Luận án hệ thống hóa được cơ sở khoa học (bao gồm cơ sở lý thuyết và thực tiễn) về bẫy thu nhập trung bình (BTNTB), xác định được các yếu tố khiến các quốc gia rơi vào BTNTB, những yếu tố giúp các quốc gia thoát BTNTB. Dựa vào các nghiên cứu quốc tế đi trước, luận án đã xác định được các cách để đánh giá khả năng vượt BTNTB của Việt Nam. Trong luận án, NCS đã kết hợp được nhiều phương pháp nghiên cứu từ phân tích, tổng hợp so sánh, tới phương pháp định lượng và phương pháp phỏng vấn chuyên gia. Những phát hiện, đề xuất mới rút ra được từ kết quả nghiên cứu, khảo sát của luận án Luận án đã xây dựng được phương pháp luận đầy đủ để nghiên cứu về BTNTB mà bất cứ quốc gia nào có đặc điểm tương đồng Việt Nam (tức là đang trong giai đoạn TNTB thấp) có thể áp dụng để nghiên cứu cho quốc gia mình, cụ thể như sau: Luận án đã khái quát hóa được những yếu tố chung và đặc trưng khiến các nền kinh tế vượt/mắc BTNTB tại châu Á; đồng thời áp dụng các cách xác định BTNTB, các tiêu chí so sánh và lấy ý kiến chuyên gia để đánh giá khả năng vượt BTNTB của Việt Nam. Luận án dùng mô hình dữ liệu bảng xác định được các yếu tố quyết định tới khả năng vượt BTNTB của Việt Nam và các nước TNTB thấp khác Châu Á. Cuối cùng, từ kinh nghiệm BTNTB trong khu vực, thực trạng kinh tế và khả năng vượt BTNTB của Việt Nam, các yếu tố quyết định tới khả năng vượt BTNTB của nhóm nước TNTB thấp tại châu Á, xu hướng phát triển kinh tế trên thế giới, định hướng phát triển kinh tế của chính phủ, luận án đưa ra hàm ý chính sách cho Việt Nam để vượt BTNTB thành công bao gồm thúc đẩy tăng trưởng nhờ tăng năng suất lao động, phát triển khoa học công nghệ, đổi mới, nâng cao chất lượng thể chế; chính phủ can thiệp linh hoạt vào nền kinh tế; nâng cao hiệu quả chi tiêu công và hiệu quả hoạt động của vốn đầu tư, tận dụng cơ hội từ hội nhập kinh tế quốc tế.
MINISTRY OF EDUCATION AND TRAINING FOREIGN TRADE UNIVERSITY SUMMARY OF DOCTORAL THESIS MIDDLE INCOME TRAP: INTERNATIONAL EXPERIENCE AND POLICY IMPLICATIONS FOR VIETNAM MAJOR: INTERNATIONAL ECONOMICS CODE: 9310106 LE PHUONG THAO QUYNH Ha Noi – 2022 The thesis is completed at Foreign Trade University Supervisors: Assoc Prof, Dr Tu Thuy Anh Assoc Prof, Dr Hoang Xuan Binh Reviewer 1: Reviewer 2: Reviewer 3: The research will be protected in front of the council meeting at The thesis could be read at National Library and Foreign Trade University Library INTRODUCTION Rationale of the research The concept of the middle-income trap (MIT) first appeared in a report by Gill and Kharas published in 2007 by the World Bank Accordingly, MIT is a situation in which countries have achieved a middle-income level but then stuck at that income level and cannot achieve a highincome level because they cannot maintain the same growth momentum as before Many countries in Latin America, the Middle East, Africa have been affected by MIT such as Brazil, Mexico, Iran, Lebanon, Egypt, Jordan, and South Africa, Thailand, etc Therefore, MIT has become a hot topic, regularly researched and analyzed by international organizations such as World Bank, IMF, ADB, OECD and receiving attention from researchers and policy-makers in the world According to World Bank’s GNI per capita, Vietnam has become a lower-middleincome country since 2009 with a per capita income of $1,120 (Beliner et al., 2013) Based on the country classification by income using the GDP PC PPP 1990 from the Maddison dataset (2010), Felipe et al (2012) show that Vietnam has become a middle-income country since 2002 with an income per capita of $2,023 In the period 2002 - 2010, the average GDP growth rate of Vietnam was 6.1% If Vietnam wants to escape the lower MIT, it must achieve an average growth rate of 4.3% in the period 2010 - 2029 It can be said that, if compared with itself on the time axis, Vietnam’s performance has been remarkable However, to become a developed economy with high income like Korea, Japan, and Singapore, there are still many difficulties Vietnam's economy in recent years has been successful with a high economic growth rate compared to the region and the world, stable inflation, much-improved trade balance, and balance of payments However, Vietnam's economy also reveals many shortcomings such as inefficient investment, low quality of labor, resource-intensive economy, underdeveloped financial market, banking system risks, environmental pollution, corruption, etc Many domestic and international studies have shown that Vietnam is currently facing the risk of falling into MIT The 13th National Party Congress also confirmed that MIT is one of the four great risks of Vietnam The 13th National Party Congress set the goals that by 2030, the 100th anniversary of the founding of the Party, Vietnam will become a higher middle-income country with modern industry By 2045 - the 100th anniversary of the country's founding, Vietnam will become a highincome country Therefore, escaping MIT to achieve the above goals is a scientific problem, a political task, and a national responsibility There have been many studies to answer the question “how to escape MIT?” with specific analysis, evaluation, and evidence However, posing the problem of comparative research to find out the determining factors of two groups "MIT escapees" and "MIT nonescapees", together with analysis from the intrinsic capacity of Vietnam's economy, thereby providing policy implications for Vietnam to successfully escape MIT, which is an academically convincing, appropriate choice - viewed from the context and practical conditions Therefore, the author decided to choose the topic: "Middle-income trap: International experience and policy implications for Vietnam" Through the thesis, the author offers a deeper study on the experience of escaping or falling in MIT of countries around the world, applying the methods of determining MIT and the model given by international studies to analyze the MIT in Vietnam and provide significant policy implications for the Vietnamese economy 2 Research objective 2.1 Overall objectives The overall goal of the thesis is to find out the difference between the group of countries that have escaped the MIT or have been trapped in the MIT, the factors that affect the MIT of countries around the world, and determine the ability to escape MIT of Vietnam Accordingly, providing policy implications for Vietnam so that Vietnam will be able to achieve sustainable growth and successfully escape MIT 2.2 