The relationship between sustainable public debt and economic growth in selected countries and polic.

The relationship between sustainable public debt and economic growth in selected countries and polic.The relationship between sustainable public debt and economic growth in selected countries and polic.The relationship between sustainable public debt and economic growth in selected countries and polic.The relationship between sustainable public debt and economic growth in selected countries and polic.The relationship between sustainable public debt and economic growth in selected countries and polic.The relationship between sustainable public debt and economic growth in selected countries and polic.The relationship between sustainable public debt and economic growth in selected countries and polic.

MINISTRY OF EDUCATION AND TRAINING FOREIGN TRADE UNIVERSITY SUMMARY OF DOCTORAL THESIS THE RELATIONSHIP BETWEEN PUBLIC DEBT SUSTAINABILITY AND ECONOMIC GROWTH IN SELECTED COUNTRIES AND POLICY IMPLICATIONS FOR VIETNAM MAJOR: INTERNATIONAL ECONOMICS CODE: 9310106 PHAM XUAN TRUONG Hanoi - 2022 The thesis is completed at: Foreign Trade University Science instructor: Assoc Prof, Dr Hoang Xuan Binh Assoc Prof, Dr Nguyen Viet Dung 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 Public debt or government debt is a problem that all countries have to face Public debt by itself is not a bad thing The government with its function needs to spend on infrastructure, ensure the education system, health care, maintain national security and social order, which has been increasing during the development of economy, while it cannot raise taxes continously to offset these increases in spending As a result, the government has to borrow However, the rate of public debt growth is higher than ever In Vietnam, a developing country also experienced a similar situation when public debt increased from 54,5% of GDP in 2013 to 61,4% of GDP in 2017, this figure decreased slightly in 2018 at 58.4% Public debt per capita increased respectively from 23 million VND in 2013 to 32 million VND in 2018 (GSO, 2018) It is noteworthy that even the public debt decreased in 2018 at 58,4% of GDP, it was still 10% higher than the average level in the group of developing countries As the scale of public debt is too high, accompanied with the increase in short-term liquidity risks such as reduced tax revenue, high interest rates and foreign debt, public debt will fall into an unsustainable state Economists all agree that unsustainable public debt will create a negative macro-environment for the economy, causing economic growth to decline The public debt crisis in Latin America in the 1980s, the public debt crisis in Europe in the period 2009 2011 followed by an economic recession are clear examples In contrast, a moderate debt-to-GDP ratio, stable interest rates, and reasonable debt structure (domestic/foreign, short-term/long-term) will be the necessary conditions to reach the state of sustainability in public debt Strong public debt sustainability (PDS) will create a strong fiscal foundation to help the economy prosper and reach its potential The Nordic countries are good examples when they have maintained strong PDS with a prosperous economy (Calmfors, 2020) However, in developing countries, governments with a dominant role in the economy will inevitably increase spending to promote growth According to the Keynesian view, this would increase output in the short run In turn, due to increased spending leading to higher debt scale and higher interest rates, public debt will become less sustainable Fast economic growth is therefore accompanied by less sustainable public debt The above fact shows that it is necessary to have an overview study on the relationship between PDS and economic growth of countries in the world and especially by group of countries to come to a unified conclusion on the relationship As a developing country, Vietnam needs to know the relationship in its group of countries in order to make reasonable fiscal consolidation policies in the future In addition, most of the current empirical studies only focus on exploiting the relationship between public debt over GDP ratio, one aspect of PDS and economic growth Meanwhile, PDS is a composite object that includes many components such as public debt structure or short-term debt solvency (liquidity) expressed through indicators for instance foreign debt over total debt ratio, short- term public debt service over total tax revenue, foreign debt over export ratio Quantifying the level of PDS from its components’ value, thereby considering its correlation with economic growth becomes a necessary task in the study of public debt in particular and macroeconomic research in general Stemming from pratical problems and research gap, the author has chosen the topic "The relationship between sustainable public debt and economic growth in selected countries and policy implications for Vietnam" in which the degree of PDS is quantified through index method Research objective missions 2.1 Research objective Identify the relationship between PDS and economic growth in countries in the world, thereby drawing policy implications related to the PDS in Vietnam 2.2 Research missions To achieve the above general objective, the thesis sets out the following misions: - Making research overview and theoretical background on the relationship between PDS and economic growth - Developing an index method to measure the degree of PDS - Determining the quantitative model to evaluate the relationship between PDS and economic growth - Proposing policies and some implementation solutions based on the results from quantitative model and pratical situation to optimally solve the relationship between PDS and economic growth in Vietnam Research subject and scope 3.1 Research subject The relationship between PDS and economic growth based on the data from 151 countries around the world 3.2 Research scope + Space: the thesis examines the relationship between PDS and economic growth of 151 countries and four groups of countries, which are high-income country group (39 countries), upper middle-income country group (39 countries), lower middle-income country group (47 countries) and lowincome countriey group (27 countries) + Time: the thesis examines the relationship between PDS and economic growth in the period 2000 – 2018 since this is the duration that data of public debt of all available countries was fully collected by the World Bank (WB) and the International Monetary Fund (IMF) + Content: the thesis only evaluates the economic relationship between PDS and economic growth That means only financial components of PDS are taken into account, whereas political and geographical components are not considered In addition, the thesis will focus on the quantitative characteristics of