STATE BANK OF VIETNAM MINISTRY OF EDUCATION AND TRAINING BANKING UNIVERSITY OF HO CHI MINH CITY TRAN LONG GIANG Impact of bank credit accessibility to Vietnamese household income Major: Banking and Finance Code: 9340201 Tp Ho Chi Minh - 2019 The dissertation was completed at Banking University of Ho Chi Minh City Supervisor: Assoc.Prof.Dr Hoang Thi Thanh Hang Reviewer 1: Reviewer 2: Reviewer 3: The dissertation will be defended in front of the Council at banking University of Ho Chi Minh City at: hour day month 2019 The dissertation can be found at: Banking University HCMC’s Library National Library, Tp Ho Chi Minh CHAPTER INTRODUCTION 1.1 Problem statement According to the report of General Statistics Office in 2016, the proportion of people living in rural areas of Vietnam accounted for about 68%, equivalent to 61.2 million people Therefore, agriculture and rural areas have been of special interest to the Government in recent years In particular, the regulation of the bank credit market is considered as an important policy of the Government to help reduce poverty in rural areas where agriculture, forestry and fishery activities are essential Specifically, Decree No 55/2015 / ND-CP dated 09/6/2015 on credit policies for agricultural and rural development or Decision No 813 / QD-NHNN dated April 24, 2017 on the loan program to encourage the development of hi-tech agriculture, clean agriculture, etc has been implemented However, rural households have difficulties in accessing credit in general and access to bank credit in particular Theoretically and previous research in the world and in Vietnam related to the impact of access to bank credit on household income shows that there are conflicting results Several studies show that access to credit has a positive effect on household income (Dinh Phi Ho and Dong Duc, 2015; Li et al., 2013) However, other studies suggest that access to credit does not improve household income (Phan Thi Nu, 2013; Takahashi et al., 2010) In this study, the author fills the following research gaps: Firstly, the author chooses the time of study in 2014 and 2016 to eliminate the shock of the economic crisis in 2008 that affects the sustainability of the estimated coefficients This is a new point in the scope of the study time that previous studies have not updated Secondly, the research uses the difference in difference (DID) method to combine pooled regression data (Pooled-OLS) to assess the effect of bank credit access on household income that rarely use in Vietnam (Dinh Phi Ho and Dong Duc, 2015) Thirdly, almost no studies have been systematically conducted in Vietnam to measure the factors that affect access to bank credit, using the control group of households only access to non-bank credit to comprehensively assess other control factors such as household financial and credit characteristics Fourthly, another research gap that the study will fill is the large sample with 598 households in 61 provinces / cities from the Vietnam Household Living Standards Survey Data Set (VHLSS) of 2014 and 2016 were repeated Therefore, the estimation results will be more reliable with smaller samples that the studies use Finally, in this study, the credit characteristics of households are more diversified, taking into account various aspects such as the amount of time they borrowed, unpaid loans, and the cost of obtaining loans, etc Therefore, this study was conducted to identify the factors affecting access to bank credit and the impact of bank credit on the income of rural households in Vietnam 1.2 Objectives and research questions 1.2.1 Resreach objectives Originating from practical and theoretical meanings, the overall objective of this study is to identify the factors affecting access to bank credit and the impact of bank credit on household income From there, two specific goals are set as follows: (i) Identify and measure the factors affecting the access to bank credit of rural households in Vietnam (ii) Measure the impact of bank credit on the income of rural households in Vietnam 1.2.2 Research questions (i) What is the reality of households in 2014 and 2016? bank credit accessiblity of Vietnamese rural (ii) What factors affect the bank credit accessiblity of rural households in Vietnam? (iii) How does bank credit affect the income of rural households in Vietnam? 1.3 Research subject and scope of the study Research subject: Factors affecting to bank credit accessibility in two aspects (i) ability to receive bank loans compared to the control group are households that only access credit non-bank; (ii) total borrowed money from rural households Income of rural households in Vietnam is also a main subject of the research Scope of the study: The study focused on rural Vietnamese households in 2014 and 2016 in 61 provinces / cities The impact of credit accessibility for production activities on household income 1.4 Methods and research data 1.4.1 Methodology To identify and measure the factors affecting to the bank credit accessibility of rural households in Vietnam, the study uses two-step Heckman method (Heckman, 1979) The difference or double difference (DID) method is used to measure the impact of bank credit accessibility on the income of rural households in Vietnam DID method allows to compare the difference between the results of participants (intervention group) and non-participants (control group) program or policy 1.4.2 Research data This study uses secondary data from the Vietnam Household Living Standards Survey Data Set (VHLSS) of 2014 and 2016 The survey scope for this data set is 61 provinces and cities that are actually carried out by the General Statistics Office (Appendix 1) Survey subjects are households, members of households and communes / wards in rural Vietnam 1.5 Contribution of the dissertation Firstly, an overview of the current situation and comparison of credit accessibility and income of rural households in Vietnam in 2014 and 2016 Secondly, the study provides new evidence on the impact of bank credit accessibility on the income of rural households in Vietnam in 2014 and 2016 Thirdly, the research findings complement the scientific justification for other relevant studies under two aspects of the problem: (i) Factors affecting the bank's ability to access bank credit rural households in Vietnam with the control group are households with access to non-bank credit only; (ii) The impact of access to bank credit on the income of rural households in Vietnam Fourthly, the results of the study suggest a number of appropriate policies for relevant organizations to improve access to bank credit, efficient use of capital, and increase income of Rural households in Vietnam 1.6 The structure of the dissertation The content of the thesis