1 Introduction

The World Bank and other global actors were impressed by the fact that the relatively poor Southeast Asian country Vietnam made significant success in the Programme for International Student Assessment (PISA) evaluation in 2012, which tests literacy in mathematics, science and reading among 15-year-olds (approximately 5000 students in each country participate). In total 65 countries participated in this particular survey, and Vietnam placed as 17, ahead of several significantly wealthier Western countries such as Australia, Sweden, Norway, UK, and France [36]. In the 2015 survey, the science scores were high (525) whereas reading scores (487) and mathematics scores (495) had decreased somewhat [42]. This has led a number of researchers within for example education economics and psychology to analyze the achievement-related factors within Vietnam’s educational system and consequently explain why the cross-nationally measurable educational results are relatively good even in a global context, what characterizes its education system, and to some degree also what other nations, policy specialists and individuals might learn from Vietnam in this regard. Moreover, some have investigated the slight decline of scores (e.g., [10]).

At this point, a plethora of critical academic works have emphasized the downsides of PISA, both on methodological and other grounds (e.g., [47, 57]). Further, PISA scores are, according to some researchers, not good indicators of educational quality [26, 55]. Moreover, the PISA machine is a costly procedure in which the benefits are by no means clear [55] and educational policy borrowing can be quite problematic [56, 57, 66]). Hence, one may prefer a more balanced, critical yet grateful approach—if the Organization for Economic Co-operation and Development (OECD) has done something positive it is to provide researchers, politicians, and laymen with more data pertinent for cross-national comparisons [56]. Due to the richness of data, there is advanced econometric and multivariate research which disentangles relevant input factors associated with PISA scores among high performers such as Finland, Shanghai, and South Korea (e.g., [17, 34, 40, 45]).

Whereas the education system in Vietnam has been highlighted in earlier social science and historical research written in English, for example in William Duiker’s work Vietnam: Revolution in Transition (1995), the major focus seems to have been the broader historical development, the Vietnam War, and the political-economic dimensions rather than specific educational topics. However, there is growing number of scholarly works which, at least in part, treat the Vietnamese educational system in depth. One example is the journalist Bill Hayton’s work [19]. Another is Skilling up Vietnam [3], which is compiled by economic researchers and published by the World Bank. This work focuses to a large part on economic development, but examines broader societal development, including education in relation to vocational demand on the local and regional job market. Welch [67] analyzes Southeast Asian countries, including Vietnam, in relation to China’s economic growth but the study is mainly concerned about higher education and corruption.

Therefore, a deeper and more precise thematic review of the educational achievement-related literature on Vietnam—related to PISA in particular—is required. In particular, systematic literature review approaches are lacking in regard to Vietnam’s PISA performance. What is pertinent is to tease out the factors which have made Vietnam perform well, relative to its economic developmental level, and what may be improved as the scores are substantially below for example China, Japan, Taiwan, and Estonia (e.g., [41]). Thus, the current article aims to fill the gap and synthesize results from different pertinent sources to answer the following research questions:

RQ1: What has led to Vietnam’s high level of educational achievement, such as in PISA 2012 and 2015?

RQ2: What factors may improve Vietnam’s educational achievement?

After an historical overview relevant for the education system, the method and data, and result sections are presented. A tentative conceptual framework is followed by a discussion which summarizes the findings and concludes the article.

1.1 Education in Vietnam from a cultural-historical and economic perspective

As a consequence of war and turbulence during more than two centuries, the political and economic development has been stagnant in Vietnam. Only after 1975 the country has become stable and enabled material and economic development [59].

An important step in Vietnam’s political-economic, cultural and material development was, however, the reforms, Đổi Mới, that took place during the 1980s and forward and were launched by the Communist Party of Vietnam, which de facto has ruled the country without political pluralism or competition since 1976, roughly a year after the end of the Vietnam War. Similar to China and the reforms of Deng Xiaoping in 1978 onward, the Vietnamese economy was restructured from a central-planned and to a largely agrarian such, to the gradual incorporation of market reforms and increased integration into the global economy, which occurred in 1986 [2]. On average, the Vietnamese economy grew with 7.5 percent between 1995 until the financial crisis in 2007/2008, which is a few percentage points below the Chinese average total GDP growth rate throughout the same period. Between 2008 and 2012 the growth rate has been 5–7 percent [3], which reflects structural challenges and an infrastructure which not really matches the needs and demands of the growing middle class [70]. However, the educational system has to a large degree been developed and reformed during the decades that have been characterized by economic growth. Further, poverty rates have significantly decreased: from 37% percent in 2000, to below two per cent in 2018 [61].

With regard to continuous curriculum reforms, the demand of English and Korean among the future workforce has not yet been reflected upon these. Instead, the major changes throughout the last 10–15 years largely concern the increased quality of teachers, smaller classes, and a higher degree of modern instruction modes and strategies. The current education system consists of a 5-4-3 structure, which implies five years of primary or elementary school education, four years of middle school or lower-secondary school education, and three years of high school or upper-secondary school education ([36]). Furthermore, over the least years the influence from the United Nations and UNESCO’s Sustainable Development Goals by 2030 have become imbedded in educational policy discourse [61].

