Demand analysis of televisions owership in vietnam

University of Economics Institute of Social Studies Ho Chi Minh City The Hague Vietnam The Netherlands VIETNAM- THE NETHERLANDS PROGRAMME FOR MASTER OF ARTS IN DEVELOPMENT ECONOMICS DEMAND ANALYSIS OF TELEVISIONS OWNERSHIP IN VIETNAM A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMEN1S FOR THE DEGREE OF MASTER OF ARTS IN DEVELOPMENT ECONOMICS HO CHI MINH CITY, SEPTEMBER 2003 University of Economics Ho Chi Minh City Institute of Social Studies The Hague Vietnam The Netherlands VIETNAM- THE NETHERLANDS PROGRAMME FOR MASTER OF ARTS IN DEVELOPMENT ECONOMICS DEMAND ANALYSIS OF TELEVISIONS OWNERSHIP IN VIETNAM A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS IDR THE DEGREE OF MASTER OF ARTS IN DEVELOPMENT ECONOMICS Student: Supervisor: HO CHI MINH CITY, SEPTEMBER2003 LE QUOC THANH NGUYEN HUU LOC CERTIFICATION "I certifY that the substance of this thesis has not yet been submitted for any degree and is not being currently submitted for any other degrees I certify that to the best my knowledge, any help received in preparing this thesis, and sources used, have been acknowledged in this thesis" Le Qu6c Thanh September 8, 2003 i ACKNOWLEDGEMENT Today, my thesis has been completed for official submission In order to complete the thesis; I have duly received many enthusiasm supports from many people I, therefore, would like to be thankful to all of them, especially the following people Firstly, I would like to express my thanks to my supervisor, Mr Nguyen Huu Loc who has advised me a lot of valuable recommeniation to improve the thesis Secondly, the special thanks shall be devoted to Dr Haroon Akram Lodhi, Project Leader of the Vietnam-Dutch MA programmes, for his teaching works, especially for very useful topic: Research Methodology that seems to be unique in Vietnam so far I would like to thank all teachers and staff of the Project such as Mr Tran Vo Hung Son - Project Director, Mdm Nguyet, the Project secretary and Mdm Chi, the Project librarian The special thanks is duly devoted to Dr Karel Jansen, Dr.Youdi Schipper, not only for the teaching works, but also for granted scholarships which is contributed by Dutch people Thank to the contribution of Dutch people, many Vietnamese students and I have chance to access very academic ccurse in development economics Finally, I am deeply indebted to my family members such as my parents and my wife who always encourage me during the studying Ho Chi Minh City, September 8, 2003 Le Quoc Thanh ii ABBREVIATION AFfA : Asian Free Trade Area CEPT : Common Effective Prefereme Tariff Agreement FDI : Foreign Direct Investment GSO : General Statistical Office JN : Joint Venture JICA : Japanese International Cooperation LPM : Linear Probability Model SIDA : The Swedish International Development Agency TV(s) : Television sets VLSS :Vietnam Living Standards Survey VND :Vietnamese Dong (the currency unit ofVietnam) VElA : Vietnam Electronics and Informatics Association VLSS : Vietnam Living Standard Survey VCR : Video Cassette Recorder WTO : World Trade Organization iii TABLE OF CONTENTS Certification i Acknowledgement ii Abbreviation iii Table of Contents iv List of tables v List of figures and appendices vi CHA.PTER 1: INTRODUCTION ! 1.1 BACKGROUND OFTHE THESIS 1.2 OBJECTIVE OF THE THESIS 1.3 SCOPES AND FOCUS OF THE THESIS .4 1.4 DATA COLLECTION AND METIIODOIDGY .4 1.5 STRUCTURE OF THE THESIS CHA.PTER 2: TIIEORETICAL FRAMEWORK 2.1 SUPPLY- DEMAND 2.2 DEMAND DETERMINANTS 2.2.1 PRICES 2.2.2 INCOME 11 2.3 THEORY OF CONSUMER'S CHOICES 14 2.4 HOUSEHOlDS 23 2.5 THE QUAliTATIVE RESPOND MODEL AND DEMAND ANALYSIS 25 2.6 EMPIRICAL STUDIES 26 CHAPTER REMARKS • 29 CHAPTER 3: AN OVERVIEW OF VIETNAM'S ELECTRONICS INDUSTRY AND TELEVISIONS 31 3.1 VIETNAM'S ELECTRONICS INDUSTRY 31 3.1.1 THE HISTORY 31 iv 3.1.2 CURRENT SITUATION OF VIETNAM ElECTRONICS INDUSTRY 32 3.2 SUPPLY OFTV FROM DOMESTC PRODUCTION ANDIMPORTATION 34 3.2.1 TELEVISION ASSEMBlERS 34 3.2.2 DOMESTK:: SUPPLY OFTV AND IMPORTATION 36 3.2.3 VIETNAM'S DOMESTIC PRICE IN COMPARISON WITH THE WORLD PRICE 37 3.3 ELECTRONICS INDUSTRY ANDGLOBALINTEGRATION 38 3.4 POLICIES RECOMMENDATION 38 3.5 STATISTICAL DESCRIPTION OF TV OWNERSHIP IN VIETNAM 39 3.6 CHAPTERREMARKS 41 CHAPTER 4: EMPIRICAL ANALYSIS 42 4.1 MODELSPECIFICATION 42 4.1.1 MODEL SELECTION 42 4.1.2 MODEL SPECIFICATION 