Specific objectives Analyze the characteristics that help Japan, Korea, Taiwan, Singapore, Hong Kong successfully escape MIT Analyze the factors that make the Philippines, Indonesia fall into MIT Draw common and different characteristics between these countries and find out lessons that can be applied to Vietnam Quantifying the influence of factors in the economy on the ability to escape MIT of lower-income countries in Asia including Vietnam, India, Cambodia, Laos, Philippines, Myanmar, Mongolia, Bangladesh, Pakistan, Nepal, Indonesia, Sri Lanka, and Bhutan for the period 2002 - 2020 Based on the definitions given by international studies and consulted experts to determine the possibility of escaping MIT in Vietnam Finally, provide policy suggestions for Vietnam 2.3 Research questions Research needs to answer the following questions: i) What is the difference between the group of countries trapped in MIT and the group that successfully escape MIT?; (ii) What is the current economic situation of Vietnam and its ability to escape MIT?; (iii) What are the factors affecting the ability of Vietnam and other lower-income countries in Asia to escape MIT?; (iv) From international experience and Vietnam's economic situation, policy implications can be drawn for Vietnam to grow sustainably and quickly catch up with higher middle-income countries and then high-income countries in the world? Research subject and scope 3.1 Research subject The thesis studies the experiences of countries around the world for MIT; Vietnam's ability to escape MIT; study the factors affecting the ability to escape MIT of countries around the world and Vietnam 3.2 Research scope Scope of content: The thesis researches the experience of MIT escapees and nonescapees from countries around the world, estimating the factors determining the probability of MIT exit from the group of countries with lower-middle-income in Asia, determining Vietnam's economic status and ability to escape MIT, thereby providing policy implications for Vietnam Although there are many causes and manifestations of MIT, in this thesis, the author focuses on the manifestation of MIT through the indicator of per capita income and economic growth of a country and focuses on the factors that help countries escape MIT successfully Research space: The author researches about MIT in the world, however, to find policy implications for Vietnam to successfully escape MIT, the thesis chooses to focus on MIT research in economies in Asia such as MIT escapees (Korea, Japan, Taiwan, Singapore, Hong Kong), MIT countries in Asia (specifically the Philippines and Indonesia) and the low-income group in Asia (includes Vietnam, India, Cambodia, Laos, Philippines, Myanmar, Mongolia, Bangladesh, Pakistan, Nepal, Indonesia, Sri Lanka, and Bhutan) The choice to focus on Asian economies is because these countries have geographical, cultural, historical, and economic similarities with Vietnam, thus minimizing bias in the results Research results and policy implications for Vietnam will be more appropriate Research duration: The thesis researches the experience of MIT in the world and Asia in the period from 1950 to 2020 The thesis evaluates and estimates the factors affecting the ability to escape MIT in lower-middle-income countries in Asia in the period from 2002 to 2020 The thesis conducted expert interviews to find out the decisive factors for Vietnam's ability to escape MIT and build scenarios for Vietnam's economy from 2021 to 2030, the interview period is March 2021 - June 2021 Research method In this thesis, the author uses a combination of two research methods, qualitative and quantitative 4.1 Qualitative research methods 4.1.1 Research data The data used by the author in qualitative research includes primary and secondary data: Secondary data was collected from websites of World Bank, IMF, Penn World Tables, Transparency.org, General Statistics Office of Vietnam, Ministry of Finance, from 1950 to 2020 Primary data is collected through in-depth interviews with 15 experts with experience working and researching in the field of economics, collecting primary data period is from March 2021 to June 2021 4.1.2 Analytical methods - Comparative meta-analysis method for secondary data with the following purposes: First, analyze the experience of Asian countries towards MIT (find out the factors that help the group of countries Japan, Korea, Taiwan, Singapore, Hong Kong successfully escape MIT; analyzing the factors that cause the Philippines and Indonesia to be trapped in MIT) Second, compare the factors affecting the ability to escape MIT between Vietnam and the MIT escapees and non-escapees, and compare with some specific countries such as Korea, Singapore, Japan, etc ), thereby assessing Vietnam's ability to escape MIT - Analytical method for developing scenarios according to experts' opinions by scoring: In this study, the author used a questionnaire, gathering expert opinions on factors affecting the ability to help Vietnam escape MIT thereby building Vietnamese economic scenarios from 2021 to 2030 Compare the results of the expert interview with the above comparative meta-analysis and the quantitative study to determine the possibility of escaping the MIT of Vietnam 4.2 Quantitative research methods 4.2.1 Research data The data used by the author in the quantitative research is secondary data collected from the websites of the World Bank, IMF, Penn World Tables, Transparency International, and Maddison (2018) for the period from 1950 to 2020 4.2.2 Analysis methods Regression method: The researcher used the regression method for panel data to evaluate and estimate the factors affecting the ability to escape the MIT of lower-income countries in Asia (Vietnam, India, Cambodia, Laos, Philippines, Myanmar, Mongolia, Bangladesh, Pakistan, Nepal, Indonesia, Sri Lanka, and Bhutan) period 2002 - 2020 Model results are the basis for making policy implications in chapter Forecasting method: Based on different concepts of MIT, the author uses the ARIMA model (Box and Jenkin, 1976) to forecast the income and growth of Vietnam to 2030 to determine the possibility of escaping MIT of Vietnam The author also used Eview software to evaluate the stationarity of the data series on the relative income of Vietnam compared with the US, thereby assessing Vietnam's ability to escape MIT New contributions of the thesis 5.1 Theoretical