Impact of remittances on the households expenditure structure in vietnam

UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS IMPACT OF REMITTANCES ON THE HOUSEHOLD’S EXPENDITURE STRUCTURE IN VIETNAM BY LE THI NHU AN MASTER OF ARTS IN DEVELOPMENT ECONOMICS HO CHI MINH CITY, DECEMBER, 2015 UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS IMPACT OF REMITTANCES ON THE HOUSEHOLD’S EXPENDITURE STRUCTURE IN VIETNAM A thesis submitted in partial fulfilment of the requirements for the degree of MASTER OF ARTS IN DEVELOPMENT ECONOMICS By LE THI NHU AN Academic Supervisor: Dr Nguyen Ngoc Thuy HO CHI MINH CITY, DECEMBER, 2015 ABSTRACT Since the Vietnamese economic reform called “Doi Moi” in 1986, Vietnam has witnessed a blooming in economy as well as the huge remittances flow from the internal and external migrants Many researches choose the remittances behavior for their study Learning from the way people send back their income and the way people in the homeland use this source of financial support is not only support the valuable theory about remittances, but also bring out a practical experiment for one of the developing economy, Vietnam Using the VHLSS 2012 data, this study explore the relationship between remittances and the six categories of household expenditure These categories are food product, non-food (consider for the consumer goods), medical services, education payment, housing facilities payment, durable goods The method using is Tobit regression for the education factor, and OLS for the other Our results show that overall, remittances have statistically significantly impact for food, housing, durable and the medical expenditure Meanwhile, consumption of remittance recipient households is not different from non-recipient households TABLE OF CONTENT CHAPER 1: INTRODUCTION -Page 1.1 The problem statement Page 1.2 Research objective Page 1.3 Thesis structure Page 1.4 Scope of supply - Page CHAPTER 2: LITERATURE REVIEW Page 2.1 Household utility and expenditure theory - Page 2.2 Review of empirical studies Page 10 - International studies -Page 10 - Related studies in Vietnam -Page 14 CHAPTER 3: PATTERNS OF MIGRATION AND REMITTANCE FLOWS IN VIETNAM - Page 16 3.1 Overview of Vietnamese remittances -Page 16 3.2 International remittances -Page 16 3.3 Internal remittances -Page 18 CHAPTER 4: RESEARCH METHODOLOGY - Page 22 4.1 Analytical framework -Page 22 4.2 Econometric model Page 23 4.3 Description of variables Page 24 4.4 Data sources Page 26 CHAPTER 5: EMPIRICAL RESULTS - Page 29 5.1 Statistical result from VHLSS data -Page 29 5.2 Estimation results Page 37 5.3 Interpretation of the results Page 39 CHAPTER 6: CONCLUSION & POLICIES IMPLICATIONS- Page 41 LIST OF TABLES AND FIGURES List of figures Figure 1: Top ten remittances countries Page 18 List of tables: Table 1: Possible Positive and Negative Impacts of Remittances -Page 11 Table 2: Ratio of immigration and migration in 1999 and 2009 Page 20 Table 3: Remittances and household expenditure -Page 22 Table 4: Details of dependent variables Page 29 Table 5: Dependent variables Page 30 Table 6: Independent variables Page 31 Table 7: Number of recipient household Page 33 Table 8: Sources of remittances Page 33 Table 9: Non-parametric analysis Page 34 Table 10: Impact of some household characteristic on the total expenditure Page 35 Table 11: Preliminary regression model of remittances impact on dependent variables -Page 37 Table 12: Regression results Page 38 CHAPTER 1: INTRODUCTION 1.1 THE PROBLEM STATEMENT Nowadays, with the explosion of economic transaction between country and region, the numbers of people go to another place different from the birthplace for earning the living has been dramatically increased This migration human resource flow certainly leads to one of the most important monetary flow- remittance Remittances are the money and goods that called internal remittance if migrants living in the city bring back a part of their income to their families in rural area, and external remittance if migrants living abroad send to their families in the mother countries Whether these funds come from internal or external, they can involve a change in consumption patterns of the migrants' households back home Remittances are one of two sources (along with inflows of direct investment FDI) shown in the balance of international payments as funds transfers (net), and represent for more than 10 percent of gross domestic product According to World Bank statistic represented on April 2015, the total remittances flow in all developing countries over the world was 351 billion USD in 2011, reach 436 billion USD in 2014 During these period (2011-2014), all developing regions except for regions in Europe and central Asia- due to the Russia weak economy status and the depreciation of rupee, have been recorded positive growth on the remittance flow Early this year, 2015, economics have expected the total remittance flow in developing countries will have a moderate increase to 440$ billion as the result of the strengthening US dollar and tightening immigration policy Remittance flows are expected to recover in 2016 to reach $479 billion by 2017, in line with the more positive global economic outlook For both estimation of remittance from developing and developed countries, the total remittances transfer will reach to 586$ billion in 2015 and 636$ billion in 2016, according to the report of World Bank Also, according to this report, five largest destinations for migrant are as follow: United State, Saudi Arabia, Germany, Russia and the United Arab Emirate Remittances nowadays play an important role and increase significantly during recent decades In developing countries, remittances is not only the source of income, it is also has a more important role than foreign direct investment Remittances are associated with the significant development impacts At the macro point, remittance has several impacts on the financial system of country and the economy They are considered one of the sources of national income, foreign currency exchange together with the FDI They have a dominant impact on providing financial support on the remittances recipient Orozco, et.al 2005 have made an empirical study and lead to the conclusion that, remittances have lift people out of the poverty and tend to have more savings than those who not receive remittances One side of remittances impact is for the global financial crisis During the recession, remittances to have positive impact on economy, unlike the capital flows or FDI, tends to fall during the downturn period Another saying, remittances are much more stable than many kinds of financial flow The explanation for this stability is that, remittances are migrant-to-family monetary support flow, targeted to help their family- recipient remittances household have a better life At the household level, remittances lead to a higher income that recipient can consume, save or invest Secondly, they help diversify the sources of income depending on the usage of remittances For the developing countries, they are one of the important sources of income for necessary need such as consumption goods, health care, education, ect and reduce poverty With the up-level households, they can use remittance as a source of capital for small businesses and entrepreneurial activities Thirdly, at the micro level, have important implications and financial stability, especially for the international remittance Especially during hard time or in recession period, migrant tend to send more and more remittances to support their family and friends Transparently, remittances directly help to reduce the poverty, increase the welfare of the recipient household Remittances are also used to settle small business, create job for many people The important question is whether the remittances impact on the structure of expenditure of each recipient household or not? If the answer is yes then remittance promotes or inhibits the expenditure and how it affects to consumption section Studying this impact of remittance flows on recipient households has important implications Understanding the impact of remittance flows to expenditure will help raise awareness about the positive as well as the negative that we can promote appropriate policy 1.2 RESEARCH OBJECTIVES The major objective of this study is to identify the impacts of remittances on the expenditure structure of Vietnamese recipient households The study also aims at offering policy recommendations regarding foreign employment and an effective use of remittances Base on the question, we will focus on the two hypotheses tested by the OLS and Tobit regression - Remittances significant improve the likelihood of spending on the six categories of household expenditure - The impact of remittances on the household spending after adding the control variables 1.3 SCOPE OF STUDY As introduced above, this thesis mainly focus on the household expenditure structure affected by the remittances, which can be increase or decrease, or even unchanged according to the fluctuation of remittances receipt To examine at the household level, we use the VHLSS – Vietnam Household Living Standard Survey conducted by The General Statistics Office (GSO) of Vietnam every two year In the thesis, we will use the most updated VHLSS data published in 2012 The study is narrow to the spending of the household, classified into two groups, receive or not receive the remittances Source of remittance are come from the international migrants and the internal remittances from domestic migrants The expenditure structure is classify to six main categories include food consumption, consumer goods, education, health, housing facilities and durable goods 1.4 STRUCTURE OF THIS THESIS The paper is structured in chapters and each chapter will cover the following content Chapter gives the introduction about the research topic including the overall information of remittances and theirs effects; the research objectives