the two objects to evaluate the relationship between them in econometric models With PDS, it is the degree of public debt sustainability measured by debt sustainability index and with economic growth, it is the economic growth rate Research questions Research results will answer the following questions: - What is the theoretical background of the relationship between PDS and economic growth? - How to measure the degree of PDS? - What is the practical relationship between PDS and economic growth in overall countries and in group of countries? - What are the solutions that help Vietnam strengthen its PDS in accordance with economic growth? New contributions of the thesis The thesis has theoretical and practical contributions through the construction of a composite index on PDS, thereby determining the relationship between PDS and economic growth in each group of countries classified by income by econometric models, based on that, drawing policy implications for Vietnam, specifically as follows: - The thesis proposes a new method - the index method using max normalization to measure the degree of PDS The new method evaluating most of the important components of PDS can quantify the degree of PDS of countries and groups of countries by income at a given year - Based on the PDS index, the thesis finds out the relationship between it and economic growth rate This is a new finding Subsequent studies related to PDS can use the index to conduct other quantitative models, thereby enriching the practical knowledge of PDS - The thesis uses VAR/VECM models to quantify the relationship between PDS and economic growth, thereby indicating the specificity of the relationship in each group of countries - Based on the relationship between PDS and economic growth in the group of lower middle-income countries, the thesis also draws some policy implications to help Vietnam increase its PDS but still ensure a harmonious relationship with its economic growth Thesis structure Besides introduction and conclusion, the thesis includes chapters, as follows: Chapter 1: Literature review on the relationship between public debt sustainability and economic growth Chapter 2: Theoretical background on the relationship between public debt sustainability and economic growth Chapter 3: Reserach method Chapter 4: Empirical analysis on the relationship between public debt sustainability and economic growth Chapter 5: Policy implications for Vietnam CHAPTER 1: LITERATURE REVIEW ON THE RELATIONSHIP BETWEEN PUBLIC DEBT SUSTAINABILITY AND ECONOMIC GROWTH 1.1 Studies on public debt sustainability Researches on PDS focus on three topics: what is PDS, how to evaluate PDS and its impact on other variables in the economy According to international financial institutions, public debt is considered sustainable when the debt is repaid without major adjustments in the budget, the government does not need to declare default or renegotiate its loan (UNCTAD, 2016; IMF, 2013; ECB, 2012) In Vietnam, according to the Law on Public Debt Management in 2017, public debt is considered sustainable if public debt safety indicators are not violated This is also a view commonly used by policy makers (Debrun, 2017) In general, the higher the scale of public debt is, the more likely it is to fall into an unsustainable state To assess whether public debt is sustainable, academic studies often use the binary tree method of Manasse and Roubini (2005), the IMF's Debt Sustainability Framework (DSF) and the method testing the stationarity of time series data on public debt proposed by Corsetti and Roubini (1991) Regarding the impact of PDS on other macro variables, it can be said that the impact of PDS on economic growth is the topic that many studies focus on and analyze However, due to the lack of a comprehensive measure of PDS, these studies often use one aspect of PDS for instance the size of public debt to GDP or the size of foreign debt to GDP and then consider the impact from the aspect on growth Typical studies evaluating the impact of public debt size on economic growth include Reinhart and Rogoff (2010), Checherita et al (2010), Imbs and Rancierce (2005) In addition, the size of public debt is also considered assumingly representing PDS in relation to other examined factors such as institutions, the development of financial markets and inflation 1.2 Studies on the relationship between public debt sustainability and economic growth 1.2.1 Studies on the impact of economic growth on public debt sustainability Most of the studies looking at the impact of economic growth on PDS are the studies that examine the criteria IRGD (Interest Rate-Growth Differential) Specifically, studies will try to find empirical evidences that as the difference between interest rates and growth decreases, or equivalently, the difference between growth and interest rates increases, the size of public debt will tend to stabilize or decrease in the future, thereby indirectly showing that the country’s public debt becomes more sustainable Abbas et al (2011) used the data of 178 IMF member countries since 2001 to find out how much factors contribute to the change of public debt size The results showed that in the case of 60 countries with high public debt, with the initial debt size originating from more than 60% of GDP, the difference between interest rates and growth contributed the most to the change in the public debt size with 18,1% compared to the initial budget deficit of 8,9% In addition to many studies that have found a one-way effect as above, there are also a few studies that consider the opposite impact from PDS on the difference between interest rates and economic growth Turner and Spinelli (2012) not only showed a unidirectional effect of the interest rate growth differential on the size of public debt, but also indicated the opposite effect in OECD member countries Turner and Spinelli (2012) showed that having a very low interest rate or in other words a small difference between interest rates and growth rates had helped to keep the size of public debt in OECD countries stable during the period 1995 – 2005 1.2.2 Studies on the impacts of public debt sustainability on economic growth The author has not found a study directly showing the relationship between PDS and economic growth due to the lack of a quantitative measure of PDS However, there are many studies on the relationship between an aspect of PDS and economic growth, specifically the component public debt/GDP ratio There are three standpoints about the impact of this indicator on economic growth: the effect is linear (either positive or negative), the effect is not detected and the effect is non-linear (both positive and negative) in which the effect will reverse when the public debt/GDP ratio crosses a certain threshold value In fact, most of the studies found a non-linear