Category Archives: CIES 2011

Complementary, not Supplementary Education?

This post is another result of conversations with Victor Kobayashi at the CIES meetings.

I continue to wrestle with terminology (no good at all other than for very quick communication: ‘cram school‘; better, but perhaps no longer true: ‘shadow education‘; even better, but not evocative for general public or many colleagues and what do I call the actual schools: ‘supplementary education’).

Kobayashi raised the possibility of referring to juku as “complementary education”. He based this partly on the history of juku and its premodern focus on erudition rather than education. The argument then is to say that juku complement conventional schools rather than supplementing them. In my mind, this captures the “shadowing” part of ‘shadow education’ better as it hints at the extent to which juku follow the official school curriculum with very few exceptions.

My hesitancy about ‘shadow education’ stems in part from the fact that much of these activities are no longer in the shadow and that also applies to complementarity. When students across different contexts are reporting that the ‘real learning’ (whatever that is, exam success seems to be hinted at) occurs at juku not in conventional schools, then neither ‘complementary’, ‘shadow’, nor ‘supplementary’ education works any longer.

Just as research on supplementary (or shadow) education may be establishing itself with this label, the brand may be becoming obsolete through the intervention of pesky empirical reality.

“Enduring Contexts”: Bjork/Fukuzawa

Chris Bjork and Becky Fukuzawa presented a fascinating attempt to compare school-based research on “guidance” in the 1980s and the 200s. Diachronic fieldwork without resorting to the use of a time machine, wonderful!

The gist of the much more elaborate and interesting discussion in the paper was that discipline, sometimes backed up by corporal punishment, prevailed in 1980s schools and that “guidance” meant the enforcement of school rules covering hair length, clothing, etc. By the 2000s schools had largely given up on the enforcement of this kind of guidance. While the nature of a diachronic comparison of ethnographic fieldwork precludes conclusions about the extent to which this conclusion would apply across Japan and across school types, it’s an observation that rings true to many scholars who have been involved in research on education in Japan for a longer period.

Bjork and Fukuzawa in their discussion of the 1980s refer back to Tom Rohlen’s classic (certainly for our field) Japan’s High Schools (Berkeley: UC Press). In his discussion, Rohlen summarizes the reliance on discipline as an element in guidance under the motto of “intimacy coupled with severity” (p. 201)

I frequently see remnants of this kind of attitude in juku as well. While I see much genuine warmth in the relationship between juku students and teachers/operators, some of these teachers also do not shy away from moralistic admonishment of students in a way that takes me aback at times, just because some of these lectures seem to belittle or browbeat students more than anything else. Rohlen’s original understanding of this “intimacy coupled with severity” approach channeled by Bjork and Fukuzawa helps me make a bit more sense of this dynamic.

By contrast, however, the operators and teachers of small juku at least, put very little stock in outward appearances like uniforms that were and to some extent are the focus of guidance within schools.

This kind of guidance provided especially by some of the “charismatic educator” type of juku-cho clearly goes much beyond subjects or matters directly related to lessons, homework or schooling. Moral guidance thus becomes an element in some of the in loco parentis consulting that juku increasingly seem to offer parents and students.

Vic Kobayashi contributed a wonderful footnote to Bjork and Fukuzawa’s discussion of “guidance” when he pointed out that in the immediate postwar years, guidance was generally used in Japanese schools as an imported word, i.e. ガイダンス. Privately, this notion was often derided as 外ダンス (a foreign dance). While this is admittedly a bit of a nerdy pun, I hadn’t left as hard about anything in a discussion of Japanese education in quite some time. The notion that supported this pun, was of course that Japanese teachers had traditionally (i.e. in the prewar/wartime education system) engaged in plenty of moral guidance and that this was now being performed in the guise of a foreign dance or perhaps a dance for foreigners.