to find out what should be focused on this paper Chapter briefly introduces the migration patterns and remittances in Vietnam, Chapter is some literature review covered the impact of remittances on the household consumption structure from the previous researchers Next, Chapter aims at explaining the methodology of research Chapter5 is the statistical models to discuss about the empirical analysis and present regression results and detailed explanation Finally, chapter gives a conclusion for what has been found and some policies recommended based on these finding CHAPTER LITERATURE REVIEW 2.1 HOUSEHOLD UTILITY AND EXPENDITURE THEORY The methodology of this paper is conducted base on the Engel curves Ernest Engel (1821-1896) firstly introduced in his study published in 1857 for the basic theory between food consumption and the income of household He stated that, the spending on food would decline when the household income increase The relation between household expenditure and household income are mathematized as follow: Ci=fi(X,Y,Z,U) (1) where, Ci: expenditure on category i; X: total expenditure; Z: household characteristic; Y: Income of household; u: unobservable variation It represents the relationship between the household bud get shares changed by the specific types of good to total household expenditure Using this model, the difference in consumption between households with different income can be observed To apply this function to this study, income here can be understood by the remittances definition Based on the Engel curve, several functional forms have been proposed for the remittances variable For example, Deaton and Muellbauer (1980) used the Engel curves to exploit the relationship between remittances and spending; they found when income increases, spending for food decrease, spending for clothing, fuel, lightning fee remain the same, whereas the share of luxury goods increase One of the most famous practical studies using the Engel curve, Working-Leser (Working, 1943, Leser, 1963) represented the model which is used as a foundation econometric model of this study The model describes the linearly relationship between the budget shares of one certain expenditure category to the logarithm of the total household spending: (2) Table 9, most of the remittances are sending to the households living in the rural area, account of 60.78% per total households’ observation One considerable difference between two group, receive or not receive the remittances can be well pointed with the household size With the non-recipient’s household, the household size tends to be larger than the recipient household (have been proved in the table with Pr(|T| > |t|) = 0.0001 < 0.05 (5%) In the preliminary literature review, the motivation for migration and sending remittances back to home is mentioned Result in table can be considered a proof for this conclusion The more dependent individuals in the household, the higher probability of people migrate and send remittance back to home Also, people who living in the rural and have migrants in the family will have a higher probability to receive the remittances The results also show that, t test between the gender and employment has no significant meaning For the preliminary step, we run the regression with the dependent variable is total per capita expenditure and the independent variables are the households’ characteristic Table 10: Impact of some household characteristic on the total expenditure LOG_TOTALEXP hhead_employment hhead_male headmarried hhsize urban migration -0.0301 (-0.90) 0.0519 (1.22) -0.0146 (-0.32) 0.191*** (39.41) 0.531*** (31.92) 0.0347* (2.13) LOG_TOTALEXP (with robust standard error) -0.0301 (-0.93) 0.0519 (1.29) -0.0146 (-0.33) 0.191*** (33.95) 0.531*** (31.90) 0.0347* (2.14) schooling_primary schooling_avarage prob_adult _cons N t statistics in parentheses ="* p|t| 0.000 0.123 0.000 0.000 0.050 0.000 0.000 0.006 0.302 0.034 0.877 0.174 0.717 0.000 6147 170.88 0.0000 0.2659 0.2643 11387 [95% Conf Interval] -.0234851 -.0511953 -.0558952 -.1163552 -.0127191 0208141 0161297 -.0137852 -.0054846 -.0234803 -.0053036 -.0125703 -.0130341 1.673995 -.0071436 0061304 -.0189407 -.1065595 -5.28e-06 0249726 029888 -.0023483 0176857 -.0009377 0062161 0022687 00897 1.777359 Table 2: OLS regression for NON-FOOD variable reg NONFOOD IN_REM EX_REM IN_EXREM LOG_TOTALEXP migration hhsize urban self_business hhead_male hhead_employment > schooling_avarage prob_adult prob_child Source Model Residual Total SS df MS 2.99517797 41.9319718 13 6133 230398306 006837106 44.9271498 6146 007309982 NONFOOD IN_REM EX_REM IN_EXREM LOG_TOTALEXP migration hhsize urban self_business hhead_male hhead_employment schooling_avarage prob_adult prob_child _cons Coef -.0028287 0039428 -.0073217 0265648 -.0073105 -.0029563 0164885 0078318 0030398 -.000199 -.0006416 002485 -.0087237 -.0857853 Std Err .0030265 010617 0068442 0018142 0023547 0007702 0025481 0021182 0042913 004175 0021335 0027482 0040753 0191436 