***  0.570***  0.562*** 
   (0.190)  (0.194)  (0.190)  (0.191) 
Occupation: Professional  ‐0.867**  ‐0.880**  ‐0.885**  ‐0.861** 
   (0.386)  (0.374)  (0.384)  (0.386) 
Occupation: Housewife  ‐1.109***  ‐1.170***  ‐1.131***  ‐1.106*** 
   (0.251)  (0.247)  (0.250)  (0.251) 
Occupation: Laborer  ‐0.517***  ‐0.532***  ‐0.527***  ‐0.517*** 
   (0.181)  (0.182)  (0.182)  (0.181) 
Workunit: August 3rd unit   0.486*  0.465*  0.500*  0.484* 
   (0.259)  (0.260)  (0.260)  (0.259) 
Workunit: Army  0.692*  0.625*  0.686*  0.691* 
   (0.353)  (0.353)  (0.351)  (0.355) 
Education: Post college  1.404**  1.386**  1.378**  1.411** 
   (0.639)  (0.634)  (0.642)  (0.640) 
Left North Korea post‐reform period     ‐0.218       
          (2003~)     (0.159)       
Left North Korea post‐judicial reform        ‐0.142    
          (2005~)        (0.182)    
Left North Korea post‐retrenchment period           0.045 
          (2006~)           (0.217) 
Constant  ‐0.762***  ‐0.650***  ‐0.727***  ‐0.769*** 
   (0.204)  (0.215)  (0.211)  (0.209) 
Observations  300  300  300  300 
Pseudo R‐squared  0.115  0.120  0.116  0.115 
Log likelihood  ‐171.4  ‐170.4  ‐171.1  ‐171.3 
Chi‐squared  43.80  47.71  46.19  43.81 
*** p<0.01, ** p<0.05, * p<0.1             
Robust standard errors in parentheses             
35
Table 4. Length of Imprisonment by Detention Facility
Kwan-li-so Kyo-hwa-so Jip-kyul-so
Ro-dong-dan-
ryeon-dae
Less than 1 week Freq 1 2 11 10
Pct 33.33 22.22 16.42 20.41
Less than 1 month Freq 0 1 31 17
Pct 0 11.11 46.27 34.69
Less than 1 year Freq 1 2 22 20
Pct 33.33 22.22 32.84 40.82
Between 1 and 5 years Freq 1 3 3 2
Pct 33.33 33.33 4.48 4.08
More than 5 years Freq 0 1 0 0
Pct 0 11.11 0 0
Total Freq 3 9 67 49
Pct 100 100 100 100
Table 5. Experiences of Violence in the North Korean Prison System
(Share of those imprisoned in each type of facility)
Kwan-li-so
N=3
Kyo-hwa-so
N=9
Jip-kyul-so
N=68
Ro-dong-dan-
ryeon-dae
N=49
While you were detained or imprisoned did you see with your own eyes:
Executions 66.7 77.8 50.8 47.9
Forced
starvation
33.3 66.7 73.1 83.7
Death from
being tortured or
beaten
33.3 55.6 33.3 30.61
Killing of
newborns
0 11.1 7.7 8.3
36
37
Appendix: Sample Characteristics
This paper draws on two refugees surveys. Neither of these surveys was random;
neither we nor anyone else knows the underlying characteristics of the refugee population,
and cluster-type techniques used in other contexts to correct for these problems were
infeasible. Nonetheless, a comparison of the composition of the survey with underlying
demographic characteristics of the country and what we know about patterns of egress
suggests that the two surveys are probably a reasonable reflection of the North Korean
refugee population. The Chang, Haggard and Noland (2008) survey of 1,346 refugees
was conducted from August 2004 to September 2005 at 11 sites in China by 48
individuals trained by one of the authors before conducting the interviews.19
Most of the
respondents were prime age adults, with a median age of 38 years and females slightly
outnumbering males (52 to 48 percent). As in other surveys, members of lower-income
classes and residents of the northeast provinces were both overrepresented (cf. Robinson
et al. 1999, 2001a, 2001b; Lee et al. 2001; Chon et al. 2007 Lee 2007; Kim and Song
2008; and Lee et al. 2008). Most respondents were laborers (54 percent), with farmers
(34 percent) the next largest occupational group. Most respondents were from North
Hamgyong province (57 percent), followed by South Hamgyong province (19 percent);
these two provinces both felt the brunt of the famine and are geographically proximate to
the border. Although this distribution of responses actually makes these provinces
somewhat less overrepresented than in earlier surveys, these provinces account for only
about 23 percent of the North Korean population (United Nations Population Fund, 2009).
19
Shenyang, Changchun, Harbin, Yangbin, Tumen, Helong, Hunchun, Dandong, Jilin,
Tonghua, and Wangqing.
38
The survey of 300 North Korean refugees living in South Korea was conducted in
November 2008. Again, the overwhelming majority of the second survey was prime age
adults, with just over half between the ages of 35 and 50, and a larger majority than in the
Chinese survey accounted for by women (63.3 percent). Residents of the northeast
provinces were again overrepresented, with North Hamgyong province accounting for 50
percent of respondents followed by South Hamgyong province with 14.7 percent. It is
important to underscore, however, that while this overweighting of the northeast limits
the conclusions that can be drawn from the sample with respect to the North Korean
population as a whole, it does not necessarily present a problem for drawing inferences
about the North Korean refugee communities in China and South Korea, which almost
certainly are similarly skewed.
The occupational status of the respondents in the second survey is complicated
somewhat by the large number of women in the sample; 52, or 17.3 percent of
respondents report that they are housewives. If we look only at those in the economically
active population—excluding housewives, students and retirees (73 respondents, or just
under one quarter of the sample)—the largest category among those in the workforce is
laborers (40.1 percent), followed by government (18.9 percent), and merchants (7.9
percent, with nearly two thirds of those women); the occupational distribution of the
South Korea survey thus differs somewhat from the China survey and is more diverse.
However, a closer inspection reveals that a substantial share of those listing their
profession as laborers in fact work on collective farms or cooperatives, resulting in an
occupational mix that is closer to the Chinese survey than it first appears.
39
With respect to political classification, the bulk of respondents were categorized
as “wavering” (61.7 percent) with 11 percent “hostile” and 13.7 percent reporting that
they did not know. Nonetheless, 13.7 percent reported being in the “core” group,
suggesting that even privileged political status did not provide benefits adequate to deter
migration.