is structured into chapters: Chapter "Introduction" In this chapter, the author presents important contents of the topic including research issues, research objectives and questions, subjects and scope of research, research methods and data, and contributions and the structure Chapter "Overview of theory and previous studies" In this chapter, the author reviews the theories and previous studies related to factors affecting access to bank credit and the impact of bank credit on income Therefore, asymmetric information in financial markets, economic development theory, sustainable livelihood theory, the economic basis of microfinance are considered fundamental theories Chapter "Research Methods" The research process, experimental models and selected methods corresponding to each objective will be presented in detail in this chapter Chapter "Research results and discussion" This chapter will present descriptive statistical results, quantitative analysis, and discuss research results to answer the first research question This chapter will also present descriptive statistical results, quantitative analysis, and discuss research results to answer the second research question Chapter "Conclusions and recommendations" The author summarizes the main research results Since then, a number of relevant policy implications are also recommended to improve rural household's access to bank credit in relation to income CHAPTER OVERVIEW OF THEORIES AND PREVIOUS RESEARCH 2.1 Theoretical basis 2.1.1 Theoretical basis on the factors affecting credit accessibility (i) Information asymmetric Asymmetric information or imperfect information was first introduced by Akerlof (1970) that led to adverse selection problems and adverse selection and moral risks hazard) are considered fundamental theories explaining the limited access to credit of households, especially poor households with low income levels (Stiglitz and Weiss, 1981; Williamson, 1987) According to Stiglitz and Weiss (1981), for the credit market, the negative option is also a consequence of asymmetric information which means that potential borrowers who are more at risk are often looking for loans and are selected, in which the distinction between the less risky and the more risky borrower is reflected in the interest rate However, raising interest rates to compensate for the high costs of credit risk can push less risky borrowers out of the market As a result, lenders only lend for high risk projects Asymmetric information also leads to moral or psychological risk, which is the situation where the borrower no longer has the incentive to try or act appropriately as before the transaction occurred by the person Loans bear a portion of the risk Consequently, the borrower cannot repay the debt In general, the decision on how much credit to grant and how much to give depends on the information the lender gets from the borrower As a result, not all borrowers have access to credit (ii) Factors affecting access to credit Although financial development has a broad impact on the economy, credit accessibility of households is necessary for adequate research Credit accessibility in developing countries in general and in Vietnam in particular is improving but there are still many factors preventing households from using the financial system in a comprehensive way Indeed, in countries with underdeveloped financial systems, poor households have less access to credit than wealthy households 2.1.2 Theoretical basis on the effect of access to credit to income (i) Development Economic Theory One of the common methods used to assess the contribution of factors to GDP growth is to use a production function with two basic inputs: capital and labor The increase in output in the economy is due to two main components: (i) the increase of inputs; (ii) the increase in productivity is measured by the Total Factor Productivity (TFP) In addition, Claessens and Feijen (2007) point out that access to financial services is a measure of the development of the financial sector The analysis of the financial, income and millennium goals of this study shows that access to financial services has a positive effect on income Income affects health, education and gender equality and vice versa In addition, higher incomes reduce poverty (Figure 2.1) Figure 2.1 The relationship between finance, income and the millennium goals Health, education and gender equality Development of the financial sector Access to financial serviceschính Income (growth) Poor (income poverty and malnutrition) Source: Claessens Feijen (2007) (ii) Sustainable Livelihood Theory Today, the livelihood approach has been widely used when studying the socio-economic characteristics of households in developing countries Livelihood (Chambers) is defined by Chambers & Conway (1992) as follows: “Livelihoods include capacity, assets (reserves, resources, requirements and access) and the activities required to ensure the Livelihoods: livelihoods are only sustainable when they can cope with and recover from shocks, maintain or improve their capacities and assets, and provide sustainable livelihood opportunities for generations to come next; and to contribute net benefits to other livelihoods at the local or global level, in the short and long term ”(Chambers & Conway, 1992, p.6) (iii) The economic basis of microfinance One of the first laws of economics is the law of diminishing marginal returns to capital 2.2 Review previous studies 2.2.1 Previous studies on factors affecting credit accessibility There are many domestic and foreign studies on the factors affecting access to credit with various methods, stages of research and scope of research However, the research results also not have high consensus 2.2.2 Previous studies on the effect of credit accessibility to income According to Claessens and Feijen (2007), access to financial services in general and access to credit in particular have an important contribution to income, especially for poor households A number of studies are currently being conducted to assess the effect of credit accessibility and other factors on the income of households in many countries around the world and local perspective 2.3 Current situation of bank credit accessibility in the world Based on the Financial Access Survey (FAS) of the IMF (2017), the results show the disproportionate or imbalance in gender in financial access in general and credit accessibility in particular, especially in official financial services This directly affects household income and investment opportunities for everyone, resulting in a negative effect on a country's growth and economic development potential FAS survey data in 2016 provides financial accessibility information from commercial banks of countries