“Enduring Contexts”: Macdonald

Larry Macdonald (Soka University) reported on one of the aspects of yutori education that has been largely overlooked in the hoopla and panic associated with the presumed impact of a reduction in textbook content on academic abilities: the (gentle) introduction of integrated teaching and a more interrelated curriculum, 総合的な学習の時間 (sôgô tekina gakushû no jikan, general study period). Under this banner, three hours per week were to be dedicated to study that reaches across the different subjects in schools. This has been reduced to two hours this year in the context of the yutori reversal for the current curriculum.

Macdonald reported on some of the more creative, ambitious and far-reaching focus areas that particular schools have selected for the general study period. A number of Osaka-area schools with a strong tradition of dowa education thus focused on an integrated discussion of human rights.

The most exciting approach to the general study period that I had never heard of previously is that the Japan International Cooperation Agency (JICA) has apparently developed curriculum materials to introduce students to its work and thus to developing countries. Fascinating!

Even though juku instruction would lend itself very well to a turn towards integrated study, I have not seen much of this in any of my fieldwork. Especially subject matter from arithmetic/math and science, but also from Japanese and social studies could well be discussed in an integrated fashion, but neither instructors nor any of the teaching materials commonly used in juku have pursued this to my knowledge.

Japan Panels at Different Conferences

For those of us actively conducting research on contemporary Japan, the extent to which Japan captures the public imagination as well as the imagination of our non-Japan-specialist colleagues obviously has a significant impact on our professional lives.

Beyond the widespread political lament about “Japan-passing”, there are more direct impacts on our professional activities in terms of attendance at Japan-related events we organize and probably also in terms of our and our students’ opportunities in granting competitions, etc.

In the context of the question of any decline in interest in Japan, it was very interesting to attend the AAS/ICAS conference and the CIES conference within a short time span and to compare Japan panels at these conferences.

The bottom line is that AAS conference participants continue to divide along country-specialization lines, while a country focus (at least on Japan) seems to be declining at CIES.

In terms of my own scholarly interests, I do not have a preference for either form of organization, and find both very useful.


On the one hand, I very much enjoyed the “Enduring Contexts” exchanges in Montreal that were exclusively focused on Japan, on the other hand, I tend to learn more in terms of arguments/theories/explanations from cross-national contexts such as the panel that Kathryn Ibata-Arens (PoliSci, DePaul Univ) organized for the AAS “Innovations in Education in Asia: Private Sector Growth, Government Reform and Emerging Models of Best Practice?” where I presented on “The Impact of Private-Sector Innovations on Public Primary and Secondary Education in Japan”. The AAS has been promoting these cross-area panels for some years now and most participants would agree that intellectually and in terms of moving research along, this is the more desirable form of organization. However, in my experience, attendance at these panels remains anemic while panels with a specific country/region focus continue to see strong attendance.

I realize that the latter phenomenon (strong country panel attendance) may be a function of the size of the Japan-crowd within AAS, and that the former phenomenon (weak attendance at interarea panels) may be due to the small number of Asian regionalists or intra-Asia comparativists among AAS attendants. Nevertheless, that seems to continue to be the reality at AAS meetings.


The situation at CIES seems to be the opposite. Attendance at Japan panels, including the one I presented on, is relatively weak, perhaps signaling a decline in the number of Japan specialists, while cross-national comparative panels, like the one I served on as a discussant for papers on shadow education in Cambodia, Hong Kong, and Malaysia, saw strong participation, though this may have been due to Mark Bray’s prominence.

Again, for my needs, I depend on both of these contexts. While it is very important for my understanding of juku to have information on and a good understanding of yutori education, for example, as it was a topic of discussion among the “Enduring Contexts” participants, it is also essential for me to have an opportunity for interaction with researchers who a) are investigating supplementary education in other countries/cases, and b) are looking at supplementary education on a more global scale through a comparative lens.

I thought that the discussions at CIES were very productive, so to the extent that I’m able to do so, would hope to attend more of these conferences in coming years, in part to have opportunities for discussions with supplementary education comparativists, but I will also continue to attend AAS, though perhaps more for interactions with the Japan-specialist crowd (in addition to the Japan Studies Association of Canada meetings, of course).