2 Methods

2.1 Overall procedure

System literature review is a conventional method in the social sciences, inclusive of educational research, which is used to synthesize research on a given phenomenon or related phenomena. Some systematic reviews include meta-analyses, which quantify and compare effect size relationships between different studies (e.g., [37, 60]), but when study designs are different and research questions do not correspond to quantification particularly well a qualitative synthesis without meta-analysis might be preferred. Some have criticized the confounding or merging between systematic and narrative literature reviews [16] but this interaction is mostly a consequence of the necessity of more objective features of all literature reviews. The current study took advantage of the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) statement’s criteria for reliability and transparency [37] and combined these with qualitative features such as a thematic analysis (e.g., [28] discussed below).

2.2 Search strategy and inclusion criteria

Below, I describe the procedure and search strategy, which also is manifested in the flowchart figure (see Flowchart Fig. 1), largely in accordance with the PRISMA flow diagram [37]. Three databases, Web of Science, ERIC, and Google Scholar, were systematically searched using a comparable search strategy in June 2020. These databases were selected because they are some of the major ones in the fields education and social science research. These included the years 2012–2020, although with minor variations in regard to the constitution of the search engine or database. The aim was to identify articles around 2012 and later, which highlight research on Vietnam’s educational achievement and PISA performance in 2012 and 2015 and also includes knowledge on the educational system in Vietnam more broadly but nevertheless could provide cues on Vietnam’s PISA achievement. A specific inclusion criterion was that the study either had a particular focus on PISA performance, or its content was relevant for explaining Vietnam’s PISA performance.

Fig. 1
figure 1

Flowchart figure

Full size image

The search words were “Educational achievement Vietnam”, “Vietnam PISA”, and “Vietnam PISA education”. Since there might be relevant conference papers or working papers related to Vietnam’s educational system or PISA performance, the author included not only peer reviewed articles, books, and book chapters but also non-refereed data of high quality and relevance. A quality criterion was that they included new knowledge on Vietnam’s PISA performance.

The screening for eligibility in Google Scholar after the search, which resulted in thousands of hits, ended when it was no longer relevant to continue. Therefore, these specific search results were not included in Flowchart Fig. 1. In addition, the reference sections in eligible studies were examined for additional studies which had not already been included or excluded at that stage. All studies had to be written in English for consistency [28, 29].

2.3 Thematic analysis

Inspired by Krippendorff [30] and Fereday and Muir-Cochrance [12], the included studies were also thematically coded based on in particular the analytical discussion or results sections.

Fereday and Muir-Cochrance [12] emphasize the consistency between the researchers typifications and “actors understanding in everyday life”, in the present study implying that a non-specialist should be able to grasp the meaning of the main themes. The themes identified in the included studies are mainly descriptive and closely situated in each study’s text. Specifically, coding in this regard meant to summarize its main findings into a coherent theme, as well as to categorize each study’s main method and identify the sample size. This is manifested in Table 1. It is notable that many studies in education economics use a mix of regression and decomposition analyses, but for the sake of brevity and simplicity these studies were coded as having regression analysis as its their main method. All in all, the descriptions and categorizations of the literature review rely on a basic thematic content analysis (e.g., [30], see also [28]). From the basis of each included study, a number of main themes were discerned. Hence, each study’s theme might be regarded as sub-themes [4, 5].

Table 1 Included studies
Full size table

In addition, a descriptive and critical summary of each major theme was presented (e.g., [35]). The length of such commentary depended on the chronological order, the novelty and importance of the theme (some findings in later study echo that of earlier counterparts), as well as the limitations of the article format. As a number of of these studies might be thematically connected, similarities and differences are discussed in relation to some of these studies, as well as to other related literature.

2.4 TALIS [58]

Aside from PISA, TALIS (Teaching and Learning International Survey) constitutes an important educational survey (e.g., [23]). Hence TALIS [58] was examined as it is important to critically discuss why Vietnam’s PISA results have declined, to the extent that it is possible to find an explanation for this recent negative trend. Since Vietnam is not an OECD country it is not included in the general OECD reports (e.g., [43]) but these sources provide some pertinent macro-social and educational data in that respect. In each country, a representative sample of approximately 4000 teachers and their school principals responded to a survey covering nine major aspects of the school system. In Vietnam 3825 lower-secondary school teachers and 196 principals completed the TALIS survey [58].

3 Results

After excluding duplicates and irrelevant search hits and coding the potential studies in accordance with the inclusion criteria (see Flowchart Fig. 1), in total 20 studies were included (including TALIS). Overall, five main themes were identified, below listed in alphabetical order: character skills (N = 1 study), cognitive ability and culture (N = 2 studies), culture and curriculum (N = 1 study), equity and quantity (N = 9 studies), and improvement of the education system (N = 7 studies). Because cognitive ability and culture are closely studied, especially in Rindermann et al. [50], they are merged under a single headline in the narrative review.

In purely quantifiable terms it seems that equity and quantity is the most researched factor(s) in this regard, followed by improvement of the education system. That is likely because Vietnam is still characterized by relative poverty and equity discrepancies between the urban majority Kinh group on the one hand and the rural ethnic groups on the other hand. Researchers in development economics and similar disciplines belief that these explain PISA differences. In other words, the more affluent groups (i.e., in particular the urban Kinh group) has benefitted from economic development and educational reforms the 1980s onward whereas the mostly rural minority groups have not benefited to the same extent. Hence, the second research question might simply be answered by the fact the if these equity disparities were drastically reduced, Vietnam’s PISA scores would likely increase as the low-performing fractions of participants would increase their test scores (e.g., [10]). However, in the cross-national context Vietnam’s greatest comparative advantage, at least relative to countries with similar developmental levels, seem to be related to culture and cognitive ability [25, 44, 50]. These, sometimes intricate relationships, are further discussed below in relation to the six main themes and 20 included studies.