including Cabo Verde, Chile, Costa Rica, Malaysia and Poland in the period 2010 - 2016 shows that the percentage of female borrowers accounts for less than 50% In particular, the lowest average rate is Costa Rica and the highest is Poland According to the World Bank's Global Financial Database (GFD) in 2017, half of adults in the world reported having borrowed money in the past year This ratio is higher in developed countries where most of the formal credit is provided by financial institutions or credit cards In contrast, in developing countries, borrowing in the non-bank market (through family and friends) accounts for a significant proportion 2.4 Lessons for Vietnam (i) For banks Firstly, there is greater transparency in the financial performance and market information of banks: there is very little data on market demand for credit accesssibility of rural households as well as information on banks The financial efficiency of banks is widely communicated to rural households This reduces to bank credit accessibility Secondly, sharing credit information can reduce risk and increase credit accessibility: Collecting information from a variety of sources and providing records of individuals and businesses' repayment behavior for a variety of uses Through this activity, asymmetric information is cut down significantly, thus allowing lenders to screen borrower information at a lower cost As a result, lenders can make credit decisions faster and reduce risks, increasing access to credit Experience from countries with better credit information systems often provides more loans to individuals and businesses, even after controlling per capita income and contract enforcement Although the credit information system has many benefits, it is still in its infancy in developing countries In addition, the information sharing among lenders is still weak because they are afraid of information leakage to customers Experience in many countries around the world shows that the coverage of credit information systems is directly proportional to credit access (ii) For households Firstly, households need to invest in education and self-improvement of job skills to effectively use bank credit to increase household income Secondly, proactively accessing information on the credit market to improve access to bank credit Thirdly, actively participate in mass organizations to share information and receive community support in accessing bank credit Fourthly, legalizing collateral to improve access to bank credit Fifthly, improve management capacity in production activities Sixthly, use bank credit for the right purpose (iii) For Government Firstly, promoting financial infrastructure: The banking and financial system is increasingly developing, especially in developing countries Therefore, financial infrastructure needs to be upgraded to keep pace with the development of the market Typically, the application of modern information technology in credit will improve to bank credit accessibility Secondly, promote consumer protection: As more financial service providers enter the market, competition is increasingly fierce even in rural areas With increasingly diversified and complex financial services, consumer protection is increasingly urgent, especially in rural areas when education levels are significantly lower than in urban areas Thirdly, collaborative efforts: Need for closer coordination between central banks, banks and local mass organizations to bring more access to credit for rural households CHAPTER RESEARCH METHODS 3.1 Research process The basic research process consists of steps: Step 1: Research theory and previous studies related to factors affecting credit access and the impact of credit access on income Step 2: Building models and research methods Step 3: Collect and process data Step 4: Results and discussion Step 5: Conclusions and recommendations 3.2 Research method 3.2.1 Estimating the factors affecting the bank credit accessibility of Vietnamese rural households To measure the factors affecting the bank credit accessibility of rural households in Vietnam, the study uses two-step Heckman method (Heckman, 1979) The outstanding advantage of this method is that the estimation coefficients in the regression model are reliable because the use of information from the control group is that households not have access to bank credit (Gujarati, 1995) In addition, the two-step Heckman method not only assesses the ability to access bank credit but also shows why some households borrow more while others borrow less 3.2.2 Estimate the effect of credit access on income To measure the impact of bank credit access on the income of rural households in Vietnam, the study used a difference method or a difference (DID) This is a method commonly used in studies to assess the effectiveness of a program or policy (Khandker et al., 2009) DID method allows to compare the difference between the results of participants (intervention group) and non-participants (control group) program or policy In order to implement this method, it is necessary to identify the intervention and control groups and make an initial investigation, then continue to investigate further for both groups after the implementation of the program or policy Based on that, the DID method calculates the average difference between the intervention and control groups before and after affected by the program or policy This study uses the DID method (Pooled - OLS) to evaluate the effect of bank credit access to household income The regression model has the following form: Yit  0  1Ti  2ti  3 X it  i   it (4) In which: Y is the income of the household; T is the intervention variable, where T is if the intervention group (access to bank credit), equals to if the control group (does not have access to bank credit, only non-bank credit is accessible line); t is a dummy time, equal to if it is the previous time participating in bank credit and if it is the time after the influence of the bank credit program; X is the control variable that can affect household income including personal 11 (2015) Proportion of dependents Dep The percentage of children Dinh Phi Ho and under 16 years old and the East Germany elderly over 60 years old on (2015) the total number of household members Financial characteristics Average income Natural logarithm of household annual income from all sources Unit: Thousand dong Kiplimo et al (2015), Development Economics Research (2010) Personal InF Dummy variable of personal sponsorship (m4dc2_01/m4dc2_06) sponsorship, receiving value of if the household has received money from relatives, zero if not Research on Development Economics (2010), Dinh Phi Ho and East Germany (2015) Saving tthubq Sav (m8c3b) The dummy variable for Dinh