contributions The thesis systematizes the scientific basis (including theoretical and practical basis) about MIT, identifies the list of MIT escapees and non-escapees, the factors that make countries fall into MIT, the factors that help countries escape MIT Based on previous international studies, the thesis has identified methods to assess Vietnam's ability to escape MIT In the thesis, the author has combined many research methods from analysis, synthesis, and comparison, to a quantitative method and expert interview method 5.2 Practical contributions The thesis has built a full methodology to research about MIT that any country with similar characteristics to Vietnam (ie is in the period of lower-middle-income) can be applied to research for their own country, specifically as follows: The thesis generalizes the general and specific factors that cause countries to escape or fall into MIT in Asia; and at the same time apply definitions of MIT, comparison criteria, and gets expert opinions to assess Vietnam's ability to escape MIT The thesis uses a panel data model to identify the determining factors to the ability to escape MIT of Vietnam and other lower-middle-income countries in Asia Finally, from the experience of MIT in the region, the economic situation and the ability to escape MIT in Vietnam, the determining factors for the ability to escape MIT in the group of lower-middle-income countries in Asia, the trend of economic development in the world, the government's economic development orientation, the thesis makes policy implications for Vietnam Thesis structure The thesis is divided into six chapters as follows: Chapter 1: A literature review on the middle-income trap Chapter 2: Theoretical basis of the middle-income trap Chapter 3: Experiences of Asia's Middle-Income Trap Escapees and Non-escapees Chapter 4: Vietnam's economic situation and ability to escape the middle-income trap Chapter 5: Estimating factors affecting the ability to escape the middle-income trap for Vietnam and other lower-middle-income countries in Asia Chapter 6: Policy implications for Vietnam CHAPTER A LITERATURE REVIEW ON MIDDLE-INCOME TRAP 1.1 Different approaches to the concept of MIT concepts in studies Although Garett (2004) does not point out the concept of MIT directly, the author observes a stagnation in the growth of developing countries He argued that MIT occurs when a country is unable to compete with high-income economies (with knowledge-based economies and high-quality institutions) and low-income countries (those with high-quality institutions) The advantage in low-skill jobs at the lowest possible cost) According to Gill and Kharas (2007), countries that fall into MIT grow more slowly than rich or poor countries, these countries are caught between poor countries that have advantages in natural resources, cheap labor costs, and developed countries that have advantages in science and technology Ohno (2009) has a different definition of MIT, arguing that countries fall into MIT because they are stuck in the second phase of industrialization because of insufficient investment in human capital Ohno emphasized that countries with natural resources need to move up the value chain and described countries that fall into MIT as being too dependent on growth strategies based on natural resources or FDI inflows Kharas and Kohli (2011) also point out that MIT countries are "unable to make a timely transition from a resourcedriven growth model with cheap capital and labor to a productivity-driven growth model" Pruchnik et al (2017) propose a framework to assess the likelihood of falling into MIT based on the following factors: unfavorable demographics, low level of economic diversification, inefficient financial markets, inadequate advanced infrastructure, low level of innovation, weak institutions, inefficient labor market What all the descriptive approaches have in common is the initial recognition that the structural shift of the workforce from low-productivity agriculture to more productive manufacturing and services helps a country grow well but then due to lack of a new growth engine, that country fell into MIT These descriptive studies did not have clear criteria for determining a country's MIT acquisition or exit status One of the typical studies on MIT using quantitative methods is that of Eichengreen et al (2013) The authors question whether countries with natural disasters are more likely to experience a growth slowdown than others They conclude that there are two income ranges in which countries are susceptible to slowing growth: one between $10,000 and $11,000 and the other between $15,000 and $16,000 The authors emphasize the importance of accelerating technological progress to avoid slowing growth Based on research by Spence (2012), Felipe et al (2014) identified two levels of income: one in the range of $2,000 to $7,500, and an income range of $7,500 to $11,500 (1990 PPPs) If a country stays in the first income range for longer than 28 years or the second income range for longer than 14 years, the country is classified as MIT Aiyar et al (2013) use the conclusions of the Solow growth model The authors identify and review 123 periods of growth decline since 1960 and find that it is true that the MIT non-escapees have a greater frequency of growth slowdowns than the low-income or advanced countries The author provides several explanatory variables for the slowdown in growth in the Pacific Rim countries, such as poor infrastructure and limited regional integration This evidence is consistent with previous studies that suggest that the slowdown in growth is, in fact, particularly severe in countries with natural disasters Thus, the authors have very different approaches to the concept of MIT, however, the most common point of the studies is that the problem of growth is closely related to MIT 1.2 Factors affecting MIT in the research There are many studies on the factors that affect MIT, but in general, most of the studies focus on the variables controlling growth or growth slowdown, which mainly emphasizes the role of the human capital factor, natural resources, technology, investment capital, institutions, and integration In other words, the factors selected are the determinants of long-term economic growth, which are very important for developing countries to maintain economic growth Table 1.1: Factors affecting MIT Factors affecting MIT