relationship Table 1.1: Summary of empirical studies on the impact of public debt on economic growth Nghiên cứu – Kết nghiên cứu Modigliani (1961), Friedman (1988), Warner (1992), Fosu (1996), Kumar & Woo (2010) , Karagol (2012): negative relationship Elmendorf & Mankiw (1999), Gale & Orszag (2003), Linear Baldaci & Kumar (2010): negative relationship in the long relationship run between public Diamond (1965) Eisner (1984) Paul (1992): positive debt relationship and economic Nguyễn & Trần (2014), Phạm (2018): positive relationship growth Abbas & Christensen (2007): positive relationship with conditions Presbitero (2005): negative relationship in high-income country; positive relationship in low-income country Patillo et al (2004): negative relationship in the short-run but reverse in the long-run Non-linear relationship between public debt and economic growth Calderon & Fuentes (2013): negative relationship and the degree depends on institutional quality Reinhar & Rogoff (2010): 44 countries in 200 years, threshold of turning point is 90% GDP Baum et al (2013): 20 European countries in 1990 – 2020, threshold of turning point from 67% to 95% GDP Egert (2013): 44 countries in 200 years, threshold level from 20% to 60% GDP Pattillo et al (2002): 93 developing countries from 1969 to 1998, threshold level from 35 to 40% GDP Clements et al (2003): 55 low-income countries from 1970 to 1999, threshold level (regarding external debt) around 20 – 25% GDP Chang & Chiang (2006): all OECD countries from 1980 to 2010, threshold level is 66,63% GDP Cecchetti et al (2011): 18 OECD countries from 1980 to 2010, threshold level is 85% Padoan et al (2012): 28 OECD countries from 1960 to 2011, threshold level is 90% Cecchetti (2011): 18 OECD countries from 1980 to 2010, threshold level is 85% Caner et al (2010): 101 developed and developing countries in 1980 – 2008, general threshold level is 77% GDP, this threshold for developing countries is 64% GDP Checherita & Rother (2010): 12 European countries in 1970 – 2011, threshold level from 90 to 100% GDP Checherita & Rother (2012): 12 European countries in 1990 – 2008, threshold level from 90 to 100% GDP Greeenidge & et al (2012): Caribbean region, threshold level from 55 to 66% GDP Lê & Thái (2015): developing countries in South-East Asia in 1995 – 2003, threshold level is 68% GDP Caner et al (2010), Kourtellos et al (2013), Teles & Mussolini (2014), Ahlborn & Schweickert (2015), Imbs & Ranciere (2005): threshold levels are different due to different institutions Phạm et al (2020): ASEAN + (China, Japan, South Korea) in 2004 – 2015, there are two threshold levels 29,96% and 72,53% for the region None – relationship between public debt and economic growth Bernheim (1989) Barro (1990) Schlarek (2004) Panizza & Presbitero (2014) Pescatori & cộng (2014) Source: Compilation by author Regarding the research method, because public debt data is usually annual, if only one country's data is used, the number of observations will not be enough (except the case of some developed countries such as the United Kingdom, the United States having traditional and modern statistical system), especially with the existence of new variable such as foreign exchange reserves Therefore, studies often use panel data method covering many countries over a given period to increase the number of observations Regarding the scope of study, a small number of studies use data from various countries with different levels of development and these studies have subsequently received a lot of criticism (Reinhart & Rogoff, 2010) The rest of the studies choose specific group of countries with similar development level such as the study of Clements et al (2003), Caner et al (2010) or geographically close countries such as study of Le and Thai (2015), Checherita and Rother (2010) 1.3 Research gaps Due to the limitation of quantifying PDS to a specific number, the studies only take some common criteria in assessing the sustainability of public debt such as public debt/GDP ratio, external public debt/GDP ratio to examine its relationship with GDP growth, income per capita growth or real GDP In other words, these studies only consider one component in PDS to evaluate the relationship between PDS and economic growth When there is a lack of a method that can quantify the level of PDS, it is difficult for economists to examine the object comprehensively and its relationship with economic growth The thesis will fill the gap by building an index method, a method that can quantify the PDS in order to find out the relationship between PDS and economic growth at the world level as well as at the level of country group divided by income In order to be consistent with the data of public debt taken annually in the period 2000 - 2018, the thesis will use the growth rate calculated based on real GDP year by year The calculation of GDP growth based on the real variable, excluding the effect of inflation, is also consistent with the method of taking the ratio of variables related to PDS, which will be presented in Chapter when both the numerator and denominator are estimated by current price As making the division to get the ratio, it naturally cancels out the effect of inflation from the ratio 2.3 Theoretical background on the relationship between public debt sustainability and economic growth 2.3.1 The impact of economic growth on public debt sustainability Theoretically, classical economists point out that in order to sustain debt or not to default, the most important condition is that the increase in national income must be greater than the amount of interest to be paid (Modigliani, 1961; Diamond, 1965; Barro, 1974; Blanchard, 2019) In other words, the economic growth rate of the country must be greater than the interest rate when it borrows In addition to influencing PDS through the interest rate-growth criterion of classical theory, economic growth also attracts investment for domestic and foreign private sector (Khan & Reihart, 1990) The higher the economic growth rate is, the larger the investment capital from the private sector is, thereby reducing the burden of public budget, especially since the public-private partnership (PPP) model is developed (Muhlenkamp, 2014) This will indirectly cause the government's debt to decrease, as the result, interest rate will decrease Therefore, the government will also be easier to issue long-term debt Economic growth also helps contribute to the development of the country's financial market, thereby creating many tools helping flexibly and effectively handle public debts (Ragot & Pinois, 2019) All of the above factors contribute to an increase in the degree of PDS 2.3.2 The impact of pubic debt sustainability on economic growth Public debt has a relatively complicated relationship with economic growth expressed through many channels of impact