Premodern Historical Roots of Juku

At a number of the discussions at CIES, the question of the history of juku came up (both the term, as well as the teaching/school format).

Victor Kobayashi (emeritus, Univ of Hawai’i, and our very gracious host for the “Enduring Contexts” discussions paralleling the 2011 AAS meetings) was particularly interested and insightful on this.

He pointed out – quite correctly – that juku had a long pre-Meiji history. In fact, this is one of the important observations about supplementary education in general, i.e. what we see as a process of ‘privatization’ now is really just a ‘re-privatization’ after a century or two of the extraordinary growth of public education systems.

In the context of pre-modern juku, Kobayashi pointed in particular to the use of the term 教養 (kyôyô – erudition, refinement, Bildung) as opposed to the Meiji neologism of 教育 (kyôiku – education).

Numerous private educational institutions then came to be known as juku during the Meiji era, most prominently perhaps 慶應義塾 (today’s Keiô University: English History | Japanese History) in somewhat of a departure from premodern practice.

The final chapter on historical roots is the use of the term juku in postwar Japan, but that deserves a post of its own.

“Enduring Contexts”: Nomi

As part of the “Enduring Contexts” discussions, Tomoaki Nomi, Southeast Missouri State Univ, presented some of his work on the link between socio-economic background and students’ academic achievement. He is drawing on data from Tokyo prefecture and focuses on elementary and middle school.

In his paper, Nomi referred to the general decline of academic standards in Japan as a “widely recognized trend”. I strongly object to this view, as I also expressed in my comments on Hyunjoon Park‘s contribution. In short, what is important about Japan’s PISA scores is that they have been interpreted (along with other data) to suggest a general trend of decline and this perception is pervasive, not whether this trend is actually empirically substantiated (which I doubt). Keita Takayama has written decisively about this issue and has recently been awarded the CIES’ George Bereday Award for the most outstanding article published in the Comparative Education Review. In the end, this was more of a semantic disagreement with Nomi.

For further reading, I would  recommend Nomi’s Japan Focus article: “Inequality and Japanese Education: Urgent Choices“. For a different view articulated most prominently by Takehiko Kariya, see “Misinterpreting Globalization in the Context of Japanese Education Policy” in Asia Pacific Memo.

CIES Presentation Langager et al.: Balancing CJK and English Literacy Objectives

With thanks to Mark Langager for sharing their abstract.

Wednesday May 4: Session 424, 13:45-15:15h, Floor C – Saint Laurent

Mark Langager, Hui Xu, and Jadong Kim, International Christian University

“Balancing CJK and English Literacy Objectives: A Case Study of East Asian Supplementary Schools in the Seattle Area”


Biliteracy has been broadly seen as being acquired by learners on a number of continua (Hornberger, 1989) between sets of two poles, including L1 versus L2 dominance, textual versus oral skills, communicative versus academic language and so forth. Until the recent work of Kondo-Brown and others (2006), however, most of the biliteracy literature has focused on that which occurs between languages with Latinate orthographies. Moreover, outcomes typically view language minorities as immigrants and thus favor analyses, findings, and conclusions based on skills in L2, generally the language of the host country and often the researcher (Langager, 2010).

Staking a more neutral spot on the continuum between ultimate L1 versus L2 literacy objectives, the current study uses interview and observation data to examine educational objectives of Chinese, Japanese and Korean communities providing L1 Saturday instruction at supplementary schools in the Seattle area. Students’ US residence ranges from short-term to long-term and thus literacy goals favor a range of priorities between L1 and L2. Initial findings suggest the three language communities all differ systematically in their literacy acquisition goals, albeit all three hold fast to both L1 and L2 objectives. Differences are identified among the three communities in the quality of governmental overseas assistance, circumstances bringing families to the US, educational career paths, and the place of entrance exams in the homeland. Accordingly, while children’s education varies considerably among individuals, separate patterns can be traced for the three communities.