Physical capital Human capital Technology Institution Integration Explanation Author The marginal productivity of capital declines, making investment less efficient than in the past and countries fall into MIT The low quality of education makes the workforce lack knowledge and skills – this is the reason countries fall into MIT Kharas & Kohli (2001); Agenor (2012); Warr (2011), Lembaran (2018) Investment in education helps improve the quality of human resources, causing positive effects to promote technology, attract investment capital thereby helping countries to successfully surpass MIT Promote innovation, develop technologies that increase productivity, help countries achieve higher economic growth, and gradually move away from MIT Countries that want to exit MIT need to have a comparative advantage in many exports, especially need to increase the technological content of exports Government policies play an important role when it comes to directing and intervening in the economy, addressing market failures Good policies and institutions help countries surpass MIT Integration helps to expand markets, attract FDI, absorb technology, thereby helping countries have the opportunity to exit MIT Acemoglu et al (2006); Eichengreen et al (2011) Stone and Shepherd (2011); Felipe et al (2012) Bozkurt et al (2014), Agenor and Canuto (2015) Stone & Shepherd (2011), Atalay (2015), Agenor and Canuto (2015), War (2011), Tho (2013), Dinh (2014), Phan The Cong & Pham Minh Uyen (2016) Eichengreen et al (2012), Agenor and Canuto (2012), Felipe et al (2012), Aiyar et al (2013), Nguyen Minh Phong (2014) Felipe et al (2012), Eichengreen et al (2013), Lin and Treichel (2012) Kharas and Kohli (2011) Kanchoochat (2014) Paus (2014) Barrios et al (2010), Tho (2013) Nguyen Minh Phong (2014), Nguyen Quynh Huy (2019), Dinh (2014) Tho (2014) Vo Tri Hao (2016) Source: Author's compilation from studies, 2021 1.3 Research gap There is no unified definition of MIT, each study has a different definition and approach, but quite a few studies are showing that the decline in productivity growth and the slowdown in growth are directly related to MIT In addition, studies have concluded that most of the countries with MIT are located in Asia and Latin America MIT problem is being interested and studied by many scholars in Vietnam, mainly studies using qualitative and descriptive analysis methods Most of the studies show that Vietnam is at risk of MIT, so a comprehensive solution is needed to help Vietnam successfully surpass MIT; studies emphasize the role of changing the growth model through improving labor productivity, developing technology, and integrating the economy Currently, there is no study to synthesize the experience of the group of MIT escapees and non-escapees in Asia based on the national classification of Felip et al (2012), no quantitative research on the factors affecting the ability to escape the MIT of the lowermiddle-income countries in Asia (LMICA) (including Vietnam), there is no research to verify the ability of Vietnam to escape the MIT based on quantitative methods Therefore, the thesis will fill in these gaps by: First, the thesis analyzes and synthesizes the experience of the group of countries escaping and being trapped in Asia, focusing on the differentiating factors between the two groups Second, the thesis uses the concepts of MIT to determine the ability to escape MIT in Vietnam (by qualitative and quantitative methods) The thesis also uses the method of collecting expert opinions and developing scenarios about growth and MIT of Vietnam Comparing the results of the methods, thereby clarifying Vietnam's ability to escape MIT Third, the thesis uses a quantitative model to determine the influence of factors on the ability to escape MIT of Vietnam and other lower-middle-income countries in Asia Fourth, from international experience, Vietnam's economic situation, Vietnam's ability to escape MIT, and identify factors affecting Vietnam's ability to escape MIT, the thesis provides policy implications for Vietnam to have sustainable growth and successfully escape MIT CHAPTER THEORETICAL BASIS OF MIDDLE-INCOME TRAP 2.1 Definitions and identification of the middle-income trap 2.1.1 Definitions of middle income and middle-income trap Within the scope of this thesis, MIT is understood as the phenomenon that economies after reaching the middle-income threshold thanks to resource-based growth and low labor costs, growth is slowed down and stuck in the middle income, it takes a long time (more than 42 years) to move to the high-income level This definition is based on the work of Felipe et al (2012) and Robertson and Ye (2013) Escaping MIT is understood as the fact that a country has fallen into MIT and then escaped the trap (as in the case of Greece with the time to transition from low unemployment to high income is 48 years) or a country does not fall into MIT but go straightly to highincome level (as in the case of Korea, Japan, Singapore, etc) with the transition time from low income to high income less than 40 years) 2.1.2 Identification of middle-income trap There are two main approaches to the MIT: from the absolute and relative thresholds Table 2.1: Absolute and relative approach to the MIT Author Definition Felipe et al (2012) Being in the threshold of middleincome more than: 28 years for the lower-middleincome threshold (needs growth >= 4.7% to escape the trap) 14 years for the upper-middleincome threshold (needs growth >= 3.5% to escape the trap) Relative income to the US is a stationary time series Robertson and Ye (2013) Middle-income range 2000 USD 11.500 USD (GDP per capita PPP 1990) 8%-36% Data Period Maddison (2010) 19502010 PWT 7.1 19502010 Source: Author, 2021 What the two approaches have in common is that most of them focus on Latin American and Asian countries Studies with an absolute approach point out growth slowdown are the reason for MIT The relative approach asserts countries that are trapped in the MIT are the one that fails to catch up with the high-income country Most research on MIT focuses on studying the evolution of per capita income and growth rates of a country over time and studying the factors that influence these two variables Commonly used data sources of World Bank, Penn World Tables, Maddison (2010) 2.2 MIT explanation according to growth theories There are many approaches to MIT and different classifications