44 4.1.3 ANALYTICALFRAMEWORK 45 4.2 DATA DESCRIPTION 47 4.2.1 SOURCEOFDATA 47 4.2.2 DATA SAMPLING 48 4.3 DESCRIPTIVE STATISTICS 48 4.4 RESULT OF ESTIMATION 52 4.4.1 MODEL TEST 52 4.4.2 RESULTS OF REGRESSION 52 4.4.3 COMMENTS ON RESULTS 53 4.5 CHAPTERREMARKS 54 CHAPTER 5: CONCLUSION AND POLICY RECOMMENDATION 56 5.1 CONCWSION 56 5.2 POLICY RECOMMENDATION 56 APPENDICES REFERENCES LIST OFTABLES Page Table 2.1 :Alternative market baskets 12 Table 2.2 : Summary of demand responses to a 1% increase in income 19 Table 2.3 : Estimates of income elasticity of demand in UK 20 Table 3.1 The share of each electronics product in Vietnam 33 Table 3.2 Export of electronics product in 1991-1998 33 Table 3.3 List of TV Assemblers in Vietnam 35 Table 3.4 Market Share of some biggest TV Assemblers in Vietnam by the year 2000 36 Table 3.5 Supply of TV in Vie1nam market n the year of 1990s .36 Table 3.6 Tax reduction schedule according to AFfA/CEPT 38 Table 3.7 Share of Household owning TV in Vietnam in the year 1992 and 1998 39 Table 3.8 Share of Household owning TV in Vietnam by expenditure quintile in 1998 40 Table 3.9 Share of Household owning TV in Vietnam by areas .40 Table 4.1 Description of variables 47 Table 4.2 Probability of TV ownership across household type .49 Table 4.3: Probability of TV ownership across regions 49 Table 4.4: Probability of TV ownership across Household size 50 Table 4.5: Probability of TV ownership across Educational Level of Household head 51 Table 4.6: The result ofregressions 52 v LIST OF FIGURES AND APPENDICES • FIGURES Page Figure 2.1 : Supply and Demand Curve Figure 2.2 : Shits in supply Figure 2.3: The welfare costs of a tariff 10 Figure 2.4 : Shifts in Demand 11 Figure 2.5 : An indifference curve 12 Figure : Maximizing consumer satisfaction 14 Figure : Income ani substitution effects 15 Figure 2.8(a) and 2.8(b) Effect oflncomeChanges 17 Figure 2.9 : Engel Curve 18 Figure 2.10 : Choices of two household faced with binary choice 21 Figure 2.11 :The derivation of threshold expenditure XT(3) 22 Figure 4.1 Logit an pro bit cumulative distributions 43 Appendix 1: 58 Appendix 2: 63 vi CHAPTER INTRODUCTION 1.1 BACKGROUND OF THE THESIS After over a decade of shifting to the market economy, Vietnam has achieved considerable successes in the economic development, which resulted in a higher living standard for Vietnamese From 1994 to 1999, the nominal monthly income per capita has increased by 75% (GSO, 2000) By taking the average inflation rate of 6.04% during the year of 1995-1999 (Vietnam Economics Times, 1999-2000: 4), it might be concluded that there is considerble increase in real income of people after years of development Televisions (here abbreviated as TVs) as a durable goods has become important to Vietnamese living since television services have rapidly developed nationwide It allows people to be accessed to huge source of information, basic knowledge in many fields, which is very necessary to human being and their business activities By watching television, businessmen, companies and manufacturers can update domestic or international market information, which is quite important to their business Farmers, especially in remote areas can learn more about agricultural knowledge through agricultural extension programmes, animal health care or weather forecasts which make their agricultural activities become more scientific and less risky, which leads to higher profits as a result Children, adults and elderly now may have chances to enjoy appropriate entertainment programmes that make their life more enjoyable and interesting In the same manner, young parents can learn how to take care of their babies through televised children care programmes Pupils and students can learn foreign languages, understanding more about history or geography, etc, inside and outside the country Obtaining knowledge of these fields will certainly facilitate their learning process and thus increase their human capital In the state management aspect, the Government or state management bodies can introduce new policies, legal regulation or law in quick and cheap way to people by televised law programme Companies and manufacturers can advertise their products, services to consumers and thus facilitate their sales Consumers may enjoy open chances to a wide range of products, which considerably supports