3.1 Character skills

The country case study conducted by Bodewig et al. [3] comprises five chapters, of which four focus on the skills that are required to be able to obtain education at many levels and to succeed within a plethora of vocational fields in contemporary Vietnam. Chapter one describes the economic and structural changes that have occurred in Vietnam since the 1980s, while chapter two and three predominantly highlight the crucial skills that are necessary in the current and future job market. It does also examine the so-called STEP Skills Measurement Project that Vietnam has participated in, and in conjunction with interviews with employers it indicates that Vietnamese people who are employed at international firms or corporations which operate in the country tend to lack adequate educational and character skills, for instance critical thinking and interpersonal abilities.

Regarding specific test results for adolescents, such as PISA, this study tells us little. A cue from the study, however, may be that employers within the service sector claim that adult Vietnamese people possess adequate basic literacy and numeracy skills but lack analytical abilities in relation to for example writing and compilation of more extensive texts and documents. Necessary computer skills are also inadequate. OECD measures of adult competence like PIAAC (Programme for the International Assessment of Adult Competencies) show that they perform below the average. This may to a large part be a consequence of an educational lag, relative to the West and wealthier and more well-developed countries within the Asia Pacific region. One may suspect that more recent measures will demonstrate greater skills among Vietnamese adults who participate in such surveys. It is mainly Job-related skills that are requested among employers and much less favorable character or personality skills (e.g., [20]).

Overall, this study provides the reader with an important insight of the economic development of Vietnam and how it is intimately connected to the reformed education system. Great progress has already been made but further advancement is required if the goal is to increase long-term economic growth and fully integrate the country in the regional and global economy. However, it lacks a self-reflexive, critical dimension as it does not question why the integration into the regional and global economy is supposedly the only available option as a normative course for Vietnam’s educational development. This might demonstrate the disadvantage with global organs such as the OECD and the World Bank. Moreover, it does not discuss national-cultural factors in any significant depth.

This study does to some extent answer the second research question but suggests, although indirectly, that the parents of today’s students in Vietnam (at lower-secondary level) lack some cultural resources to help their children in an adequate way with regard to academic performance. From an input-process-output perspective (e.g., [34]) this means that parental input might be somewhat insufficient, as perhaps also indicated by parenting styles [50].

3.2 Cognitive ability and culture

In Rindermann et al. [50], these researchers have studied a sample of German students in the city of Chemnitz in the State of Sachsen and compared these with two classes from the city Hai Phong in northern Vietnam, and the city of Pleiku, located in the central area of the country. The cognitive ability levels of the students and tests similar to those of PISA have been measured and analyzed. Because of that Vietnamese sample groups were on average five months younger than their German counterparts these differences have been weighted with great accuracy. Other qualitative and quantitative dimensions of education, for instance data which reflects so-called parenting styles, have been collected by means of questionnaires that both parents and students have answered.

It is notable that both countries demonstrate a learning culture, although Germany is mainly associated with the Protestant work ethic and Bildung whereas Vietnam is associated with Confucianism. These two concepts have partly been used to establish cross-culturally valid analytical categories. Such are culturally adapted to further suit national-cultural characteristics. However, differences exist since German parents own more books, are better educated, and use a more structured and warm parenting style (authoritative), whereas Vietnamese parents tend to use strategies that are located in the intersection between authoritarian and neglecting. The latter implies that they have high expectations of their children but spend less time and concern for them, which partly reflects both material conditions and larger family sizes. At least from a Western viewpoint, authoritative appears to be the most optimal parenting style, but on the other hand this real or potential advantage must be weighed against the fact that German parents are divorced to a greater extent than their Vietnamese counterparts. Divorce does generally have a negative effect on children’s educational performance. It is also noteworthy that German parents help their children with homework to a larger degree, while Vietnamese parents tend to send their children to cram schools after the regular school day.

Regarding the specific test results that are presented in the study, these two countries have similar results, which is quite remarkable since Germany has a GDP per capita level which is roughly 20 times higher compared to Vietnam. For instance, the samples have a measured IQ of approximately 99 (Vietnam) and 100 (Germany). This indicates that resource allocation and economic development is not as relevant as factors as self-discipline, diligence, culture, and the quality and quantity of education. Specifically, the Vietnamese students have obtained somewhat greater mathematical knowledge, abilities that are linked with training rather than “pure” intelligence. The researchers note that the demands are higher in Vietnam and the students seem to have possessed higher degrees of self-discipline and perseverance when the mathematic tests were conducted. This reflects a Confucian-oriented culture. Thus, students’ background factors such as culture, cognitive ability, and parental expectations matter for the output [50].

This is indeed a comprehensive and insightful study which considers cognitive, non-cognitive, cultural, and material factors. However, the total sample size is relatively small, and the German children constitute a convenient sample, since Heiner Rindermann is active as a professor in the city of Chemnitz.