of countries that fall into MIT, and the causes of MIT are studied in many directions including both descriptive and quantitative But most studies agree that the root cause of a country falling into MIT is related to its growth problems Below, the author uses growth models to explain the MIT phenomenon Table 2.2: Summary of theoretical explanation for MIT Author Solow (1956) Lewis (1954) Ohno (2009) Tho (2013) Eeckhout & Jovanovic (2007) Garrett (2004) Reasons for falling into MIT The country that achieved growth through capital accumulation was nearing or reaching a steady-state, where output and wages per worker did not increase, thus falling into MIT Countries entering middle income lose the low-wage advantage that they had when transitioning from agricultural to industrial economies Middle-income countries have failed to move from the industrial stage to the creative stage Countries fail to improve the quality of labor to move to a higher stage of development Middle-income countries have fewer advantages over high-income and low-income countries in the process of globalization Source: Author, 2021 2.3 Factors affecting escaping MIT 2.3.1 Physical capital In the early stages of economic development, a high rate of investment (especially by the public sector) generates strong benefits for economic growth (Agénor, 2012) A steadily 11 3.2 Middle-income trap non-escapees in Asia Expanding the study of Felipe et al (2012), in the middle-income economies in Asia, it can be confirmed that Indonesia and the Philippines have fallen into MIT According to the definition of Felipe et al (2012), the Philippines became a lower-middle-income country in 1975 with an income of $2033 (GNP PPP 1990) By 2016, according to Maddison’s (2018) data, the Philippines's per capita income was only $3,950, much lower than the highermiddle-income threshold that Felipe et al (2012) set at $7250 Thus, the time the Philippines stays in the lower-middle-income threshold until 2016 is 42 years, it can be confirmed that the Philippines falls into the lower MIT Similarly, Indonesia achieved a lower-middle-income status in 1986 ($2051 PPP 1990), 31 years later in 2016, Indonesia’s income was $6215, so Indonesia also fell into a lower MIT Table 3.2: MIT non-escapees in Asia Economy Inefficient investment Philippines ICOR - Indonesia ICOR - Slow growth in human capital Human capital from 1,81 to 2,68, low investment in education Human capital from 2,17 to 3,69 Lack of innovation and technology TFP growth slowdown Lack investment in R&D TFP growth slowdown Weak institution The political crisis, corruption Corruption, inequality Integration (FDI contribution to GDP is low) FDI contribution to GDP is low (1,22% GDP) FDI contribution to GDP is low (1,21% GDP) Source: Author, 2021 Thus, it is possible to draw the basic reasons why the income of the Philippines and Indonesia has not increased rapidly in the past four decades, including ineffective investment, low human capital growth, TFP growth slowdown, improper investment in technology and innovation, political instability, corrupt institutions, high social inequality and inefficient use of FDI capital All these factors make it impossible for the Philippines and Indonesia to move to high income and be stuck at MIT CHAPTER VIETNAM’S ECONOMIC SITUATION AND ABILITY TO ESCAPE THE MIDDLE-INCOME TRAP 4.1 Vietnam economic overview 4.1.1 Income and growth of Vietnam's economy According to World Bank data, Vietnam's economy from 2002 to 2019 had a growth rate ranging from 4% to 6.6% Specifically, in 2000, Vietnam's economic growth reached 5.6%, then this figure decreased in 2001 and 2002 to nearly 5%, before recovering and continuing to increase continuously until 2007 at 6.6% The period 2008-2010 witnessed a sharp decline in Vietnam's economic growth, with an increase of only 4.4% in 2010 In 2011, 2012, Vietnam's economy recovered and achieved growth of 5.4% and 5.1% respectively, before falling into a growth decline in 2013 From 2014 to 2018, the general trend of the growth rate of Vietnam's economy is increasing, the growth rate in 2018 reached 6.02% Compared with the average growth rate of the world and East Asia - Pacific countries, Vietnam's growth is 12 always significantly higher In 2019, Vietnam's growth reached 6% while the average of East Asia - Pacific countries was 3.2% and the world was 1.4% In 2020, in the context of the Covid-19 epidemic spreading around the world, Vietnam still achieved a GDP growth rate of 2.9% (according to World Bank, 2020) This is an outstanding growth compared to the world (negative growth of 3.6%) The World Bank (2021b) also has an optimistic forecast about Vietnam's economic situation with a 2020 growth rate of 2.8 %; from 2021 will recover to 6.8% 4.1.2 General assessment of Vietnam's economy in the period 2002 - 2020 4.1.2.1 Achievements Firstly, the growth rate is high compared to the world, and the per capita income increases Second, exports maintained a good growth momentum Third, macroeconomic issues such as inflation and public debt are gradually being controlled Fourth, technology has developed strongly, and the digital economy has been gradually formed 4.1.2.2 Limitations Firstly, investment efficiency is still low partly because the economy is in the phase of focusing on investment in infrastructure, including infrastructure in remote and isolated areas, and investing in hunger eradication and poverty reduction, ensuring social security In addition, investment in Vietnam is still scattered, planning work is still limited, public investment projects still have losses and waste Second, the competitiveness is not high Third, the quality of labor is not high, there is a lack of skilled human resources Fourth, the institution still has many problems: government performance, corruption control, private property ownership, etc Fifth, environmental pollution in big cities is getting more and more serious 4.2 Assessing Vietnam's ability to escape MIT 4.2.1 Assess Vietnam's ability to escape MIT by comparing criteria The author analyzes and compares the economic factors of Vietnam and the group of countries that have escaped the trap and cannot escape the trap according to Bulman's study (2014) and compare with MIT escapees in Asia such as Japan, Korea, Singapore, Taiwan, Hong Kong and two economies in the MIT which are Indonesia, and the Philippines From there, determine Vietnam's ability