Phi Ho and savings, receives a value of East Germany if the household deposits and (2015) if the household does not deposit Credit characteristics Time borrowed LoanTime The time the loan has not been repaid Unit: Year Value of Loanv (m8c9) the loan Value of the Thousand dong loan Unit: The Fee (m8c10) amount of money to pay for a loan The amount of money to pay for a loan Unit: Thousand dong Official LoanInterest interest rate Loan interest rate (monthly) Assogba et al Unit: % (2017), Ololade and Olagunju (2013), Bui Van Trinh and Truong Thi Phuong Thao (2014) The Tar0 (m8c13) purpose of the loan is according to the loan application The dummy variable for the Tran Ai Ket and loan application is a value of Huynh Trung Thoi if the loan for agricultural (2013) activities includes "Other crops", "Livestock", "Forestry", "Fisheries", " Growing rice"; is equal to if the loan is for non-agricultural 12 activities including "Education and training expenses", "Expenses for medical examination and treatment", "Non-agricultural activities", "Buying land", "Buying other assets", “Other loan repayments,” “Build / buy a home”, “Other operating expenses” and “Other” The actual Tar1 (m8c14) purpose of using the loan The dummy variable for the Development actual purpose of the loan, Economics receiving a value of if the Research (2010) loan for agricultural activities includes "Other crops", "Livestock", "Forestry", "Fisheries", "Growing rice"; is equal to if the loan is for non-agricultural activities including "Education and training expenses", "Expenses for medical examination and treatment", "Non-agricultural activities", "Buying land", "Buying other assets", “Other loan repayments,” “Build / buy a home”, “Other operating expenses” and “Other” Property mortgage Col (m8c15) Equal to if there is a Diagne (1999) mortgage; equals if there is no mortgage Amount owed Unpay (m8c12) The amount owed includes both principal and interest Unit: Thousand dong Other factors Natural shock Nasock (m4b1b) Dummy variables such natural disasters epidemics, get a value of affected by these factors zero if not affected as Dinh Phi Ho and and East Germany if (2015) and Member of Agrc (m1ac15a) Farmer Association Get a value of if you are a member of the Farmer's Union, equal to if not in this association Assogba et al (2017), Development Economics Research (2010) Member of Womec (m1ac15b) the Women's Receive a value of if it is a Assogba et Women's Union member, (2017), equal to if not in this Development al 13 Union Party member association Party (m1ac15c) Member of Veterc (m1ac15d) Veterans Association Economics Research (2010) Receive a value of if it is a Assogba et al Party member, if not a Party (2017), member Development Economics Research (2010) Receive a value of if you are a member of the Veterans Association, equal to if not in this association Source: Summary of analysis of the author Assogba et al (2017), Development Economics Research (2010) 14 CHAPTER RESULTS AND DISCUSSION 4.1 Research results and discussion of factors affecting bank credit accessibility of rural households in Vietnam 4.1.1 Descriptive statistical results Statistics show that, out of 598 households accessing credit in 2014, up to 483 households borrowed from bank credit institutions (accounting for 80.77%) and 115 households borrowed from non-bank credit (accounting for 19.23%) In 2016, the proportion of households with bank and non-bank credit were 83.44% and 16.56%, respectively This shows that banking credit institutions including Social Policy Bank, Vietnam Bank for Agriculture and Rural Development, and other commercial banks play an important role in rural Vietnam Compared to 2014, the proportion of bank credit in 2016 accounted for about 2.67% is higher (Figure 4.1) Figure 4.1 Situation of access to credit in rural Vietnam in 2014 and 2016 90.00% 80.00% 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% 83.44% 80.77% 19.23% 2014a 2014b 16.56% 2016a 2016b Notes: a - access to bank credit; b- Access to non-bank credit Source: Authors calculations based on VHLSS 2014 and VHLSS 2016 datasets The statistical results in Table 4.1 indicate that the average amount of credit each rural household borrowed was about VND 48 million in 2014 and VND 55 million in 2016 The largest loan values in 2014 and 2016 respectively is 1.25 billion and billion The average official interest rate per month is 0.87% and 0.70%, respectively for 2014 and 2016 The average duration of previous borrowings in 2014 was nearly double that of 2016 with a loan decrease in 2016 compared to 2014 The average income per capita of the household in 2016 was improved compared to 2014 Therefore, these are the highlights of the efforts of the Government of Vietnam in In recent years, the Government has provided lowinterest (or zero) credits to develop agriculture and rural areas 15 Table 4.1 Descriptive statistics results Variables Symbol Credit accessibility Average income Total income Age of household head Gender of head of household Nation Marital status Educational status of the head of household Household size Proportion of dependents Personal sponsorship Saving Time borrowed Value of the loan Foac thunhapbq Income Age The amount of money to pay for a loan Official interest rate The purpose of the loan is according to the loan application The actual purpose of the loan Property mortgage Amount owed Natural shock Member of Farmer Association Member of the Women's Union Party member Member of Veterans Association Fee Medium 2014 2016 Standard 2014 2016 Minimum 2014 2016 Maximum 2014 2016 0.81 1855.15 87723.37 46.80 0.83 2226.03 99909.87 47.80 0.40 1674.02 81902.95 11.76 0.37 2712.57 89489.91 11.58 224 10447 19 357 10999 21 17159 1029550 89 57146 1371510 91 Sex 0.81 0.79 0.39 0.41 0 1 Eth Sta Edu 0.75 0.85 2.66 0.73 0.84 2.66 0.44 0.35 1.50 0.44 0.36 1.47 0 0 1 1 Size Dep 4.17 0.37 4.04 0.37 1.47 0.25 1.48 0.26 1 10 10 Inf 0.84 0.85 0.37 0.36 0 1 0.02 4.26 47664.28 0.02 2.75 0.12 2.14 97139.68 0.14 2.65 99102.06 890 25 500 17.58333 1250000 20.41667 1000000 Sav Loantime Loanv 34.78 55012.31 33.70 250.65 222.19 0 5000 3000 Loaninterest Tar0n 0.87 0.54 0.70 0.51 1.42 0.50 0.90 0.50 0 0 12.5 10 Tar1n 0.44 0.41 0.50 0.49 1 Col Unpay Nasock Agrc 0.35 46310.68 0.84 0.45 0.38 50707.68 0.84 0.44 0.48 99071.25 0.37 0.50 0.49 92548.86 0.37 0.50 890 0 0 0 1250000 1 1000000 1 Womec 0.11 0.11 0.32 0.32 0 1 Party Veterc 0.07 0.10 0.07 0.09 0.25 0.31 0.26 0.28 0 0 1 1 Source: Authors calculations based on VHLSS 2014 and VHLSS 2016 datasets An important characteristic of households who receive loans is that they tend to be members of an organization such as the Farmers' Union, the Women's Union, etc According to the statistics in Table 4.2, more than 38.63 % of