Iyer and Moore [25] explain in depth how the compared data from the large-scale Young Lives survey in Ethiopia, India and Vietnam were construed and obtained. In Vietnam, the first wave of the sampling procedure resulted in the inclusion of 52 upper-secondary schools and approximately 9000 students. In the math test of the first wave, the high performance of Vietnamese participants is underlined. Contrary to some notions on the Vietnamese education system and curriculum, a substantial share of Vietnamese students could solve abstract reasoning items not typically aligned with the national curriculum. For this purpose, paper-and-pencil tests from the PISA 2003 survey were used. On the other hand, as expected, Indian schoolchildren are better equipped to deal with English tasks and items. Based on findings and interpretations of this study, it is a bit unclear which inputs that seem to matter for PISA achievement. However, cognitive abilities were higher in Vietnam around this time (e.g., [32, 51]) and therefore it seems to be a cognitive skill advantage.

3.3 Culture and curriculum

After applying econometric analysis, Parandekar and Sedmik [44] found that Vietnam’s specific features, which seem to be associated with diligence and discipline, explain about half of the differences of Vietnam’s high achievement relative to countries of similar economic development. Thus, such cultural and family input factors are processed in the school system and leads to a rather high output (i.e., PISA achievement). They underscore that truancy rates are lower in Vietnam relative to comparable developing nations. Moreover, in relation to educational quantity as related to student effort, they underline that Vietnamese students, on average, spend roughly two hours more on supplementary education compared to the seven compared countries. A disadvantage with this study is that it is solely based on PISA 2012 data. Instead, comparisons between the 2012 and 2015 surveys are examined by Glewwe et al. [14].

3.4 Equity and quantity

Rolleston and Krutikova [53] is the first published study which builds on Young Lives (YL) data in Vietnam, which was collected at 20 sites in five provinces in Vietnam. The sample is somewhat biased for participation of poor students, which is related to the purpose of the study and the emphasis on inequities. The authors examined learning achievement of students in Grade 5. They found that basic educational opportunities are relatively equally distributed, but in relation to more sophisticated learning environments and opportunities the inequalities between home disadvantaged and home advantaged pupils are more striking. Home advantaged means that families have more economic and cultural resources and, for instance, can afford to pay a fee for having their children in school for an entire school day and supplementary schooling, whereas poorer families can only pay for half a school day. Furthermore, advantaged students do typically have more access to superior school resources such as teachers with higher levels of subject knowledge. While not discussed by the authors, this might imply that disadvantaged schoolchildren will likely perform lower in assessments such as PISA. Hence, students’ background factors such as parental income and cultural resources (input), but also the length of the school day (process), matter for the achievement output.

Glewwe et al. [15] used Young Lives data collected in 2006 and found, among 12-year-olds as well as a younger cohort, that ethnic gaps between the Kinh majority and minority groups exist. The higher education of Kinh families, in tandem with language skills deficiencies among many minority members, explain some of the gaps in cognitive tests. However, upon a closer interpretation it is more likely that living in a mixed commune rather than language per se is what affects test scores. This is because segregated communes do not have access to Vietnamese language and culture in the same way as ethnically mixed communes. This means that ethnic minorities’ lower performance is related to a cultural (partly linguistic) input factor in the school process.

A limitation of this study was that it did not even mention the strong PISA performance in the 2012 survey and how the ethnic gap and enrollment rate disparities might affect the achievement levels. However, later studies by Paul Glewwe and colleagues (e.g., [14]) address such issues.

Glewwe et al. [14] have applied econometric analysis to Vietnam’s PISA performance in 2012 and 2015 and examined factors such as enrollment rates and socioeconomic status among participating cohorts of students, specifically child, household, and school level characteristics. While the enrollment rates may skew the results positively, meaning that Vietnam would have scored lower if a larger share of low-SES cohorts would have participated, this is insufficient to explain the high achievement levels relative to countries of similar or slightly higher GDP levels. Among “comparable countries”, only Chile’s top 50% outperform Vietnam in the 2015 reading assessment [14]. The regression analysis and PISA data have also been compared with multiple variables from the Vietnamese Household Living Standards (VHLSS). Overall, this study stresses the nexus between educational quantity and socioeconomic characteristics in relation to PISA results, which signify both input and school process factors. Vietnam performs well, despite its low GDP per capita, relative to comparable middle-income nations but the factors underlying such an achievement are perhaps better explained by Parandekar and Sedmik [44].

The study by Nguyen and Nguyen [38], situated within the fields of the economics of education and developmental studies, indicates that education plays a substantial role in intergenerational as well as intragenerational mobility. A strength of the study, apart from the advanced statistics, is that it integrates a sociological approach into a larger developmental framework. However, the authors do not critically discuss to which extent these sociological theories, which are based on the works of Pierre Bourdieu, are applicable to the Vietnamese sociocultural context. Nonetheless, education is, or at least has been, a means to reach higher on the socioeconomic ladder in Vietnam. Thus, higher access to education across all socioeconomic strata and ethnic groups might increase output in school surveys such as PISA.

According to Nguyen [39], the Kinh majority has a higher educational achievement than minority groups due to higher socioeconomic status and better geographical locations. The results of the study, which were primarily based on YLS and YLSS (Young Lives School Survey) data, show that educational achievement differences between the Kinh majority (86%) and the non-Kinh minorities are mostly attributable to inputs like household characteristics such as family expenditure and the father’s educational level. That implies that socioeconomic status, geography, and ethnicity often overlap in relation to educational achievement. Moreover, it means that even though the data sets focus more on poor groups, minority groups also perform lower in surveys like PISA.