to exit MIT Vietnam's economic growth is relatively positive compared to two countries in the MIT in Asia, namely the Philippines (average growth of 1.79%) and Indonesia (average growth of 3.65%) However, when compared with Asia's MIT escapees, Vietnam had a higher economic growth rate than Hong Kong during the middle-income period (4.93% growth rate) but lower than that of four other economies Therefore, to escape MIT, Vietnam needs to promote further growth Human capital When comparing data on secondary and higher education with successful economies in Asia during the middle-income period including Japan (1960-1965), South Korea (1985), Singapore (1970), Hong Kong (1970), Taiwan (1980), Vietnam currently has high school and university completion rates, as well as the average number of years of secondary and 13 tertiary education higher than the averages of other countries when they were in middle income Technology Using PWT 9.1 data to compare TFP between Vietnam and the group of economies that escaped MIT and got trapped in MIT in Asia showed that: the average TFP in the middleincome period in the Philippines and Indonesia was 0.47 and 0.48, respectively While the Asian exits from MIT have higher average TFP in the middle-income period: Korea (0.5), Japan (0.65), Hong Kong (0.91), Taiwan (0.80) ), Singapore (0.73) Meanwhile, the average TFP of Vietnam in the period 2016-2019 reached 0.45 (Socio-economic situation report 2019, GSO) Thus, comparing the TFP value, it can be seen that the TFP of Vietnam is still relatively lower than that of the exit group from MIT Institution Regarding corruption, although there are no specific statistics on the corruption perception index in countries that successfully escaped MIT in Asia in the period 1960 1990, however, the studies showed that Korea China, Japan, Taiwan, Singapore, and Hong Kong are the countries that focused on controlling corruption and had achieved a lot of success from the very beginning of middle-income years At present, in 2020, among the countries with the best corruption control in the world, Singapore ranks 3rd, Japan and South Korea rank 19th and 33rd respectively, Indonesia (102), Vietnam (104), Philippines (115) (Transparency International, 2021) Thus, in terms of controlling corruption, Vietnam is currently similar to the group of MIT non-escapees and far behind the group of countries that have successfully escaped MIT in Asia Integration An important indicator reflecting the integration of an economy is FDI Among the economies that successfully exited MIT in Asia, Singapore, Taiwan, and Hong Kong are the economies that have made good use of FDI capital to develop FDI not only developed industrial production, created many jobs, but also brought technology to these countries According to World Bank data, in 2019, Vietnam's net inward FDI accounted for 6.1% of GDP, while this figure in Singapore since the 1970s has reached 8-10% and continues to increase in the decades, by 2019 the number already reached 32.17% In the two Asian countries trapped in MIT (the Philippines, and Indonesia), net FDI accounts for about 2% of GDP The descriptive analysis that the author has just conducted shows that Vietnam currently has many similarities with the group of countries successfully escaping MIT (high GDP growth, increasing TFP growth rate, relative years of secondary education, economic restructuring towards industrialization, FDI plays an important role in the economy) However, Vietnam has challenges in improving the quality of human resources, promoting technology, and improving the quality of institutions to maintain its growth momentum in the coming time 4.2.2 Assess Vietnam's ability to escape MIT by type of approach 4.2.2.1 Absolute approach Research by Felipe et al (2012) concludes: A country falls into a lower MIT if it stays in the lower-middle-income level for more than 28 years; A country falls into a higher MIT if it stays in the higher middle-income threshold for more than 14 years In this study, based on Maddison data (2018), Vietnam has become a lower-middle-income economy since 2002 (with an income of $2014), to 2010 Vietnam had been in the lower-middle-income 14 range for years, with an average growth rate of 5.5% in the period 2002-2010, to escape $7250 and exit MIT within 28 years, growth in the period 2011 - 2029 needs to reach 4.3% The thesis uses the ARIMA forecasting method, the results show that Vietnam will have an income of 7285.5 USD in 2027 and not fall into the lower MIT If using the IMF's data on growth, by 2026, Vietnam's national GDP will reach 7454 USD, Vietnam also successfully escape MIT 4.2.2.2 Relative approach In the relative approach, the author uses Robertson and Ye's definitions (2013) to study the case of Vietnam Robertson and Ye (2013) develop a model to determine if a country falls into MIT The two authors argue that the convergence of growth rates in a middleincome period is a necessary condition for determining MIT (using PWT 7.1 data) The necessary condition for a country I to fall into MIT is that the relative income of country i to the reference country must be: (1) constant over time; and (2) that income is within the middle-income level If we take yi,t to be the natural logarithm of the income per capita of country i in year t, and yr,t to be the natural logarithm of the reference national’s income per capita in year t Note that if yr,t, and yi,t contain a common definite trend, then xi,t ≡ Yi,t - yr,t are stationary time series data; xi,t is the relative income of country i to the reference country The authors choose the US as the reference country because, over the past 100 years, the US's economy has always grown steadily around 1.8%/year Therefore, it can be considered that the US has reached a level of technological progress and steady growth A country is considered to have middle income when it has income between 8% and 36% of US income The thesis calculates the relative income series of Vietnam compared to the US, the results show that the relative income series has a positive growth trend and through ADF and Perron tests in Eview software, this series does not stop at level Therefore, Vietnam is likely to escape MIT if the current growth momentum is maintained 4.2.3 Assess Vietnam's ability to escape MIT through expert interviews The author conducted interviews with 15 experts (in which men accounted for 46.7%, women