“Enduring Contexts”: Akiba

Akiba’s contribution to the “Enduring Contexts” project was based on her article on bullying in Japanese middle schools: Motoko Akiba, Kazuhiko Shimizu, and Yue-Lin Zhuang. 2010. “Bullies, Victims, and Teachers in Japanese Middle Schools“, 54(3): 369-392.

One of the very interesting aspects of attempt to (ac)count (for) widespread bullying (いじめ ijime) in Japan is that the MEXT statistics suggest that the incidence of bullying is actually declining. Akiba et al. thus cite official statistics as showing that reported cases dropped from approx. 60,000 in 1995 to approx. 20,000 in 2005. Magical, that! This has got to be one of the all-time rare instances of the recognition, naming and consciousness of a social problem to lead to a reduction in incidence. It seems to me that most social ills that I can think of increase in reported incidence as they are recognized by society, the media, and the state as a category of social ill. Hate crimes would be an interesting example of such a development that a former PhD student of mine, Bern Haggerty, has written about.

There probably is a literature or at least some writings on why we would see the drop in reported cases of bullying in Japan, but the two explanations that come to mind would be that MEXT also created a category of violence in school that may be ‘attracting’ some of the reports of bullying, and that there are no incentives for principals to report on bullying in an accurate and truthful manner, especially with the introduction of school choice down to the middle school level in many Japanese jurisdictions.

Compared to these official statistics, Akiba et al.’s figures of around 30% of middle school students in one school district reporting victimization are obviously much higher. A quick back-of-the-envelope calculation: the per grade population of Japanese students must be on the scale of 1 million so that a 30% victimization rate in middle school would suggest around 3 mio victims at that level of schooling alone. That calculation makes me wonder a bit about self-reported victimization/perpetration as it seems so enormous.

One of the clearest findings in Akiba et al. is the higher rate of victimization and perpetration of physical violence for boys than for girls.

I am less convinced by some of the causality implied by Akiba et al.’s article which Motoko Akiba also included in her presentation to the Enduring Contexts meeting.

They discuss a causal relationship between negative valuation of schooling and bullying (victimization and perpetration), but it seems to me that the causal direction has to be going in both directions for this. For example, status as a teacher’s pet (presumably high value of schooling) may lead to victimization, but that victimization of a teacher’s pet might reinforce this victim’s high value for schooling as well.

When I raised the issue of causal directionality in the discussions, Victor Kobayashi (correctly, in my mind) suggested that we ought to think about (and analyze) bullying in relational terms, i.e. as a relationship between perpetrator, victim, and also teachers, esp. the homeroom teachers that Akiba et al. focus on in their analysis.

The focus on homeroom teachers immediately begged a comparison with owner-operators of juku. As is the case with homeroom teachers, jukucho take/bear a special responsibility for the development of students beyond their scholastic progress. For some jukusei enrolled in a small juku past their transitions through the educational system, their link with a charismatic jukucho may be the strongest social link and bond with anyone in education. A measurement of that link with a jukucho, especially for students who maintain a strong social tie with a juku (this would clearly be much more common in smaller juku than in the highly competitive urban corporate juku) would make for a fascinating control variable for Akiba et al.’s analysis, though the data obviously do not exist for this.

In a survey, I would expect that we would find a greater proportion of students in middle school reporting a strong bond with a jukucho than with a homeroom teacher.

The juku comparison also came to my mind because I have asked jukucho about bullying directly and have always been told that there is no bullying in juku. I have discounted some of that response as a PR claim (I suspect, for example, that bullying incidents that occur in transit between school and juku would be blamed on schools, not the juku by most parents), but have taken the explanation seriously that jukucho can threaten expulsion from the juku in a way that school principals or homeroom teachers cannot. Akiba et al.’s finding of the significance of a strong bond with homeroom teachers suggest that low rates of bullying in juku (if there were data to support this claim) could be associated with strong social bonds with juku teachers. A comparison between owner-operated and corporate juku would be illuminating in this regard.