their choices In short, TVs and television is very important to people's living, state-mana~ment works and business activities as well of household head at the same educational level TV ownership status is increased along to increase in schooling year of household head in general Although, there are some decreases in some specific educational level of household head, but the general trend is going up to 100% of household in which household head having 19 years of schooling own TV This trend reflects the real situation that higher educated people tend to own TV rather than the lower educated ones By looking at the variance of TV ownership according to age of household head, educational level of household head and household size, the general trend of TV ownership in Vietnamese households is illustrated However, in order to obtain more clear and deep economic view, it is necessary to come to econometric analysis as the following part 4.4 RESULT OF ESTIMATIONS 4.4.1 Model test: The econometric results show that the Chi/\2, the test for goodness-of -fit, with its p value= 63.17% which indicate highly statically significance of the model So, ifp value is greater than 5% the model used has a rather high level of goodness-of-fit, so it is concluded that the chosen variables included in model are statistically significant In addition R-squared is 38.2 percent, which also prove suitability of the model In case of cross-sectional analysis, this R'\2 also means high level of goodness-of fit of the model 4.4.2 Results of regression: As argued earlier, the full model ofthe study will be as below: Li ~0 + ~ (LnHH_EXP) ~s(HH_EDU) + ~ (AVG price) + ~3(HH_AGE) + ~4(HH_SIZE) + + ~k(REGION) +ui 52 Table 4.6: The result of regressions Estimated probability of TV ownership when independent variable changes by one unit and nitial probability is: (%) Coef Std Err P>Jzl 3% 7% 15% 'pependent variable Does the household own television or not? 'lfndependent variables fill AGE (years old) ~H EDU (schooling years) fiHSIZE ~nHH EXP(V1{01000) (V1{0 1000) ~nA VGPRICE 0.0131 0.0616 -0.2776 3.0416 1.3271 0.0044 0.0153 0.0256 2.9582 2.8371 0.002 0 0.078 3.04 3.18 2.29 39.31 10.44 7.09 7.41 5.39 61.18 22.11 15.17 15.80 11.79 78.70 39.95 -0.6443 -0.5759 -0.6347 -0.9141 0.0664 0.1308 0.0962 0.0739 0.013 0 1.60 1.71 1.61 1.22 3.80 4.06 3.84 2.93 8.48 9.03 8.55 6.61 ~GION Red river delta (as reference) North central coast South central coast Northwest Northeast Southeast Mekong river delta Notes: There is no observation in Central Highland while North and South Central Coast are not statistically significant Number of obs Wald chi2(9) Prob > chi2 Pseudo R2 4.4.3 2702 697.92 0.0000 0.3822 Comments on results: Effect of household expenditure and average price of television These two variables are key determinants of ownership probability due to their scope of effect It shows a positive relation between probability of ownership and household expenditure With initial probability of percent, household expenditure increases by one percent; the probability of ownership will rise to 39.3 percent, keeping other things unchanged Similarly, initially probability of 15 percent, the ownership probability will increase to 78.7 percent with a change of expenditure of one percent 53 Effects of average price of television are less important than household expenditure It reveals a positive relation with the ownership probability of television; and raises the probability to 39.95 percent with initial probability of 15 percent as average price rises by one percent, keeping other things constant The effect of age of household head The age of household head has a statistically positive relationship with probability of TV ownership: below percent level of significance However, this effect of age of household head is really small and thus not the key determinants of TV ownership Given other things constant and initial probability of TV ownership of percent, the age of household head increases by one-year-old and the probability of ownership increases to 04 percent Similarly, given initial probability of 15 percent, there is an increase of oneyear-old that lead to the probability of ownership of 15.17 percent Effect of household size In contrast to above result, the household size turns out opposite to what is expected Household size has a negative relation