accounted for 53.3%; 13.3% of the interviewees were under the age of 30; 66.7% were from 30 to 45 years old, 20% aged between 45 and 60 years old) 40% of experts work in universities and colleges; 26.67% work at state offices, 13.33% work at research institutes, and 20% work at banks and insurance companies These experts are with 5-25 years of working experience in the economic field with expertise in teaching, research, or policy consulting Regarding MIT, 86.67% of experts have studied or researched about MIT In response to the question of Vietnam's ability to escape MIT, 40% of experts said that Vietnam has a high probability of escaping MIT; 46.67% think that the ability to escape MIT is moderate and 13.33% think that the ability to escape MIT is low Thus, most experts believe that Vietnam has a medium and high ability to escape MIT Evaluation of the uncertainty of factors affecting income and economic growth in Vietnam (in the group of growth factors including TFP, urbanization, labor quality, government spending, capital investment, the proportion of medium and high-tech exports, the capacity to control corruption, income inequality, and economic integration), experts say that the digital transformation process is the factor with the highest uncertainty This means that it is difficult to predict whether the digitization process in Vietnam will take place quickly or slowly in the future In response to the question of factors that will help 15 Vietnam escape MIT in the next 10 years, experts say that labor quality is the factor that has the strongest influence on Vietnam's ability to escape MIT The next strong influential factors are TFP, digital transformation, corruption control, deep economic integration, investment capital, the proportion of medium and high-tech goods in exports, reduction of social inequality, government spending, and urbanization CHAPTER 5: ESTIMATING FACTORS AFFECTING THE ABILITY TO ESCAPE THE MIDDLE-INCOME TRAP FOR VIETNAM AND OTHER LOWER-MIDDLE-INCOME COUNTRIES IN ASIA Vietnam and the above countries in the LMICA group have many similarities in terms of geographical location and level of development, so the estimated evaluation results can be used as one of the bases to make the policy implications for Vietnam in chapter 5.1 Income and growth of Vietnam and other lower-middle-income countries in Asia (LMICA group) from 2002 to 2020 5.1.1 Income and economic growth of the LMICA group Regarding income, in the period 2002 - 2020, Vietnam's relative income o the US increased from 6.47% to 13.6%; the relative income of the LMICA group increased from 7.16% to 12.57% In which, the countries with the largest income growth in this period are Vietnam, Mongolia, Sri Lanka, Indonesia, and Bhutan (with the relative income doubling at the end of the period compared to the beginning of the period) During the research period, the GDP growth rate of Vietnam had been always higher than the average of the LMICA group (except in 2013) The growth fluctuations in Vietnam and the average of the LMICA group are also quite similar In the period 2008 - 2010, the growth rate of Vietnam and the LMICA group was only about 4% per year, but it can be said that Vietnam and the LMICA group were not affected by the global financial crisis as much as the developed countries In 2020, the Covid 19 pandemic caused the economy of the majority of LMICA countries to suffer heavy losses and fall into recession The bright spots in the LMICA group in 2020 are Vietnam and Bangladesh when they still maintained a positive growth rate of 1-2% 5.1.2 Some factors affecting income and growth of LMICA group In this chapter, the thesis focuses on assessing the influence of factors on the income and growth of the LMICA group, those factors include five factors affecting the ability of countries to escape MIT as analyzed in the above chapters (investment capital, human capital, technology, institutions, and integration), incorporating the government spending factor (reflecting government policy, aggregate demand of the economy) and urban population ratio (reflecting the urbanization of the economy) In general, Vietnam and the LMICA group in the period 2002 - 2020 have the advantage of maintaining a high and relatively stable growth momentum compared to the world, increase in investment capital played an important role in growth Vietnam has a higher percentage of internet users, net FDI/GDP, and tertiary enrollment rates than the group on average Government spending and investment in Vietnam are lower than the average of the LMICA group The common weakness of the whole LMICA group (including Vietnam) is that the contribution of high education in the economy is still limited, corruption is widespread, this may be the obstacle for the LMICA group in the process of escaping MIT and becoming developed countries 16 5.2 Quantifying factors affecting the ability of the LMICA group to pass MIT 5.2.1 Research model The author chose to study the factors affecting the ability to escape MIT of the group of lower-middle-income countries in Asia (including 13 countries, namely Vietnam, India, Indonesia, Laos, Cambodia, Myanmar, Nepal, Pakistan, Mongolia, and Bangladesh, Bhutan, Philippines, Sri Lanka) Because Vietnam and other countries in the LMICA group have similar characteristics in terms of region and level of economic development, the results of estimating the factors that determine the ability to escape MIT of the LMICA group can be used as a basis for policy implications to help Vietnam escape MIT Using Solow's economic growth model, the researcher runs econometric models for groups of factors: Capital (K), Labor (L), and Technology (T) Combining some additional factors according to Keynesian aggregate demand models such as government spending (G) and other factors such as institutions, urbanization The function will have the form: Y = F (K, L, T,…) in which the model used for estimation in the thesis is: The regression function represents the relationship between the dependent variable, the relative GDP of a country to the US (RelGDP) The variable RelGDP is a variable that represents the ability of a country to escape MIT in terms of relative MIT approach (in the study of Woo et al (2012), Agenor et al (2012), Bulman et al (2014) ), Robertson and Ye (2013)) The thesis is based on a relative approach when determining MIT based on the argument of previous