Number of obs = F( 13, 6133) = Prob > F = R-squared = Adj R-squared = Root MSE = t -0.93 0.37 -1.07 14.64 -3.10 -3.84 6.47 3.70 0.71 -0.05 -0.30 0.90 -2.14 -4.48 P>|t| 0.350 0.710 0.285 0.000 0.002 0.000 0.000 0.000 0.479 0.962 0.764 0.366 0.032 0.000 6147 33.70 0.0000 0.0667 0.0647 08269 [95% Conf Interval] -.0087618 -.0168703 -.0207386 0230083 -.0119265 -.0044661 0114933 0036794 -.0053725 -.0083835 -.004824 -.0029025 -.0167127 -.1233135 0031043 0247559 0060953 0301213 -.0026946 -.0014464 0214836 0119842 0114522 0079855 0035408 0078725 -.0007348 -.0482571 Table 3: OLS regression for MEDICAL variable reg MEDICAL IN_REM EX_REM IN_EXREM LOG_TOTALEXP migration hhsize urban self_business hhead_male hhead_employment > schooling_avarage prob_adult prob_child Source Model Residual Total SS df MS Number of obs = F( 13, 6133) = Prob > F = R-squared = Adj R-squared = Root MSE = 1.11703174 13 085925519 39.1597778 6133 006385093 40.2768095 6146 006553337 MEDICAL IN_REM EX_REM IN_EXREM LOG_TOTALEXP migration hhsize urban self_business hhead_male hhead_employment schooling_avarage prob_adult prob_child _cons Coef Std Err .0124004 0082437 0177874 010739 0128092 -.0065559 -.0093923 -.0005428 -.0041822 0041789 -.0035105 -.0004136 0045042 -.0609146 0029248 0102601 006614 0017532 0022755 0007443 0024624 002047 004147 0040346 0020618 0026558 0039383 0185 t 4.24 0.80 2.69 6.13 5.63 -8.81 -3.81 -0.27 -1.01 1.04 -1.70 -0.16 1.14 -3.29 P>|t| 0.000 0.422 0.007 0.000 0.000 0.000 0.000 0.791 0.313 0.300 0.089 0.876 0.253 0.001 6147 13.46 0.0000 0.0277 0.0257 07991 [95% Conf Interval] 0066669 -.0118696 0048215 0073021 0083484 -.008015 -.0142195 -.0045556 -.0123118 -.0037304 -.0075523 -.00562 -.0032162 -.097181 018134 028357 0307532 0141759 0172699 -.0050968 -.0045651 00347 0039473 0120882 0005313 0047928 0122245 -.0246481 Table 4: OLS regression for HOUSING variable reg HOUSING IN_REM EX_REM IN_EXREM LOG_TOTALEXP migration hhsize urban self_business hhead_male hhead_employment > schooling_avarage prob_adult prob_child Source Model Residual Total SS df MS Number of obs = F( 13, 6133) = Prob > F = R-squared = Adj R-squared = Root MSE = 361107286 13 027777484 2.11724636 6133 000345222 2.47835364 6146 000403247 HOUSING IN_REM EX_REM IN_EXREM LOG_TOTALEXP migration hhsize urban self_business hhead_male hhead_employment schooling_avarage prob_adult prob_child _cons Coef .0019492 0077401 0077174 -.0016625 0000484 -.001583 0157718 0001205 00138 -.0012269 0002297 -.0001269 -.0007078 0445605 Std Err .0006801 0023857 0015379 0004077 0005291 0001731 0005726 000476 0009643 0009381 0004794 0006175 0009157 0043017 t 2.87 3.24 5.02 -4.08 0.09 -9.15 27.55 0.25 1.43 -1.31 0.48 -0.21 -0.77 10.36 P>|t| 0.004 0.001 0.000 0.000 0.927 0.000 0.000 0.800 0.152 0.191 0.632 0.837 0.440 0.000 6147 80.46 0.0000 0.1457 0.1439 01858 [95% Conf Interval] 000616 0030632 0047026 -.0024617 -.0009888 -.0019223 0146494 -.0008125 -.0005103 -.0030659 -.0007101 -.0013375 -.0025029 0361277 0032824 0124169 0107323 -.0008633 0010856 -.0012438 0168943 0010536 0032703 0006122 0011695 0010837 0010874 0529933 Table 5: OLS regression for DURABLE variable reg DURABLE IN_REM EX_REM IN_EXREM LOG_TOTALEXP migration hhsize urban self_business hhead_male hhead_employment > schooling_avarage prob_adult prob_child Source Model Residual Total SS df MS Number of obs = F( 13, 6133) = Prob > F = R-squared = Adj R-squared = Root MSE = 4.41057906 13 339275313 103.531814 6133 016881105 107.942393 6146 017563032 DURABL E IN_REM EX_REM IN_EXREM LOG_TOTALEXP migration hhsize urban self_business hhead_male hhead_employment schooling_avarage prob_adult prob_child _cons Coef .011403 0169329 0324956 036729 0045077 -.0068903 0093815 -.0009058 0053383 -.006939 -.002321 0031308 -.0149237 -.2932995 Std Err .0047556 0166827 0107543 0028507 0036999 0012102 0040039 0033283 0067429 0065603 0033524 0043184 0064035 0300807 t 2.40 1.01 3.02 12.88 1.22 -5.69 2.34 -0.27 0.79 -1.06 -0.69 0.72 -2.33 -9.75 P>|t| 0.017 0.310 0.003 0.000 0.223 0.000 0.019 0.786 0.429 0.290 0.489 0.468 0.020 0.000 6147 20.10 0.0000 0.0409 0.0388 12993 [95% Conf Interval] 0020803 -.0157711 0114133 0311407 -.0027455 -.0092628 0015325 -.0074305 -.0078802 -.0197994 -.0088929 -.0053347 -.0274769 -.3522683 Table 6: TOBIT regression for EDUCATION variable 0207258 0496369 0535779 0423174 0117608 -.0045179 0172305 0056189 0185569 0059214 0042509 0115963 -.0023706 -.2343308 ... using this data for this thesis, one of the most disadvantages of using the VHLSS is the lack of information regarding the migration Although the survey has the questionnaire for the migration,... their migration One of the important explanations for the migration explosion in Vietnam is the support from the Government policy As is the case in many developing countries, the migration in. .. detailed information for the remittances and remittances source in this study The information regarding the remittances are designed into two question about the value of international and external remittances