with TV ownership probability: below percent of statistical significance It could be explained that an increase of number of household members in general, especially more dependent people will create more pressures on daily living, decrease household's living standard and other welfare of household, so-called negative effect Besides, the increasing size of household may generate needs of television ownership for entertainment, and self-education for children, so-called positive effect But the negative effect of increase in household size seems to outweigh the positive one The negative sign of estimation of household size reflects it Under initial probability of 15 percent, a decrease of household size by one member will reduce the ownership probability to 11.8 percent Effect of education Estimation of education on television ownership becomes correct as initial expectation It shows a positive relation between probability of TV ownership and the schooling years of household head Although its effect is greater than influence of age, educational level of household head does not also play as the key role of TV owningdecision Keep other things unchanged and the initial probability of percents; an 54 increase of education by one schooling year will raise the probability to 3.18 percent If initial probability of 15 percent, an increase by one schooling year will yield ownership probability of 15.8 percent Effect of regions As the regression result, the living standard of Red river delta seems to be better than other regions People living in Northwest, Northeast, Southeast, or Mekong river delta will have less probability of television ownership than people in Red river delta As an example of Mekong river delta, "other things equal", if given the initial probability of percent, the probability of television ownership in Mekong River Delta will be reduced to 1.22 percent in comparison with the others living in Mekong River delta Furthermore, at the initial probability of 15 percent, changing in living place will lead to a decrease in probability of television ownership to 6.61 percent, "other things equal" These are noted as inequality of income distribution amongst regions in Vietnam 4.5 CHAPTER REMARKS: The most important finding is that the household expenditure that represents household income level has strongly positive relationship with the level of TV ownership in Vietnamese museholds in whole country The second is the purchasing price effect on TV ownership of household The positive relationship is found between price of TV and TV ownership status in Vietnamese household This confirmed the fact that higher price of TV will certainly prevent a part of consumers at low income level from enjoying services of TV as stated in the Chapter In addition, Educational level of household head measured by schooling years has positive relationship with TV ownership This reflects the situation that higher educated people will have better chances of consumption due to their higher income Age of household head also has positive effect on TV ownership of Vietnamese households However, this effect seems to be weak, just percent of significance Thus, age of household head should not be considered as key determinants as the case of income and price above Final determinant of TV ownership that found from this study is the household size that has negative impact on TV ownership 55 CHAPTERS CONCLUSIONS AND POLICY RECOMENDATIONS This part will present conclusions of thesis and also some recommerrlation of policy based on findings on previous chapters The main conclusions as the followings : 5.1 CONSLUSIONS The Thesis is designed to focuses on demand analysis of TV in Vietnamese household By looking at the relationship between the ownership status of TV in Vietnamese household with a number of determinants, it is concluded that: (1) As mentioned theoretical framework, the determinants ofhousehold ownership ofTV include household income, household size, age of household head, educational level of household head, purchasing price of TV (2) Suggested model for examining relationship between these determinants and ownership status was tested and the results have confirmed the main determinants of TV ownership and its sign of relations with respect to TV ownership as findings mentioned in the Chapter Remark of the