Bui and Imai [7] weighted decomposition and regression analyses of VHLSS data (2008–2012) found that basic education is beneficial for the rural poor. Hence, the government should facilitate improved and educational opportunities, such as a helping the rural poor to enter university to a larger extent, as well as supporting the agricultural sector in which a large share of rural families works. This might improve educational output among the lowest-performing fraction of students because of the nexus between inputs like parental education and overall living conditions and schoolchildren’s learning output.

build on earlier works of the authors, as well as that of the World Bank [69], Glewwe et al. [14] and Dang and Glewwe [10]. However, the focus is on the characteristics of students in Vietnam who do or do not continue to Grade 10, which is the first year of upper-secondary school. Moreover, the authors underline that these findings indicate that access to post-basic education in Vietnam is not as equitable as at earlier stages, where access to education is almost universal. In an equitable system, where access is purely determined by merit, one may not expect wealth or caregiver’s education to be an important predictor of access after controlling for prior performance [9]. Essentially home advantaged students, which have more resources at their disposal, are overrepresented in the transition into Grade 10. However, if disadvantaged students have high academic results, they can counterbalance the wealth effects. Earlier interventions in elementary school might mitigate the equity problem and improve output (i.e., grade transition and performance in surveys like PISA). Thus, this study provides some cues for research question 2. The input factors echo those of several other studies summarized in this review.

Tran and Pasquier-Doumier [63] used Yong Lives data and investigated differences in Vietnam regarding specific dimensions of educational quantity and equity. As noted by Dang and Glewwe [10], a substantial part of Vietnamese pupils is only given access to half school days (roughly 3.5 h). Increasing instruction time would, ceteris paribus, increase low-SES children’s achievement and intranational equality. However, as noted by the authors increasing instruction time has not narrowed inequalities since there are more subtle ways (e.g., supplementary education) for high-SES family to support their children and be a step ahead of low-SES families and pupils in regard to educational investment. This questions the notion that increasing educational quantity in such regards has much of an effect, at least if it is not combined with improving opportunities of the disadvantaged schoolchildren as parallel processes (i.e., after school classes). Thus, both input and school processes matter for output in the current Vietnamese context, why narrowing inequalities of opportunity might improve academic output (e.g., grades and PISA performance) and the education system more broadly. The latter is affected by broader economic, material, and social conditions and factors such as parental income and geographical location.

Using a multivariate framework, Asadullah et al. [1] compared dependent variables related to Singapore, South Korea (which outperforms Vietnam but do also have substantially higher GDP per capita), Indonesia, Malaysia, and Thailand (which performs lower than Vietnam but only have a somewhat higher GDP per capita). After controlling for GDP, educational spending, the Gini coefficient (i.e., the income inequality index within a country), student, family, and institutional characteristics, the authors conclude that cultural factors, in tandem with institutional factors, appear to explain Vietnam’s high PISA achievement. They stress, “equalizing the socioeconomic background differences among students from the two countries (i.e., Vietnam and South Korea) would give Vietnamese students an even better performance in the PISA Science test.” However, a shortcoming of this study is that the lower mathematics scores among Vietnamese students are not examined. Vietnamese educators and policy analysts might be keen on more information about how to be as high-performing as for instance Singapore in mathematical literacy.

3.5 Improvement of the education system

Mcaleavy and Elwick [36] focuses on HCMC and thus do have a more local emphasis. At the same time, it reflects more general intranational conditions and highlights reports on Vietnam which have been conducted by international education and development institutes, which all praise the country’s good results, despite its relatively low educational level among parents to the current young generation. The quality of education within the country appears relatively high. That leads to that more people remain in the education system, which in turn leads to higher aggregated quantity in that sense that a larger share of the population are more extensively educated, but a significant portion (6.6% boys and 12.9% girls) do not continue with middle school education after the initial five years of elementary school education. The researchers summarize interviews which indicate that the equity within the HCMC is high, since the poorest pupils perform at the same level as the wealthiest counterparts in emerging markets like Ethiopia, Peru, and India. The supposed reason for this is a modern curriculum and the high quality of infrastructure, teacher competence, and instruction methods. Some of these processes may be related to the Education Development Strategic Plan 2011–2020, which has been developed and implemented in consultation with the World Bank. This has led to a large degree of modern didactics, often understood as a hybrid between traditional (teacher-centered didactics) and contemporary and modern teaching methods (for instance cooperate learning, active learning, and problem solving). This is also known as HERA (Higher Education Reform Agenda).

The educational investment of the government has increased because of a larger emphasis on education, something that generally leads to better results in emerging markets. However, large expenses are not automatically leading to increased achievement, at least after a certain threshold value has been surpassed. Part of the investment strategies has resulted in smaller school classes (20–30 individuals rather than 30–40) and better school buildings and infrastructure. However, it does not address any cost–benefit analysis of such input and econometric research which shows that input such as to raise teachers’ salaries and decreasing class sizes may have no to little effect on educational output in developmental countries (e.g., [18, 48]).