“Student guidance” may also play a role in shaping social relationships in some juku.

One of the empirical findings that Akiba reported is the decline of incidents of bullying across middle school grades (i.e., less bullying in 9th than in 7th grades). Could this be linked to the kohai-sempai relatonship? I.e. sempai are not bullied (lower incidence for grade 9) and may not consider their interactions with some kohai bullying and thus don’t report them as such.

Supplementary Education Stepping Out of the Shadow Part II: Comments and Observations

All comments based on presentations rather than papers (for now, perhaps).


Brehm and Silova characterized the fact that shadow education may be supplanting conventional schools in importance as a “uniqueness of the Cambodian context”. As the subsequent discussion also showed, this is not unique to Cambodia at all and may in fact be part of a broader trend described by the title of this post, i.e. that shadow education is stepping out of the shadow. In the Cambodian context this stepping out of the shadow is occurring (as the presentation showed) through the role of conventional schools as an entry point to tutoring. Since it is teachers themselves who are offering tutoring (this is one of the common characteristics of shadow education in developing countries where it is mainly linked to low salaries for teachers), classes in conventional schools (already curtailed by the infrastructure need to double cohorts in school facilities creating a short school day) are a funnel into gradations of tutoring, “extra study” and “extra special study” in local parlance.

Hong Kong:

The pop start character of some tutors obviously distinguishes HK and is something that is not at all widely visible in Japan. While there are certain juku and yobiko instructors in Japan who have a bit of a start following, the large billboards of teams of prominent tutors that can be found in Hong Kong make for a celebrity status that seems somewhat outlandish in most other places. A couple of years ago CNN ran a report on one such celebrity tutor that Mark Bray also referred to as part of his presentation.


I did not know anything about shadow education prior to this presentation, so it was great to learn more, even though it was not surprising to hear that supplementary education is a substantial sector in Malaysia as well.

In contrast to Hong Kong, Cambodia, Japan and elsewhere, “private tuition” in Malaysia does seem to remain in the shadow in that Kenayathulla responded to a question that there is no sense of “the real learning is happening in shadow education”.

Another very interesting aspect of shadow education in Malaysia is the different use of tutors by ethnic groups linked to language needs and preferences.

Now some themes that I saw in these papers and which I talked about briefly at the session in my role as discussant.

1. The interaction between conventional schools and shadow education seems to be shifting in many jurisdictions. When I first began with my research on juku about six years ago, juku in Japan seemed very separate from schools, public or private. Recently, we’ve seen some occasions/spaces/programs in Japan where that separation is being watered down, for example through so-called 校内塾 (kônaijuku), that is juku within schools, that are offering juku services, aka classes, on school grounds in the afternoons/evenings or on weekends. This is primarily occurring in within the 23 wards of Tokyo to my knowledge though may also be spreading.

That’s one version of shadow education leaving the shadow. The other version is the anecdotal reports (including on Cambodia and Hong Kong in the presentations and Q&A) that students increasingly (over time? cross-regionally? what ages?) hold the view that their “real learning” is occurring in juku and that they sleep in conventional school to preserve their energy for juku classes, or because they studied until late into the evening in juku and are thus tired.

Shadow education thus seems to be increasingly (primarily over time) exerting an influence on conventional schools. Mark Bray spoke of “backwash” to the school system in this context, or of a “blending” of schools and shadow education.

This blending was also a theme, by the way, in the discussions at a workshop on “The Worldwide Growth of Supplementary Education” that I co-organized with Janice Aurini (U of Waterloo) and Scott Davies (McMaster U) last June at Waterloo.

2. Parents’ and students’ choices are increasingly driven by widely held perceptions of the quality (or, generally, lack thereof) of conventional schools. In the discussion and Q&A Mark Bray used the metaphor of shadow education as a “virus” infecting school systems. In this metaphor, popular perceptions are clearly the factor that are significantly weakening school systems’ immune systems and making them susceptible to this virus. The fact that the spread of this virus is not at all based on any established facts or, God forbid, research on the efficacy of tutoring, doesn’t surprise Bray at all, since educational policy has rarely been based on real data and evidence in other areas either.