households with a bank credit are members of the Farmers' Union This percentage is higher than the proportion of households with one member participating in the Women's Union, Farmer's Union, Party Member and Veteran's Union in the sample 16 Table 4.2 Characteristics of household association with credit accessibility Characteristics household of the Access to bank credit 2014 Member of Farmer Association Is a member Not a member Access to non-bank credit 2014 Access to bank credit 2016 Access to non-bank credit 2016 38.63% 42.14% 6.52% 12.71% 39.30% 44.15% 4.52% 12.04% 8.19% 72.58% 3.01% 16.22% 9.20% 74.25% 2.17% 14.38% 6.02% 74.75% 0.84% 18.39% 6.52% 76.92% 0.50% 16.05% 8.36% 72.41% 2.01% 17.22% 7.36% 76.09% 1.17% 15.38% Member of the Women's Union Is a member Not a member Party member Is a member Not a member Member of Veterans Association Is a member Not a member Source: Authors calculations based on VHSS 2014 and VHLSS 2016 datasets 4.1.2 Regression results In this section, the empirical analysis analyzes the determinants of the access to bank credit by rural households in Vietnam Firstly, the author looks at households with major loans from bank and non-bank credit institutions, thereby analyzing the decisive characteristics to their ability to access bank credit Then analyze the differences in the size of loans received by households Results from the two-step Heckman model, the study shows factors including average household income, age of household head, ethnicity, marital status, duration of outstanding loans and membership of the Farmers' Union, the head of the household is a Party member who has a significant influence on the ability to bank credit acceessibility In contrast, the evidence does not show factors such as the gender of the head of the household, the educational status of the head of household, household size, dependency ratio, personal funding, savings, natural shocks, Members of the Women's Union and members of the Veterans' Association have an influence on bank credit accessibility 4.2 Research results and discussion on the impact of bank credit accessibility to rural household income in Vietnam 4.2.1 Descriptive statistics results Statistical results in 2014 and 2016 in Table 4.9 show that the average loan amount per rural household is about 51 million VND The largest average loan value is 1.25 billion The average official interest rate per month is 0.79% The average income per capita of rural households in Vietnam is about 2.4 million The statistical values of other variables such as the education status of the head of household, household size, dependency ratio, household financial status such as savings and personal financing are detailed in Table 4.3 17 Table 4.3 Descriptive statistics results Variables Credit access Average income Total income Age of household head Symbol Foac Averincome Income Age Time borrowed Loantime Value of the loan Loanv The amount of money to pay for a Fee loan Official interest rate Loaninterest The purpose of the loan is Tar0n according to the loan application The actual purpose of the loan Tar1n Property mortgage Col Amount owed Unpay Gender of head of household Sex Nation Eth Marital status Sta Educational status of the head of Edu household Household size Proportion of dependents Personal sponsorship Saving Natural shock Member of Farmer Association Member of the Women's Union Party member Member of Veterans Association Size Dep Inf Sav Nasock Agrc Womec Party Veterc Medium 2040.59 2040.59 93816.62 47.3 3.505 51338.3 34.24 0.785 Standard 0.38 2260.61 85960.86 11.68 2.52 98153.56 236.75 Minimum Maximum 290.5 37152.5 10723 1200530 20 90 1.125 19 695 1125000 1.20 0.50 4.5 445 0 2.66 0.49 0.48 95850.66 0.40 0.44 0.36 1.49 4.105 0.37 0.845 0.02 0.84 0.445 0.11 0.07 0.095 1.48 0.25 0.36 0.13 0.37 0.50 0.32 0.25 0.29 0.525 0.425 0.365 48509.18 0.8 0.74 0.845 1 0 0 0 0 4000 11.25 1 1125000 1 10 1 1 1 1 Source: Authors calculations based on VHLSS 2014 and VHLSS 2016 datasets 4.2.2 Regression results Table 4.4 presents the results of estimating the basic DID model of the impact of bank credit accessibility on average household income Based on the variable time (year), household income in 2016 increased compared to 2014 However, when comparing between two types of credit access (access to bank credit and access to non-credit banks), the type of credit involved has a positive or negative effect on the average household income Accordingly, the difference in the basic DID model shows that households accessing bank credit will have a negative impact on future income at the 1% confidence level This result implies that household credit participation does not improve the income of rural Vietnamese households 18 Table 4.4 Estimated results of basic DID model Variables Symbol Time year Control Treatment Differences in differences yeartreat Coefficient _cons Regression coefficient 0.3503461*** (0.0862014) -0.0210939 (0.0667346) -0.1898691*** (0.0966324) 7.290678*** (0.0583666) Source: Authors calculations based on VHSS 2014 and VHLSS 2016 datasets In order to control other factors that may affect the average income of a household in addition to accessing credit such as demographic characteristics, credit characteristics, studies combine DID model with regression Pooled-OLS Detailed results are presented in Table 4.5 Table 4.5 Estimated results of the extended DID model Variables Symbol Time Year Control Treatment Differences in differences Yeartreat Time borrowed Loantime Value of the loan Loanv The amount of money to pay for a loan Fee Official interest rate Loaninterest The purpose of the loan is according to the loan application Tar0n The actual purpose of the loan Tar1n Property mortgage Col Gender of head of household Sex Nation Eth Marital status Sta Educational status of the head of household Edu Household size Size Proportion of dependents Dep Personal sponsorship Inf Saving Sav Natural shock Nasock Regression coefficient 0.2642*** (0.0772241) 0.0644741 (0.0555328) -0.1386955* (0.0839226) -0.0098789 (0.0076798) 1.79e-06*** (2.89e-07) -4.83e-06 (0.0000731) 0.0054358 (0.0112465) -0.1453186*** (0.0560746) 0.0880791* (0.0562671) 0.1803386*** (0.0385683) -0.0003794 (0.0657022) 0.4183306*** (0.0411443) -0.0218745 (0.0617351) -0.0146438 (0.0114199) -0.0736599*** (0.0117567) -0.4836849*** (0.06376) 0.1088563** (0.0469704) 0.3457697*** (0.0898599) -0.0486025 19 Member of Farmer Association Agrc Member of the Women's Union Womec Party member Party Member of Veterans Association Veterc Coefficient _Cons R2 (0.0485788) -0.0252865 (0.0339588) 0.0313038 (0.0707495) 0.440092*** (0.0627473) -0.0277729 (0.0532742) 7.303819*** (0.1155031) 0.3792 Note: (*), (**) and (***) are statistically significant at 10%, 5% and 1% respectively Source: Authors calculations based on VHSS 2014 and VHLSS 2016 datasets Empirical research results confirm that access to bank credit will negatively affect the average household income, while enhancing the role of non-bank credit in improving the income of households Participation in non-bank credit reduces household income per capita to 13.87% with a statistical significance of 10% This result supports the judgment of Phan Thi Nu (2013) and Li et al (2013) but contrary to Dinh Phi Ho and Dong Duc (2015), Nguyen Kim Anh et al (2011) and Economic Research Development (2010) This may be explained by the credit incentive mechanism for poor households of banks and government towards rural areas In other words, poor households are easier to borrow, in line with the poverty reduction policy of the local government and the Vietnamese government to develop economy, improve income This conclusion is supported by the finding of the variable Tar0n, which shows that the applications for loan purposes for agriculture, forestry and fishery activities are often associated with low income so it is easy to get loans However, poor households often have poor financial management skills and production capacity, so a large amount of debt negatively affects income Especially for households using loans for non-production purposes This explanation received support through the variables Tar1n and Tar0n that the purpose of the loan on the loan application is quite different from the actual purpose of the loan In other words, many poor households have changed their purpose of using loans under the loan application from production activities to non-production activities and vice versa However, in reality, households using bank loans for production purposes will improve income Specifically: Firstly, households with large amounts of credit loans often have negligible higher incomes but have a statistical significance of 1% Secondly, the purpose of a loan under the loan application for agriculture forestry - fishery activities also reduces the average income of the household compared to households using the purpose of borrowing for other activities: 14.53% with the Statistical significance is 1% Thirdly, the actual purpose of using a loan for agriculture - forestry - fishery activities has a positive effect on income at 8.8% with a statistical significance of 10% Fourthly, mortgage of assets and income of households has a positive relationship This means that households with collateral have a higher average income of 18% at 1% 20 In addition to credit accessibility and credit characteristics of household members, other factors such as ethnicity, household size, dependency ratio, personal financing, savings, and head household are Party members have significant influence on the average household income In particular, ethnicity, personal financing, savings and the head of a household member have a positive influence on the average household income In contrast, the size of the household and the percentage of dependents have a negative effect on the household's income Specifically: Firstly, a Kinh household has a positive effect on the average household income The evidence shows that the Kinh has an income increased by about 41.83% compared to other ethnic groups with a statistical significance of 1% Secondly, household size and average household income are inversely related This implies that the more members the household has, the lower the income, at 43.37% of the average income (the 1% significance level) Thirdly, the higher the proportion of households with dependents, the lower the average income This is understandable because the dependents not make income but still have to spend Fourthly, households with personal finance and bank savings books often have a higher average income of about 10.88% and 34.58%, respectively, with statistical significance of 5% and 1% respectively Finally, the head of the household is a member of the Party - the elites and the status should often have a higher income than the non-Party head, with a 44% increase in income (significant level of 1%) 21 CHAPTER CONCLUSION AND RECOMMENDATION 5.1 Conclusion The study was conducted to identify factors affecting bank credit accessibility and the impact of bank credit on the income of rural households in Vietnam To measure the factors affecting the access to bank credit of rural households in Vietnam, the study uses two-step Heckman method (Heckman, 1979) In addition, the two-step Heckman method not only assesses the ability to access bank credit but also shows why some households borrow more while others borrow less In particular, the ability of households to access bank credit is assessed through two criteria: (i) the ability to receive bank credit with the control group of households accessing non-bank credit ; (ii) total amount borrowed from the farmer To measure the impact of bank credit access on the income of rural households in Vietnam, the study used a difference method or a difference (DID) DID method allows to compare the difference between the results of participants (intervention group) and non-participants (control group) program or policy In order to implement this method, it is necessary to identify the intervention and control groups and make an initial investigation, then continue to investigate further for both groups after the implementation of the program or policy Based on that, the DID method calculates the average difference between the intervention and control groups before and after affected by the program or policy The dependent variable is the household income that is represented as the average income Control variables that can affect household income include intervention variables, time dummy variables, personal characteristics (age, gender, ethnicity, marital status, education status) , demographic characteristics (household size, dependency ratio), household financial characteristics (personal financing, savings, value of unpaid loans, loan duration of loans payment) characteristics of the credit (official interest rate, loan term, total value of the collateral, guarantor, amount payable for the loan, purpose of the loan, actual use of the loan, mortgage assets), membership of associations (Farmer's Union, Women's Union, head of the Party, Veteran's Association), natural shock ên Firstly, the statistics show that, out of 598 households accessing credit in 2014, 483 households borrowed from bank credit institutions (accounting for 80.77%) and 115 households borrowed non-bank loans (accounting for 19.23%) In 2016, the proportion of households with bank and non-bank credit were 83.44% and 16.56%, respectively This shows that banking credit institutions including Social Policy Bank, Vietnam Bank for