Another aspect that is underlined is linked to the head of school’s role in relation to its employed teachers. The former tends to a large part supervise and evaluate the latter, which reflects a top-down relationship. This seems to not have had any negative effects this far, rather the opposite. Coinciding with this development the salaries of teachers have increased, and the career development possibilities improved compared to the OECD average. However, this relative increase must be understood in relation to a low GDP per capita and average income level. Teaching is regarded as a high-status occupation but does not make a Vietnamese person particularly wealthy by any global means or standards. Indeed, this is part of a broader developmental pattern in which Vietnam started from a low base.

The methodology of the study is appropriate because it investigates characteristics of the changes that have occurred within the country, but the sample is small, and teachers may – even under anonymous conditions – only highlight the positive features of the development. Yet it provides a relevant and important insight into the education system since Vietnamese teachers in general appear to have a comprehensive understanding of how it works. Indeed, the quality of education has improved, which in turn partially affects performance in both intranational and international assessments like PISA. However, the study fails to critically examine the evidence that suggests that raising educational quality, related to smaller classes and higher salaries, may only have had a meager effect in Vietnam’s development. Further, it does not address local or national-cultural characteristics (i.e., Confucianism or Confucianism/western hybridity) or cognitive factors such as IQ in relation to literacy and numeracy (i.e., PISA). Other studies (e.g., [44, 50]) demonstrate the impact of such inputs.

According to Hoang [22], Pre-modern Vietnam was characterized by a central-planned system which was based on so-called top-down relationships. The superior strata of society exert a structural pressure on the country’s inhabitants to make them responsible for their own life situation, including education. According to Hoang, it is to a large part typical even for today’s Vietnam. In the education law that was launched in 2005, and revised in 2009, autonomy and self-responsibility have been emphasized as crucial features. A certain degree of decentralization of the education system has been incorporated, which provides local education administrators and teachers somewhat more autonomy, but since the course literature is expected to be evaluated on an annual basis it is, according to Hoang, difficult to affect the educational quality as a teacher. In that respect further reforms may be required.

With regard to autonomy and self-responsibility, the system has transitioned from controlling to achievement oriented. If administrative regions and schools can demonstrate good financial results when answering to regional committees, they will keep their autonomy, which mirrors the earlier structural suggestions posed by the World Bank. The strategies are partly based on developmental aid. A negative consequence, however, has been the “result craze”, which supposedly may have affected the instruction quality in an undesirable way. This may not have affected PISA scores, whether positively or negatively, but is rather linked to the annual evaluation of the stipulated course literature. It might be that good PISA results are merely a spillover effect from this direction.

This study provides a broad but rather incomplete understanding of how a certain degree of decentralization of the education system has been implemented in a country which otherwise is characterized by far-reaching centralization. Likely, these structural changes have predominantly affected the educational quality in a positive way, although the evidence is suggestive and inconclusive. It does only sketch a broad picture of Vietnam’s educational development, which does not consider econometric and psychometric evidence related to raising educational output through relevant input. Further, it lacks a critical vantage point since it takes for granted that the developmental strategies suggested by the World Bank are obviously correct and desirable. For instance, it may lead to the imposition of homogenizing methods and ideals which undermine national-cultural characteristics. Further, it does not discuss national-cultural components, such as the impact of increasing use of smart phones and the broader pattern of gradual decline of educational performance as a country becomes wealthier e.g., [27, 54, 65). Nevertheless, it provides some cues for improving the education system by means of for instance increased decentralization and novel teaching methods.

Doan, Le, and Tran [11] shows that returns to investment in Vietnam have decreased since the last economic crisis in 2007. This might indicate that the higher education is oversupplied, somewhat similar to a wealthier country in greater East Asia, namely South Korea [24]. While this is not particularly relevant for educational achievement around Grade 9, which roughly corresponds to the PISA age, it partially answers the second research question. Indeed, this might pose a future problem for Vietnam. Perhaps in the future, a portion of students might be less motivated to reach high academic achievement, which signifies an input factor, because they or their parents expect lower returns to such investment.

The conference paper compiled by Tran et al. [62] emphasizes the improved educational quality and enrollment rates, a dimension of educational quantity, during the period 2011–2015 in Vietnam. Specific improvements include increasing number of teachers and school administrators, as well as improved quality of such labor force, and these signify external inputs relevant for school processes and educational output. Many of these notions overlap those examined by Mcaleavy and Elwick [36] and Hoang [22]. Nonetheless, as these processes coincide with the two PISA surveys (2012 and 2015) it is likely that quality and quantity improvements are, in part, underlying the high achievement in these assessments.

In this extensive review, the Dang and Glewwe [10] stress that increased participation rates in lower-secondary/middle school education may have led to more low-performing individuals taking the PISA tests, which negatively affects the total results [10]. However, in PISA 2015 relative to PISA 2012 Vietnam’s Glewwe et al. [14] stress, “First, a major reason for Vietnam’s impressive performance on the 2012 and 2015 PISA assessments appears to be the low enrollment rate of 15-year-olds in Vietnam,of the 63 countries that participated in the 2012 assessment, Vietnam’s ‘coverage index’ was only 55.7%, the third lowest of the 63 countries. In 2015, it was only 49%, much lower than in any other country.” Hence, the enrollment rates as a student input factor are insufficient to explain the slight decline in 2015 relative to 2012. Moreover, the authors emphasize that Vietnam has many basic improvements to make, such as offering a full school day for all children and improving the conditions for ethnic minorities.