with different directions Specifically, the impact of public debt on economic growth is transmitted through the following channels: - Public debt and inflation: high public debt increases inflation, thereby negatively affecting economic growth - Public debt and expectation: public debt also affects expectations With increasing public debt, there are various expectations that can be formed leading to opposite effects on economic growth - Public debt and interest rate: the increase in public debt causes crowding out effect, in which increasing government spending raises interest rates in the capital market, thereby negatively impacting the economic growth - Public debt and tax: increase of public debt normally leads to the increase of tax According to traditional view, tax increase (indirectly enhance PDS) will improve budget balance, thus promoting national saving, reducing interest rate and increasing investment According to Ricardian view, tax increase has no impact on national saving, thus having no clear effect on investment - Public debt and exchange rate: public debt affects economic growth through the exchange rate factor in two directions, respectively short term and long term In the short term, an increase in public debt causes interest rates to rise, leading to an appreciation of the exchange rate or the appreciation of the domestic currency, resulting in a decrease in net exports (Blanchard, 2009) In the long term, the continuous increase in public debt will increase inflation, leading to a devaluation of the domestic currency, which in turn benefits net exports With classification by the nature of the impact, public debt affects growth in two directions: positive impact and negative impact - Positive impact: Firstly, public debt is a resource that helps government to have enough capital to meet the needs of society Secondly, public debt helps the government take advantage of idle financial resources in the private sector Thirdly, if the source of debt is foreign, the public debt funded by foreign sources and international organizations will help the country expand bilateral economic relations, take advantage of idle capital in the foreign country with preferential interest rates (UNCTAD, 2002) - Negative impact: Firstly, high public debt will reduce national saving, leading to investment crowding out Secondly, public debt creates pressure to increase inflation Inflation is not bad if it is at a moderate level, but with the above-mentioned effects, inflation caused by excessive public debt is often high, which is harmful to the economy Thirdly, public debt will cause government increase taxes, thereby causing unnecessary distortions in the economy Fourthly, the overwhelming effect over other necessary spendings of government Increased public debt will increase the associated interest payments Other things being equal, an increase in public debt will lead to an increase in the interest payments annually These increased payments will likely replace or offset government spending on more worthwhile programs such as spending on infrastructure, education, and health Fifthly, there are other negative impacts related to politics and institutions When a country's public debt is high, it is clear that the risk of default will be greater, which will cause the country's overall credit to decline and may spread to the credit of economic entities operating in the country Basically, when public debt is sustainable, the positive effects of public debt on GDP or economic growth will outweigh the negative effects The more unsustainable public debt becomes, the larger the negative impacts will be and the positive impacts will decrease as well Thus, theoretically, we can expect a positive interaction between PDS and economic growth when sustainable public debt has created a favorable macro- environment for economic growth On the contrary, the high economic growth in a certain period of time is enough to create an improvement in the scale and ability to pay public debt, thereby increasing the PDS CHAPTER 3: RESERACH METHOD 3.1 Index method to quantify public debt sustainability According to OECD (2004), the concept of indicators is understood as numbers that may or may not be accompanied with a formula, describing the goal that an object is aiming for A composite index is a composition of various components, each of which reflects an aspect or characteristic of the object According to OECD (2008), there are five main normalization techniques in the index method to quantify a composite object in which the thesis shows that – max normalization is the most suitable to measure the degree of PDS The steps to calculate the public debt sustainability index (DSI) based on – max normalization are as follows: Step 1: Classify country to calculate DSI complying with the latest World Bank country division by income Step 2: Identify components in the composite index Step 3: Calculate the value of each component in the composite index by the formula: Component value = (Threshold value – Current national value)/(Threshold value - The best value in the group of countries) Step 4: Calculating the composite index to reflect PDS: 𝐷𝐷 𝐷 = ∑ 𝐷= 𝐷𝐷𝐷/t in which DSI is the composite index to reflect the degree of the PDS of a country, t is the number of component used to calculate DSI Seven components are selected based on the IMF's DSF and the binary tree method of Manasse and Roubini (2005) including: CV1 - Difference between growth rate and real interest rate (IRGD standard); CV2 - Budget balance; CV3 - Public debt/GDP; CV4 – External public debt/GDP; CV5 – External public debt/Exports; CV6 – External public debt service/Tax revenue; CV7 - External public debt service /Exports Hence, the formula in Step becomes 𝐷𝐷𝐷 = ∑7 𝐷= 𝐷𝐷𝐷/7 In the usual case, when the value of the component index is positive, the component and composite index will have a value between and An index value closer to indicates that the current component of the country is very close to the best value On the contrary, the index closer to indicates that the current component of the country is very close to the threshold value, which, if exceeded, is likely to cause a public debt crisis In the case the current value of the country is greater than the threshold value, the component index will be negative A negative value of the component index indicates a negative contribution of this component to the composite index assessing the degree of PDS If there are many component indices having negative values, the composite index will eventually be negative, indicating a very weak PDS It is likely that the country is going to fall into default 3.2 Mơ hình ước lượng mối quan hệ nợ công bền vững tăng trưởng kinh tế Thesis uses VAR or VECM model to assess the relationship between the degree of PDS measured by DSI and economic growth