Agriculture and Rural Development, and other commercial banks play an important role in rural Vietnam Compared to 2014, the proportion of bank credit in 2016 accounted for about 2.67% is higher The average amount of credit capital each rural household borrowed is about VND 51 million The largest average loan value is 1.25 billion The average official interest rate per month is 0.79% The average income per capita of rural households in Vietnam is about 2.4 million The statistical values of other variables such as the education status of the head household, household size, dependency ratio, household financial status such as savings, personal finance, etc are also shown in detail 22 Secondly, the estimation results show that among the factors influence bank credit accessiblity considered in the first step of the Heckman model, there are six factors that significantly influence and have statistical significance to bank credit accessibility These factors include average household income, age of household head, ethnicity, marital status, loan period and membership of the Farmers' Union The Heckman model estimation results in the second step of the ability of rural households in Vietnam to show that there are seven factors that affect the value of loans received from bank credit institutions including: average income of household, age of household head, ethnicity, marital status, loan period, membership of Farmer's Union and Party members Thirdly, the findings from the estimation of the basic DID model on the impact of bank credit accessibility on average household income show that household income in 2016 increased compared to 2014 However, when comparing between the two types of credit access (access to bank credit and access to non-bank credit), the type of credit participation has a positive or negative effect on average income household Accordingly, the difference in value in the basic DID model shows that households accessing bank credit will have a negative impact on future income at the 1% confidence level This result implies that household credit participation does not improve the income of rural Vietnamese households In addition to credit, other factors such as ethnicity, household size, dependency ratio, personal financing, savings, loan value, loan purpose according to the loan application, actual purpose of use, mortgaging assets, the head of the Party is also a significant influence on the average household income In particular, ethnicity, personal financing, savings, loan value, loan purposes under loan applications, mortgage of properties and household heads who are Party members have a positive influence on average households’ income In contrast, the size of the household, the percentage of dependents and the actual purpose of the loan have a negative effect on the household's income 5.2 Recommendations 5.2.1 For banks (i) Complete the lending process and procedures for rural households Credit accessibility generally improves the incomes of rural Vietnamese households However, the effect of income improvement is different between the group approaching the bank credit and the group approaching the informal credit An interesting finding in this study is that participating in bank credit improves income less than participating in non-bank credit Therefore, the formal financial market needs to improve efficiency such as screening and improving the quality of credit ratings for formal financial institutions After lending, formal financial institutions need to control the use of capital by households, effectively reducing the use of loans against their original purposes (ii) Deploy loans for production households through mass organizations Banks should cooperate with mass organizations in lending activities because of the reduction in the cost of human resources that are often lacking in rural areas (credit officers) for disbursement and monitoring Furthermore, mass organizations are aware of households and can provide additional information to banks in loan 23 assessment Lending through mass organizations also contributes the knowledge and skills of bank officers in taking care of borrowers, closely monitoring the status of production and business of farmers as well as propagandizing policies, credit information for farm households iii) Promote the image and brand of banks In order to improve bank credit accessibility, banks need to promote communication activities of the image and brand for rural households to identify Moreover, banks should proactively approach customers with marketing policies, promote credit programs, loan support, and credit packages to the countryside (iv) Improve the qualifications of bank officers Improving the qualifications, knowledge and experience of bank officers in the field of rural lending Bank officers must regularly contact with customers who are agriculture-forestry-fishery production households to solve problems related to credit process, disbursement progress, administrative procedures to accelerate bank credit accessibility In addition, bank officers also need to support agricultural, forestry and fishery production households in preparing business plans, accounting for production costs and ensuring debt repayment progress (v) Advice to agriculture-forestry-fishery production households In order to effectively use credit capital, banks need to be knowledgeable and have appropriate advisory measures for each customer In addition, banks need to regularly restructure loan groups for agriculture-forestry-fishery production and closely follow the Government's support programs and policies to suit the local situation 5.2.2 For rural households (i) Improve the awareness of access to credit information In order to improve the access to credit of rural households in Vietnam, credit information plays an important role In reality, the access to credit information of rural households is limited The main reason is due to low education level, people's living standard, and equal rights is still low that compared to urban areas Therefore, the mainstream information channels need to be widely disseminated to improve the access of bank credit to rural households Local authorities at all levels should establish a department equipped with sufficient resources and specialized skills to update credit information fully, accurately and promptly about the credit market to the locals The provision of information to the people must be diversified in various forms such as local radio, posted at agencies (such as People committee’s offices, cultural houses, etc.) (ii) Encourage households to join mass organizations Joining mass organizations brings many benefits to rural households, especially access to bank credit Most rural households face many barriers