3. While shadow education in developing countries (say, Cambodia), industrializing countries (Malaysia, perhaps), and developed countries (HK) may be increasingly similar in the breadth of its impact on conventional schools, one of the main distinguishing features that remains is the organizational form. In Japan, across East Asia, but also with some of the cross-border M&A activities in Europe by growing concerns like Acadomia (based in France) or Studienkreis (based in Germany), shadow education in developed countries is increasingly taking on the characteristics of highly institutionalized industrial sectors or organizational fields. In developing countries, tutoring continues to be a more personalistic affair.

4. There are some areas where public/state education policy is preserving its influence very strongly.

  • curriculum: almost all academic shadow education continues to focus on the content defined by public curricula and courses of study, even if this content is often mediated by (entrance) examinations of various kinds and thus not set in its specificity by public policy makers.
  • transitions: the progression from one level of education to another (primary to secondary, secondary to vocational, etc.) is still governed by the structure of the education system as it is determined by public actors
  • policy makers are experimenting with regulations of shadow education. The longest-standing example is the South Korean state’s battle against shadow education in the name of (in)equality, but the no more than 45 students per classroom policy in Hong Kong, or voucher systems in Malaysia, are clear examples of more widespread (albeit ineffectual for the most part) experimentation with the regulation of shadow education.

5. Inequality, inequality, inequality. All kinds of inequalities seem to be exacerbated by shadow education: economic, rural/urban, ethnic, etc. Inequality in access to shadow education is also believed to lead to inequality in education outcomes, though that is conditional on the unproven efficacy of shadow education.

Supplementary Education Stepping Out of the Shadow

Wonderful session at CIES this morning:

Markets, shadows, and schools: The impact and implications of private tutoring in Asia
Chair and Organizer: Mark Bray, Comparative Education Research Centre, University of Hong Kong

The hidden privatization of public education in Cambodia: Quality and equity implications of private tutoring” William C Brehm, This Life CambodiaIveta Silova, Lehigh University, USA

“The evolving shadow: Supplementary private tutoring in Hong Kong” Mark Bray

“Tuition syndrome: Determinants of private tutoring in Malaysia” Husaina Kenayathulla, Indiana University, USA

First, some quick summaries/observations of the presentations/panel. Then, some broader comments inspired by the panel discussions.

Brehm/Silova on Cambodia:

They argue that private tutoring is a conduit by which traditional social relations (primarily hierarchical) are reproduced in education even though the public education system is committed to the provision of free education to Cambodian students.

Data come from fieldwork in Cambodia that include interviews with teachers and students and focus on the costs and organizational structure associated with private tutoring. Through this fieldwork Brehm/Silova are able to offer summary data on the cost of tutoring in different categories, i.e. by level of education and rural/urban location. These costs range from 200 Riel/hr (approx. C¢5) to 16,000 Riel/hr (C$3.75).

The presentation highlighted the fact that students are classified into different achievement groups in public schools and that this classification may well have come to be based on students’/parents’ willingness/ability to pay their school teachers for tutoring.

This intertwining of classroom practices, teachers’ salaries and private tutoring is increasingly turning public schools into a mere ‘façade’ on an entry point to private tutoring.

Bray on Hong Kong:

Bray relied on visuals to provide a striking impression of the current context of private tutoring in Hong Kong. From videos focused on celebrity tutors and their students, to photographs of the splashy advertising that the tutoring industry decorates Hong Kong with, this industry clearly has become a very visible part of schooling in Hong Kong.

As an anecdotal aside Bray mentioned that classroom size is limited to <45 students (by regulation), but that celebrity tutors circumvent this by having glass divisions between classrooms staffed with “dummy tutors”.

Kenayathulla on Malaysia:

Kenayathulla presented her models of the likelhood of spending on private tutoring and the amount spent.