As educational quality and output is partly related to school process factors—inclusive of both teacher salaries and competence, school leaders, and learning environments—it is relevant to consider TALIS, the OECD Training and Learning International Survey. Whereas TALIS [58] majorly reports statistics on factors such as the age structure and gender distribution of teachers and school heads, it offers some insights about school environments and other indicators of educational quality which serve as a complement to the above-mentioned studies. For instance, the report underlines:

In terms of classroom environments, relations between students and teachers are positive overall, with 95% of teachers in Viet Nam agreeing that students and teachers usually get on well with each other. However, 2% of principals report regular acts of intimidation or bullying among their students, which is lower than the OECD average (14%).

On the other hand, other sub-variables such as the need for teachers to tell students to quieten down, show that Vietnam is situated slightly below the OECD average in that regard. A problematic dimension of this type of cross-national comparison is the inherent risk of reference bias [20, 49]. For instance, Heckman and Kautz ([20], pp. 352–355) notice that South Koreans define themselves as the laziest and most unconscientious national group, despite working the most hours within the OECD after Japan. In the Vietnamese classroom context, it might imply that teachers simply are more prone to convey to students that they have to quiet down while for instance their Western counterparts allow more talkative students, or let them study outside the classroom to a larger extent [58] where they are prone to use their smartphones (e.g., [54]).

On the other hand, the data in TALIS does also show that Vietnamese students are expected to deal with tasks more on their own compared to the OECD average, which yet again questions the simplified stereotype that Confucian-oriented students rely solely on teacher-centered education [21]. Confucian elements tend to go hand in hand with authoritarian didactic methods and learning contexts, at least in the classroom and in relation to parents [50], but it does also imply that individual self-actualization occurs by means of diligence and effort. This indicates that the learning environments in Vietnam thus far are relatively beneficial, both in regard to attentiveness and agreeable and conscientious behavior. It also emphasizes the relationship between macro culture (Confucianism-oriented learning) and micro-cultural elements, such as class environments but the picture is fuzzy and inconclusive. For example, the study by Rindermann et al. [50] does not consider student’s behavior in class or relates it to Confucian-oriented culture. It is also impossible to measure the extent to which a country becomes more “Westernized”, although smartphone use serves as a rather insufficient indicative proxy [65], and how it affects output in assessments such as PISA.

Tran et al. [64] examined how variations of a contextualized task allow students to experience the mathematical modelling process. The authors explored qualitatively how Grade 10 students (15–16-year-olds) respond to tasks with different levels of authenticity when they first encounter them. The study suggests ways to adapt contextualized tasks to make them more authentic, which allows students to experience a more holistic modelling process. While this study is interesting it may only have a weak indirect link to achievement in mathematical literacy such as PISA. However, as PISA is concerned with the application of knowledge in “authentic” situations it might be important to learn more contextualized tasks. This signifies a school process that may affect learning output measured by tests and surveys. Some studies reviewed in this article underline that Vietnamese students have a rather deep abstract knowledge [25]. However, the scores in the 2015 PISA surveys were only slightly above the OECD average.

3.6 Conceptual framework

A general theory linked to academic performance is the input-process-output (IPO) model, which is a part of a school effectiveness framework [34, 40, 45]. This shares some similarities and does partly build upon the estimation of production function of education framework, which is often used to analyze PISA studies (e.g., [17]). The IPO implies that students bring with them for example their cognitive and non-cognitive abilities, and parents’ expectations, and in tandem with various school factors (e.g., teacher competence, curriculum, school material) processes occur that lead to achievement output of various degrees [40]. Such achievement is linked to for example PISA tests and pupils’ performance [34]. Ma [34] stresses:

School effectiveness research employs an input‐process‐output model to describe how school experiences influence educational outcomes of students with differing family backgrounds. Inputs such as family characteristics, home influences, and family social and cultural values are what students bring into their schools. Schools, with different contexts and climates, then channel (or process) students with different inputs into different categories of educational outcomes (outputs).

In this regard, schools play an important role in how for example boys and girls are treated and perform, but also the home environment, culture and family characteristics matter as inputs into the school process [34].

Furthermore, Perelman and Santin [45] underscore that in quantitative research which builds on individual and classroom data, both inputs (e.g., motivation) and school factors (e.g., pedagogical choices, organizational structure) might be captured. In other words, it can be difficult to identify and differentiate between student and school characteristics. However, economic, and social factors may exclude segments of students within relatively poor countries which affect student outcomes in assessments such as PISA, why this might be a larger issue in Vietnam compared to for example Japan or the US (e.g., see [34]). Nonetheless, with regard to a developing country like Vietnam, there is no reason to believe that this generic school achievement model is not applicable—the question is how rather than if such processes within its school systems occur (e.g., see [44]).

Some of these relationships are illustrated in Fig. 2. There we may see that the earlier economic and educational reforms have, both the earliest in the 1980s and those in the early 2000s, have led to school improvements that virtually affected the bulk of families and students. Nonetheless, because of the many socioeconomic and educational inequities which are highlighted in numerous studies, the PISA scores will likely increase if these disparities are substantially reduced.