rate The followings are the steps to estimate the relationship between the level of DSI and economic growth rate (g) through VAR or VECM models Depending on the results of specific steps, VAR or VECM model will be selected Step 1: Check for the stationarity of the variable Step 2: Determine the optimal lag of the variable Step 3: Determine if co-integration correlation exists or not and if exists how many co-integration correlation it has based on optimal time lag Step 4: In case of co-integration correlation confirmed, use VECM model to estimate Step 5: Verify residuals from VECM model Step 6: Use Granger causality test to determine the direction of the relationship Step 7: Use the impulse response function (IRF) and variance decomposition function (VDF) to measure the effect of the causal variable on the outcome variable At step and step 3, if the variable stops immediately without making first difference (step 1) or does not exist co-integration correlation (step 3), then use VAR model to estimate The VAR model is estimated after determining the optimal time lag as follows: Step 1: Estimate the VAR model with the optimal time lag Step 2: Check the stability of the model Step 3: Test Granger causality and adjust the VAR estimation if there is only one-way effect In this situation, the variable that is the cause becomes an exogenous variable (independent variable), the endogenous variable only comprises of the lag variable of the dependent one, then interpret the results obtained Step 4: After interpreting the results, to study the response of the dependent variable to shocks, IRF and VDR are also examined CHAPTER 4: EMPIRICAL ANALYSIS ON THE RELATIONSHIP BETWEEN BETWEEN PUBLIC DEBT SUSTAINABILITY AND ECONOMIC GROWTH 4.1 Descriptive analysis on the relationship between public debt sustainability and economic growth All countries Figure 4.1: Average economic growth and average DSI of all countries in 2000 - 2018 Source: Calculation by author Looking at the above figure, we can see that the positive correlation of the two variables is quite strong Except for three years 2005, 2014, 2015 two variables have opposite changes, rest of the time in the period as the average growth rate of countries increases, the average DSI also increases and vice versa High-income countries group The relationship between DSI and economic growth rate in high-income countries is not clear, while DSI is divided into three distinct periods: stability before the 2008 crisis, decline during the 2008-2009 crisis and gradually recovering after 2009, the economic growth rate in the same period is very unstable, which increased and decreased alternatively before the 2008 crisis and during the recovery period after 2009 Only during the crisis period, the economic growth rate moved in the same direction as DSI Upper middle-income countries group For the group of upper middle-income countries, the relationship between DSI and economic growth rate is clearer than in the overall sample Sometimes the two lines seem to be parallel (in the period 2013 – 2018) Lower middle-income countries group The relationship between DSI and economic growth in the group is divided into two phases In the first period from 2000 to 2007, these two variables have a clearly proportional relationship, as the economic growth rate increased, the DSI increased and vice versa However, since 2007, for most of the time, the two variables have an inverse relationship Specifically, in the period 2011 to 2016 when the growth rate increased, the DSI decreased Especially in 2008 when the growth rate dropped sharply from 6,28 to 4,8, the DSI of the group increased from 0,204 to 0,271 Low income countries group The relationship between DSI and economic growth rate in the group is not as clear as the counterpart in LMI group There are periods where DSI and economic growth rate have a positive relationship (2003 - 2005, 2011 - 2013), there are periods where DSI and economic growth rate have a negative relationship (2001 - 2002, 2006 – 2008, 2016 – 2018) And these phases are intertwined Specifically, in the period 2000 – 2001, the two variables agreed But in 2002 the DSI increased while the economic growth rate decreased, the two variables did not agree The positive relationship returned in the next period from 2003 to 2005 But right after that, when the economic growth rate decreased, the DSI showed the sign of increasing continuously The agreement between the two variables was quite strong in the period 2011 - 2015, but soon after the growth rate showed the sign of improvement as the DSI decreased 4.2 Quantitative analysis on the relationship between public debt sustainability and economic growth 4.2.1 Quantitative model and research hypothesis Quantitative model VAR model to estimate the relationship between DSI and g has the form as follows: DSIt = f(DSIt-1, DSIt-2,… ,DSIt-p, gt-1, gt-2,….,gt-p) + c + εt gt = f(gt-1,gt-2,…,gt-3, DSIt-1, DSIt-2,…, DSIt-p)+c’+ε’t where DSIt- i and gt-i are the value of DSI and g at lag i, c and c’ are intercept coefficient, εt and εt are the white noise of the two equations respectively If two variables are not stationary at level but stationary at any difference and have at least one cointegration relation between them, then VECM model to estimate the relationship between DSI and g will be used and have the form: ∆DSIt = f(∆DSIt-1, ∆DSIt-2,… , ∆DSIt-p, ∆gt-1, ∆gt-2,…., ∆gt-p) + π*ECM + εt ∆gt = f(∆gt-1, ∆gt-2,…, ∆gt-3, ∆DSIt-1, ∆DSIt-2,…, ∆DSIt-p)+ π’*ECM’+ε’t where ∆DSIt-i and ∆gt-i are the differences of DSI and g at lag i, respectively; ECM is the error correction component determined by co-integration analysis; π and π' are respectively the coefficient corresponding to the error correction component; εt and ε’t are respectively the white noise of the two equations Research hyphotheses Based on the research of Reinhar and Rogoff (2010), Baum et al (2013), Egert (2013), Pattillo et al (2002), Ciarli et al (2019), the thesis proposes the following four hypotheses: Hypothesis 1: The more sustainable/unsustainable the public debt in the past is, the higher/lower the current economic growth rate is (H1) Hypothesis 2: The higher/lower the economic growth rate in the past is, the more sustainable/unsustainable the current public debt is (H2) Hypothesis 3: The more sustainable/unsustainable the public debt in the past is, the more sustainable/unsustainable the current one is (H3) Hypothesis 4: The higher/lower the economic growth rate in the past is, the higher/lower the current one is (H4) 4.2.2 Research data and statistical description of data The thesis collects data from IMF and WB to calculate DSI for 151 countries with the most complete information in the period 2000 - 2018 Specifically, 151 countries are divided into groups according to income classification criteria of WB, in which the