in accessing credit such as low educational level, low income, high number of dependents, limited time for social activities due to subsistence activities, geographical distance, negatively affecting access to credit Participation in local mass organizations such as the Women's Union, Farmer's Union, Veteran's 24 Association, etc…will help people to have more opportunities to share information and create close relationships among Households then support each other for economic development The mass organizations are also a channel connecting with banks, state agencies, so updating information fully and promptly on the credit market (iii) Encourage households to complete the land use rights certificates Most rural households have the largest land assets, so the land use right certificate is attached to the land as collateral for bank loans However, the number of rural households with land use certificates is still low compared to urban households Therefore, the Government should have policies to encourage households, especially in rural areas, to make land use rights certificates, creating favorable conditions for households to access credit Bank (iv) Enhancing management capacity in production activities In order to improve the management capacity in production activities of households, local authorities need to organize training on management methods such as cost accounting, calculation, allocation of investment capital and loans Production households need to determine their family's capital needs on that basis to calculate the amount of money to borrow, avoiding the case of borrowing beyond the household's financial capacity Making a clear plan of production and business activities, a plan to repay debts in the future of the family is very necessary (v) Use credit for the right purpose The research results show that the improper use of capital has a negative effect on the income of production households, especially production loans for consumption Therefore, households need to carefully consider the purpose of borrowing before borrowing Expenditures of rural households need to be carefully planned and managed to limit cases of misuse of production purposes, causing shortage of capital for production (vi) Production linkages in agriculture, forestry and fisheries The linkage of production in agriculture, forestry and fishery is known as a development trend of modern agriculture In the current period of international economic integration, the basic requirements in competitive agro-forestry and fishery production include products that must be clean, delicious, large quantities, low-priced and have market strategies good The foundation for achieving these requirements comes from the planning, production location, good seed supply, the linkage and support between farmers - businesses - the state - scientists (or called "Linking the four houses") 5.2.3 For the Government (i) Promote the implementation of credit policies for agricultural and rural development Although there has been an improvement in the proportion of bank credit over the years, the non-bank credit market is still significant Therefore, on the one hand, there should be policies to encourage formal financial institutions to participate in rural areas On the other hand, the Government needs to create conditions for the 25 Bank for Agriculture and Rural Development, the Bank for Social Policies and other commercial banks to continue expanding their network to remote areas (ii) Complete the procedures for issuing land use certificates The Government has implemented long-term stable land allocation to rural households, while speeding up the process of issuing land use right certificates so that farmers have enough assets to secure bank mortgages This is one of the main barriers to access to bank credit by rural households today (iii) Restricting other forms of access to credit When the supply-demand balance is balanced, the credit market will be stable In order to boost bank credit, local authorities need to widely disseminate to the people about the operation of banks with the forms of loans, support packages for rural agriculture of the Government Banks must be the main channel of capital supply in the rural market In addition, the Government needs to minimize unhealthy forms of credit, especially the black credit market, and current usury lending through appropriate regulatory policies (iv) There is a policy to support linkage production model in rural areas The Government encourages and creates an appropriate mechanism to link production and processing in cross-linking between farmers and farmers, as well as between enterprises and farmer households In particular, the enterprise is responsible for supporting inputs for farmers, establishing raw material production areas, directly producing and consuming agricultural products, supplying seeds, materials and techniques to support farmers in commodity production chemical products, and consumption of outputs for the population (this is the main role) Farmers produce goods at the request of the enterprises, according to the technical process and sell products to enterprises The scientist studies new breeds of animals and plants of high productivity and good quality suitable to the ecological conditions of each region and region; high-tech cultivation technology process, post-harvest preservation processing technology; bringing machines and tools of production solutions suitable to each object and each production condition to raise labor productivity, lower production costs and create products up to national, domestic and regional standards (v) Improve the quality of agricultural extension work In order for the agriculture, forestry and fishery production activities to be effective, the agricultural extension activities of the localities should be focused Agricultural extension measures can be considered such as organizing extension training courses with appropriate forms and attracting more people to participate, building a typical production model for households to visit and learn from, providing professional documents for households to refer to in production ... accessiblity of Vietnamese rural (ii) What factors affect the bank credit accessiblity of rural households in Vietnam? (iii) How does bank credit affect the income of rural households in Vietnam? 1.3 Research... Average income tthubq Natural logarithm of household annual income from all sources Unit: Thousand dong access to bank credit Foac (m8c7) Equal if borrowed from Social Policy Bank and Vietnam Bank... on household income that rarely use in Vietnam (Dinh Phi Ho and Dong Duc, 2015) Thirdly, almost no studies have been systematically conducted in Vietnam to measure the factors that affect access