Fig. 2
figure 2

Conceptual framework

Full size image

4 Discussion

The study focused on what has led Vietnam to perform well in PISA 2012 and 2015, as well as on the factors which may improve the PISA test scores even more. Common for several research reports that have been critically summarized in this article is that the rapid economic development in Vietnam, in the aftermath of the Đổi Mới reforms during the mid-1980s, as well as educational reforms in the early 2000s, have paved the way for a set of structural, material and didactic changes that have increased both quality and quantity of education. The economic growth trajectory has been cumulative but steady. Elementary school enrollment is nearly 100% in today’s Vietnam. The World Bank and other global actors have been involved in such developmental processes (e.g., [22, 69]), including research endeavors [3]. This, in later stages, have likely affected learning outcomes in PISA 2012 and 2015. However, as many studies highlight socioeconomic disparities and inequities it is still much that needs to be addressed in this regard. Smaller disparities imply smaller differences between low- and high-performing PISA participants in Vietnam as well as, likely, higher average scores (e.g., [10]).

Moreover, the high levels of cognitive abilities, which seem to be on par with the West and parts of Confucian Northeast Asia might be another likely explanation, although not included as a variable among PISA researchers. Instead, diligence and discipline and more hours spent on homework are identified explanations for the higher achievement levels among Vietnamese students compared to those in other developing countries [44]. Thus, it seems as if culture and diligence are two main strengths that do not require much further enhancement.

With regard to cognitive abilities, on the other hand, there is a rich literature on cognitive improvement, processes that occur more or less automatically as a country develops and improves the quantity of schooling, nutrition, and family composition (smaller families) (e.g., [13, 51, 68]). Families that transition into a middle-class status might be informed about such strategies because they are also dependent on conscious decisions to some extent. To exemplify how well measured cognitive ability correlates with PISA performance, the case of China’s urban regions Beijing, Jiangsu, Shanghai, and Zhenjiang is illustrative. In the 2018 PISA survey, these countries had the highest scores, almost 579 aggregated average PISA points [41]. According Lynn, Cheng, and Wang [33], the average aggregated IQ in these four regions is 113.8, which is almost a full standard deviation higher than Germany and Vietnam. Singapore, South Korea, Taiwan, Hong Kong, and Japan do likewise have higher scores and outperform Vietnam in PISA [32], see also [4,5,6]. Due to cultural similarities, it seems reasonable that Vietnam may narrow the gap to China and other East Asian countries if economic and institutional development continues but the emphasis on education continues within the wider macro culture and among families and individuals. In other words, if these economic, material, and structural improvements occur (as in wealthier urban regions and countries such as Beijing, Shanghai, Taiwan, Hong Kong and South Korea), then the cognitive ability may increase which in turn likely leads to improved PISA scores.

As underlined in some studies, more Vietnamese people finish elementary school and lower-secondary education, although an obstacle is the extent to which Vietnamese children or adolescents drop out from the education system in order to focus on low-skilled work and/or family life. Income inequality and access to education beyond elementary school are other difficulties for a substantial part of the country’s inhabitants [10]. Apart from offering a full school day for virtually all schoolchildren in the country, free of charge after school classes, similar to Taiwan (e.g., [8]) might narrow the inequality of opportunity gap between different ethnic and social groups. Moreover, teacher and instruction quality must be of rough equivalence if full day schooling and after school classes should have more substantial effects on test scores, whether in international surveys like PISA or various assessments at the international or school level. Thus, such improvements may lead to both higher scores in surveys like PISA and enhance the education system more generally.

With both the impressive PISA 2012 results and the slight decline in PISA 2015 kept in mind, it is worth to ask if the Confucian education ethos will remain constant or gradually diminish over time. Indeed, the most optimal middle ground would perhaps be a combination of traditional diligence and modern creative thinking (for an in-depth discussion on Finland and East Asian countries in such regards, see [54]).

4.1 Limitations and directions for future research

This study has several limitations. In part because of its qualitative research design and the includes studies’ various designs, effect sizes were not compared and calculated. Moreover, studies in Vietnamese or any other language besides English were not included. However, the reference sections of the included studies show that most Vietnamese researchers or researchers on Vietnam from other countries active in these fields cite only other papers in English. A search round in Vietnamese in Google Scholar (search words: “giáo dục việt nam PISA”, “giáo dục việt nam”) found some interesting studies and essays relevant for future literature reviews (e.g., [31]). However, it must come with an understanding that most journals associated with higher education and research in Vietnam and the Vietnamese language might be of lower quality (e.g., [29]).

Furthermore, included studies were limited between 2013 and 2020, which however reflects the emphasis on research more or less relevant for PISA performance in the 2012 and 2015 surveys. Moreover, input-process-output interpretations might, in some cases, be somewhat far-fetched and lack an empirical basis. However, the aim was nonetheless to present the most likely or relevant explanation for the phenomena and findings highlighted in each study.

Further research may benefit from more comparisons between Vietnam and other high-performing Asia–Pacific countries such as South Korea, Taiwan, Singapore, and China [46]. It may also explore more cultural and psychological factors, as many studies critically summarized here are situated in education economics. Quantitative research may focus on larger and more representative samples of cognitive tests among both poorer and wealthier regions of Vietnam. The same goes for parenting styles and other cultural input factors. Such research endeavors could be relevant for educational psychologists and economists alike. Moreover, development and economic research with more recent survey data is relevant, especially if it focuses on factors pertinent for the education system and not just poverty reduction and economic improvement. For instance, geographical distance to schools and cities may capture important nuances relevant for academic attainment and performance. Qualitative studies, in which student voices are being heard, may also be of significance as such are typically lacking in the research literature.