group of high-income countries (HI) has 39 countries; the group of upper middle-income countries (UMI) has 38 countries; the group of lower middle-income countries (LMI) has 47 countries; the group of low-income countries (LI) has 27 countries Each country has 19 observations Totally with 151 countries there are 19*151 observations, of which HI has 19*39 observations, UMI has 19*38 observations, LMI has 19*47 observations, LI has 19*27 observations Similarly, data on the annual economic growth rate (g) of 151 countries in the period 2000 - 2018 will be obtained from the World Development Indicator data source of WB Table 4.1: Statistical description of g and DSI Overall sample (2869 obs) Mean Median Max Min Std Dev g 3,85 3,84 34,47 -30,2 3,92 DSI 0,36 0,48 0,99 -5,64 0,49 Skewness -0,33 -4,09 HI group (741 obs) UMI group (722 obs) LMI group (893 obs) Nhóm LI (513 obs) g DSI g DSI g 2,47 0,68 3,69 0,48 4,37 2,53 0,69 3,85 0,53 4,34 25,16 0,99 26,11 0,86 34,5 -14,4 0,37 -14,8 -0,83 -14,8 3,08 0,096 3,87 0,24 3,89 DSI 0,24 0,34 0,90 -1,89 0,39 0,11 7,96 -2,13 17,75 -0.53 -0.40 -1,47 0,52 g 5,16 5,56 26,42 -30,2 4,44 DSI -0,06 0,19 0,65 -5,65 -3,3 Source: Calculation by author 4.2.3 Estimation results The thesis estimates the relationship between DSI and economic growth for the overall sample and each group of countries All variables stop at the level, so the model used is the VAR model The final results are summarized in the table below: Table 4.2: Estimation results summary All countries (overall sample) HI group Only Only impact Two-way impact of effect of g on DSI DSI on g x Unexpected results 2,3,4 i) DSI with more lags behind negatively impacts on current DSI ii) x UMI group LMI group LI group Confirmed hypothesis 2,3 x x x 1,4 2,3 past DSI negatively impacts on current g i) g with more lags behind negatively impacts on current g i) past DSI negatively impacts on current g i) DSI with mores lag behind negatively impacts on current DSI Source: Compilation by author In each group, the relationship between economic growth and PDS is different In the HI and LI group, we only find the evidence of the economic growth’s impact on PDS, where as in the group UMI and LMI the only opposite direction which is impact from PDS on economic growth is found This reflects the particular features of each group in the relationship between PDS and economic growth In most cases, a positive effect is found from the past growth rate to the current growth rate (H4) Similarly, the past DSI also has a positive effect on the current DSI (H3) Meanwhile, H1, the most considered hypothesis in theory and empirical studies, is only found in the group UMI This demonstrates that it is unlikely as an increase in PDS has promoted economic growth Some cases even indicate that an increase in DSI at any time will reduce the future economic growth (LMI group) In the remaining groups other than LMI group, an initial improvement in the DSI probably leads to an immediate increase in economic growth, but as time goes on this effect turns negative With the variables with higher lags, the likelihood of a transition from positive to negative effect is greater For example, in the HI group, the longer the economic growth in the past lags behind, the more negative the impact will be on the current economic growth In the LI group, the longer DSI in the past lags behind, the more negative the impact will be also on the current economic growth This implies that there is a possibility that the promotion of economic growth or improvement of DSI carried out by these groups has unsustainable components and/or impatient policy implementation Therefore, in the short term, there may be a positive effect, however, as time flies the policy's limitations have been revealed, the impact gradually turned negative 4.2.4 Result dicussion The above results show that the relationship between economic growth and PDS is specific for each group of countries Depending on the group, the results may vary This is explained by the fact that the interaction between economic growth and PDS is not direct but comes through many intermediary factors as shown in Chapter The results from quantitative analysis are in agreement with the previous qualitative analysis Basically, the trend of the relationship in the overall sample graph and each country group is upward Similarly, the results from the model estimation indicates that most of statistically significant coefficients show the positive effect among the two variables In addition, all samples use VAR estimates, showing the short-term relationship between economic growth and PDS This is consistent with the results of in the research of Devarajan et al (1996) in which PDS is represented by the size of public debt and when the size of public debt increases due to increased government spending, there is only a positive effect in the short term Regarding the tested hypotheses, H1 is only found in UMI group, showing that it is not always necessary that promoting the PDS mechanically by increasing government revenue, reducing domestic and foreign debt will lead to higher growth rates in the future H2 is also only found HI and LI group, showing that not every increase in economic growth would promote the PDS For the self- reinforcing hypotheses of PDS and g (H3 and H4), most of the groups studied show the self-reinforcing nature of PDS and economic growth While H3 (self- reinforcingness of PDS) is found in the overall sample, HI and LI group; H4 (self-reinforcingness of economic growth) is found in the overall smaple, UMI and LMI group A new finding of the thesis is that in the overall sample and the LMI group, the results show that the past DSI has a negative effect on the current economic growth Similar results are found in the study of Lof and Malinen (2014) but using data from 20 developed economies This signals that the improvement of PDS does not necessarily lead to an increase in economic growth in middle and low income countries Another new finding of the thesis is that in the LI group, the longer the DSI lags behind, the more likely effect on the current DSI turns to be negative Similar to the explanation for the impact of economic growth with longer lag on DSI, it is possible that the unreasonable increase in DSI in LI countries would harm the economic growth incentives after a certain period of time, thereby causing the future DSI to decrease A good example is Greece Although it is not a member of the LI group, the reality after the public debt crisis in the period 2011-2015 shows that austerity policies implemented in Greece (increasing taxes, reducing government spending) to strengthen the country's fiscal stance at the time could have helped improve public debt criteria, but after a few years it pushed Greece into a period of persistent deflation, the highest unemployment rate