studies that the development goal of each country is to achieve a high-income level like developed countries (typically the US) The higher the relative income of countries to the US, the smaller the income gap between that country and the US, and the country's ability to escape MIT will be higher Therefore, using a country's relative GDP to the US is appropriate to assess the country's ability to escape the MIT The independent variables in the model include: investment capital (gross capital formation % GDP - gcf); human capital (measured by the indicator of tertiary enrollment (% gross) -edu); technology (measured by individual internet user (% of population - int); integration (measured by foreign direct investment inflows (% of GDP) - fdi); institutions (measured by corruption perception score - cor); government spending (% GDP) - gov; urban population (urb) 𝑅𝑒𝑙𝐺𝐷𝑃𝑖𝑡 = 𝛽1 + 𝛽2 𝑔𝑐𝑓𝑖𝑡 + 𝛽3 𝑒𝑑𝑢𝑖𝑡 + 𝛽4 𝑖𝑛𝑡𝑖𝑡 + 𝛽5 𝑓𝑑𝑖𝑖𝑡 + 𝛽6 𝑐𝑜𝑟𝑖𝑡 + 𝛽7 𝑔𝑜𝑣𝑖𝑡 + 𝛽8 𝑢𝑟𝑏𝑖𝑡 Research period: The author chose the period from 2002 to 2020 because 2002 is the year Vietnam started to achieve middle-income status (as defined by Felipe et al., 2012) and 2020 is the year that the World Bank's data is most updated so far 5.2.2 Estimation method The data used for the study is array data Therefore, the model used to estimate this type of data includes three models: Pooled OLS regression model, fixed effects model (FEM), and random effects model (REM) Specifically, each model is described as follows: Pooled Ordinary Least Square - Pooled OLS: This regression method combines all observations, ignoring space and time factors, and using regular OLS regression However, ignoring the spatial and temporal factors seems to be extremely limited assumptions of the model and Pooled OLS has the disadvantage that when used to estimate panel data, the model may be misrepresented or too tightly bound to the cross-unit Therefore, REM and FEM models are more often used for panel data to solve the above problems 17 Fixed Effects Model (FEM): With the assumption that each unit has distinct characteristics that can affect the explanatory variables, FEM analyzes this correlation between the residuals of each unit, position with the explanatory variable, thereby controlling and separating the effects of the individual characteristics (constant of time) from the regressors so that we can estimate the real effects of the regressors on the dependent variable However, this FEM model does not evaluate the effects of timeconstant variables on the dependent variable Random Effects Model (REM): there is a difference between the units that affect the dependent variable In which, the residual of each unit (uncorrelated with the explanatory variable) is considered as a new explanatory variable The REM model has the feature that the set of observations must be random and has the disadvantage that the assumption about the uncorrelatedness between the unobserved variables and the independent variable is too tight To analyze the factors affecting the relative income of the LMICA group, the process of estimation and model testing is carried out as follows: First, regression REM model, conduct Breusch - Pagan Lagrange Multiplier Test (xttest0) to choose between Pooled OLS and REM with the hypothesis pair: H0: var(vi) = 0, the error of the rough estimate does not include inter-subject deviations H1: var(vi)# 0, the error of the rough estimate includes inter-subject deviations If p_value is small, then H0 is rejected, there is a bias between observations over time, so the REM or FEM model is more suitable Continuing, to choose between FEM and REM, we conduct Hausman Test with a pair of hypotheses: H0: REM model is more efficient H1: FEM model is more efficient If p_value is small, rejecting H0 means FEM is more suitable Finally, for the research results to be reliable, the estimated results must satisfy some assumptions such as no multicollinearity, constant error variance, no autocorrelation, no correlation If the test results detect that the model has one or more of the above defects, we can use the Robust standard error estimate to control or we can use the general least squares (GLS) estimation method with controls for violations such as variable variance and autocorrelation 5.2.3 Statistical description and correlation The statistical description of variables shows that all surveyed countries have statistics on relative GDP (compared to the US), government spending (% of GDP), gross capital formation (% of GDP), urban population (% of the total population), high education (measured by tertiary enrollment to the total number of enrolled students), internet user rates, corruption perception score, and FDI The correlation matrix shows that all variables in the model are positively correlated with relative GDP Urbanization and high education are the two variables with the highest correlation with the dependent variable, 0.551 and 0.613 respectively Gross capital formation, institutions (corruption perception score), and technology (number of internet users) are also highly correlated (>0.4) The two variables with the lowest correlation are government spending (0.38) and FDI (0.179) In addition, the relationship between the independent variables is not too large, so the risk of multicollinearity can be eliminated 18 5.2.4 Results and discussion Table 5.1: Estimation results RE Variables inv edu int cor fdi gov urb Constant 0,04933*** (0,01154) 0,16724*** (0,02171) 0,03529*** (0,00733) 0,01754*** (0,00539) -0,03152* (0,01651) 0,05638 (0,04461) 0,03949 (0,04339) 2,78097* (1,47231) R-square Number of observations 150 Number of nations 13 Group of variables for model selection Xttest0 Chibar2(01) = 481,15 p-value = 0,000 Hausman’s test Chi2(7) = 42,81 p-value = 0,000 The test group violates the hypothesis Variance of error Autocorrelation FE Relgdp 0,04659*** (0,01144) 0,16597*** (0,02213) 0,03422*** (0,00728) 0,01673*** (0,00534) -0,02989* (0,01625) 0,04597 (0,04461) 0,05633 (0,04908) 2,22090 (1,37898) 0,843 150 13 GLS 0,07892*** (0,01330) 0,07200*** (0,01805) 0,05053*** (0,00988) 0,02137** (0,01021) -0,04067* (0,02118) 0,23698*** (0,03393) 0,09023*** (0,01867) -0,65159 (0,47107) 150 13 Chi2(16) = 3,8e+30 P-value = 0,000 F(1,14) = 229,622 P-value = 0,000 Note: values in parentheses are standard errors of regression coefficients, ***,**,* regression coefficients are significant at 1%, 5%, and 10% levels *** p