in Europe (25.5% in 2015) (Karanasos et al., 2017) This indirectly increases the size of public debt and the risk of debt repayment in the short term CHAPTER 5: POLICY IMPLICATIONS FOR VIETNAM 5.1 The current state of the relationship between public debt sustainability and economic growth in Vietnam Unlike most of the countries selected for analysis, Vietnam's DSI tends to improve in the period 2000 - 2018 This result is similar to China, India, and Cambodia High economic growth in the range of 6% to 8% helped the period 2000 - 2008 witness a strong increase of DSI However, in 2009 due to the impact of the economic crisis of 2008-2009, Vietnam's DSI decreased by 25,4%, the highest decrease among surveyed countries (including China, USA, UK, India and Cambodia) The reason can be attributed to the relatively high openness of the economy, especially after Vietnam's accession to the WTO in 2007 After that, the DSI recovered above pre-crisis levels in 2011 and reverted to a slight decline like in other countries China, India, Cambodia This shows the unstable post-crisis recovery process of developing countries However, thanks to the speed of economic growth accelerating again in the period 2014 - 2018, its’ DSI was also improved At the end of the period 2018, Vietnam's DSI reached 0,456, equal to that of 2008 2000 2001 Figure 5.1: Economic growth and DSI of Vietnam in the period 2000 - 2018 Source: Calculation by author In addition, as a lower middle-income country, the quantitative results in Chapter shows that DSI with more lags behind has a negative impact on economic growth several years later Based on the graph, we see this most clearly in the period 2000 - 2007 when DSI continuously increased in the early 2000 - 2004 and by 2005, the economic growth rate showed signs of decline just before the global economic crisis 2008-2009 The period 2014 – 2018 also saw DSI was continuously improved, which could potentially lead to a decline in growth in the later period 5.2 Policy implications related to the relationship between public debt sustainability and economic growth for Vietnam 5.2.1 Expand and upgrade the index used in public debt management We can see that DSI is an effective indicator to help the government manage the budget in a scientific way The volatility of the DSI tells the government a lot of important information such as where exactly the country's PDS is (weak, medium or strong), what components of PDS make it change the most, if the budget plan is properly implemented for next year then the DSI will be at what level, whether the predicted value of the DSI has fallen into the danger zone or not In fact, Vietnam has used public debt safety indicators in its 5-year and annual public debt repayment and borrowing plans as well as its 3year public debt management program However, if compared with the criteria to evaluate the level of PDS such as the components in the DSI index, or the criteria used in early warning system for debt crisis of the IMF and EU, the above criteria need to be supplemented Some important indicators must be added as follows: the difference between the growth rate and the real interest rate (IRGD), the ratio of the external public debt to GDP, exports, budget revenue and foreign exchange reserves Some recommendations for the government to implement the solution: Firstly, improve the quality of data collection, promote budget reports to guarantee their adequateness and timeliness Secondly, improve the proficiency of government’s staff assigned to the task of public debt management Thirdly, apply technology in public debt management to improve the efficiency of the data process Fourthly, legislate technical indicators like DSI as the indicators to evaluate the performance of ministries and agencies from central to local government in the activities of budget revenue and expenditure 5.2.2 Sound public debt adjustments in accordance with economic growth Vietnam is the member of LMI group The results for the group show that an increase in DSI will reduce the economic growth later Therefore, Vietnam needs to be very careful when implementing fiscal consolidation or increasing its PDS Obviously, with its current public debt situation, although it is at a safe level but tends to approach the threshold, Vietnam cannot help considering the improvement of its PDS At this time, for Vietnam, the two highest risk factors for PDS are the efficiency of public investment and the external public debt Therefore, in addition to maintaining a favorable indicators such as the size of public debt, the structure of public debt, reasonable solutions to improve PDS of Vietnam should also focus on addressing the two above-mentioned factors Some recommendations for the government to implement the solution: Firstly, improve the efficiency of public investment Secondly, regulate the composition of external public debt reasonably Thirdly, rational use of budget deficit financing tools 5.2.3 Reform the institution of public debt management The consolidation of the legal system and institutional reform in general as well as in public debt in particular will help increase the degree of PDS In addition, based on the compatibility between the public debt management model and the level of development of an economy analyzed in section 5.1, Vietnam also needs to improve the legal and institutional aspects of public debt management when the economy is moving towards the upper middle income level Some recommendations for the government to implement the solution: Firstly, supplement legal documents for more modern public debt financing tools Secondly, it is necessary to establish an independent specialized public debt management agency and gradually shift to manage public debt by the method of asset - debt management in a modern model In addition, it is also possible to establish a National Budget Council with the participation of many ministries and to give appropriate advice to the Prime Minister Thirdly, with the fact that Vietnam's public debt management model gradually shifting to the model of a transitional economy, Ministry of Finance needs to establish more market-based debt financing tools such as interest rate swap bonds, forming national investment funds that can issue convertible bonds or national certificates of deposit Fourthly, the government also needs to define a sound PDS strategy accompanied with the business cycle The appropriate strategy for Vietnam is to increase PDS in favorable economic periods and accept the decline of PDS to create policy space for economic recovery in difficult economic periods In addition, the government can set commitments on budget targets in general and public debt safety indicators in particular using public debt indexes like DSI at the beginning of each fiscal year and use these commitments as a measure to evaluate the effectiveness of the whole administrative system as well as determine the additional income at the end of the year for government’s staff like Singapore's experience CONCLUSION Balancing PDS and economic growth is always a difficult problem for countries to solve The relationship between the two variables can be various depending on the characteristics of each group or even individual countries in different times After studying the relationship between the degree of PDS (through DSI) and the economic growth, the thesis found the following results: The first is about the method to assess PDS There are three traditional methods to assess public debt safety: the public debt ceiling method, the binary tree method and the IMF's DFM However, the above methods, from simple to complex, only mainly assess the safety of public debt through each criterion The more safety criteria there are, the stronger the PDS is, but how the degree of the PDS changes overtime, the above methods cannot solve The thesis shows that the index method with - max normalization will overcome the problem when the degree of PDS is quantified into a specific number year by year Then this quantified result can be used to examine the trend of PDS change of each country or group of countries, answering the question of what factors contribute to this trend as well Furthermore, we can look at the relationship between PDS and other macro variables to have better understanding how they interact in the economy The second is about the relationship between PDS and economic growth The results show that in the overall sample of 151 countries, there exists a causal relationship between DSI and economic growth rate However, in the remaining groups of countries, only one direction of impact has been confirmed Hypothesis and are confirmed the most in all groups of countries showing the self- reinforcing nature of the macro variables, if the past variable has good quality, it will positively affect itself in the future However, there are still exceptions For example, in the HI groups, the past g with more lags behind has a negative effect on the current g or in the LI group, the past DSI with more lags behind has a negative effect on the current DSI All of the specific results above stem from differences in development levels, institutions and even factors such as socio- cultural in each group of countries This suggests that the following studies, if using DSI, need to be conducted on smaller group of countries that share as much in common as possible in oder to have highly practical outcomes The third is about policy implications for Vietnam The change of DSI value in the period 2000 - 2018 is relatively consistent with qualitative analysis on the size and structure of Vietnam's public debt Therefore, Vietnam can consider implementing public debt policies based on technical indicators including DSI To this, Vietnam needs to improve the collection and processing of budget-related data by investing in human resources and technology Vietnam is a country in the LMI group, the results of the model show that increasing DSI through sudden change like cutting public spending will have negative effects on growth Therefore, it is necessary to consider increasing DSI through appropriate measures such as improving the efficiency of public investment, choosing a reasonable form of budget deficit financing Finally, in order to catch up with the development level of the economy, Vietnam needs to reform institutions and build a model of public debt management of a transition country Besides the obtained results, the thesis also has objective limitations in terms of data when the data on the public debt has not been updated to the present time and is only available in the period 2000 - 2018 As a result, the fact that theVAR estimation applied to all cases indicates that there is no longterm relationship between economic growth and PDS Regarding the DSI calculation method, the quality of the index will be better if the author applies the same method to determine the threshold values in each group of countries However, the author's collection of threshold values based on reputable studies along with compliance with the principle high-income countries group having a higher threshold value contribute to reducing this limitation The future research direction proposed by the author is to continue updating the data in the period of 2019 - 2021 to see the impact of the Covid-19 pandemic on the relationship between PDS and economic growth In addition, the author can evaluate the relationship between PDS and other macro variables such as inflation, unemployment, and interest rates with the same research method The author will also continue to monitor the PDS evolution of countries around the world to find out the warning value of DSI for the national case of being default from public debt LIST OF PUBLICATIONS Phạm Xuân Trường, Phương pháp số đánh giá mức độ bền vững nợ công áp dụng cho trường hợp Việt Nam, Tạp chí Kinh tế Dự báo, số 22 tháng 8/2020, trang 16 – 21 Phạm Xuân Trường, Mức độ bền vững nợ công nước khu vực Đông Nam Á: cách tiếp cận theo phương pháp số, Tạp chí Kinh tế Dự báo, số 26 tháng 9/2020, trang 20 – 24 Phạm Xn Trường, Các mơ hình quản lý nợ cơng giới hàm ý sách cho Việt Nam, Chuyên san Nghiên cứu sách quản lý ĐHQG Hà Nội, Vol 37 No 4/2021, https://doi.org/10.25073/25881116/vnupam.4278 Phạm Xuân Trường, Mối quan hệ nợ công bền vững tăng trưởng kinh tế nước khu vực Đơng Nam Á, Tạp chí Nghiên cứu Đông Nam Á, số (250) 2021, trang 12 – 25 (đồng tác giả) Phạm Xuân Trường, Education, Industry 4.0 and Earnings: Evidence from Provincial – Level Data of Vietnam, Journal of Asian Finance, Economics and Business, Vol No (2021) 0675-0684, doi:10.13106/jafeb.2021.vol8.no2.0675 (đồng tác giả) Phạm Xuân Trường, The effect of Covid 19 on Public debt sustainability in selected countries, Kỷ yếu Hội thảo cấp Trường “Socio-economic issues of Vietnam beyond the pandemic”, Trường Đại học Ngoại Thương, 6/2021 (đồng tác giả) Phạm Xn Trường, Phân tích tính bền vững nợ cơng Việt Nam số hàm ý sách, Tạp chí Quản lý Kinh tế Quốc tế, số 139, 8/2021 Phạm Xuân Trường, Bài tập Kinh tế vĩ mô nâng cao, NXB Khoa học Kỹ thuật, 6/2020 (Thành viên biên soạn) ... Hypothesis 2: The higher/lower the economic growth rate in the past is, the more sustainable/ unsustainable the current public debt is (H2) Hypothesis 3: The more sustainable/ unsustainable the public. .. al (2019), the thesis proposes the following four hypotheses: Hypothesis 1: The more sustainable/ unsustainable the public debt in the past is, the higher/lower the current economic growth rate... review on the relationship between public debt sustainability and economic growth Chapter 2: Theoretical background on the relationship between public debt sustainability and economic growth Chapter