Chapter Based on these findings, some policies below would be suggested: 5.2 POLICY RECOMMENDATIONS: As mentioned in the Chapter 1, consumption of TV will provide many benefits, not only for people but also for the Government and business sectors Therefore, the main aim of policy should be to increase TV ownership among Vietnamese household In order to increase the probability of TV ownership in Vietnamese household, there might be some measures as : (1) Price of TV in general should be lower, allowing low-income household to be accesses to consumption of TV also (2) Lower import tax will create decrease in domestic price of TV and thus it would be a suitable policy for Vietnam where import tax of complete TV is too high, at 60% at 56 the present In addition, Government can influence pricing policy of TV makers by providing subsidy for TV that would be sold in remote area or to the poor people who buy TV Such policy would certainly faciliate poor people in owning TV (3) In long-term, income of Vietnamese household should be improved as much as possible to allow more low income household to be able to buy TV and benefit from services that TV provide (4) Household size represented by number of household member should be reduced to improve household income This will support for current policy from the Government of Vietnam: family planning policy that encourage young married couple should have around two children to avoid poverty, improving children health and nutrition also (5) Higher educational level ofhousehold head will facilitate household income and thus should be improved by providing them better chances for informal education because it might be difficult for them to attend formal education (6) Specific policies aiming at increasing household income should be applied for some regions in the North of Vietnam such as Northwest, Northeast and Central Highland where the probability of TV ownership is rather low compared with other regions Limitations and further studies At the present time, the price of television broadcasting services is not imposed and not available to put into the model for econometric analysis Operating cost of using TV should refer to electricity consumption using for TV when people is watching TV By surveying price of electricity set by authority at the time of survey , operating cost would be calculated and combined into the model It would be useful to use data in VLSS 1992-93 for econometric analysis to see how demand determinants of TV affect TV ownership in that period The remaining problem is the price of TV was not surveyed in VLSS1992-93 This problem would be solved by calculate price of TV in the period 1992-93 by using price of TV in 1997-1998 plus price consumer index 57 A number of households surveyed in VLSS 1992-93 was repeated in VLSS 19971998 Some of them might not own TV in the period 1992-93, but would own TV in the period 1997-98 thank to increase in their income We might examine whether increase household income is main determinant of such ownership or not? Further studies are thus suggested based on above remaining problems of this study 58 APPENDIX logit Y AGE EDU HHSIZE Iteration Iteration Iteration Iteration Iteration Iteration Iteration 0: 1: 2: 3: 4: 5: 6: log log log log log log log HHEXP AVGPRICE REG2 REG3 likelihood likelihood likelihood likelihood likelihood likelihood likelihood REG4 REGS REG6 REG7 -1861.1153 -1404.1426 -1266.6651 -1214.757 -1208.3945 -1208.2838 -1208.2837 Number of obs LR chi2(11) Prob > chi2 Pseudo R2 Logit estimates Log likelihood = -1208.2837 y Coef I z Std Err 2702 1305.66 0.0000 0.3508 [95% Conf Interval] P>lzl -+ -AGE EDU HHSIZE HHEXP AVGPRICE REG2 REG3 REG4 REGS REG6 REG7 cons ·1 I I I I I I I 0113312 0782204 -.2176477 0002065 0004665 -.5874586 -.5940821 2287932 -.09235 -.5346338 -.7368828 -3.747227 0041664 014174 0324359 0000112 0002978 1290393 2511817 2776582 2180707 1895096 1855456 8320098 2.72 5.52 -6.71 18.45 57 -4.55 -2.37 0.82 -0.42 -2.82 -3.97 -4.50 0.007 0.000 0.000 0.000 0.117 0.000 0.018 0.410 0.672 0.005 0.000 0.000 0031652 0504399 -.2812208 0001846 - 0001171 -.840371 -1.086389 -.3154069 -.5197607 -.9060658 -1.100545 -5.377936 0194972 1060009 -.1540746 0002284 0010501 -.3345462 -.101775 7729934 3350608 -.1632019 -.3732202 -2.116517 -note: failures and 13 successes completely determined logit Y AGE EDU HHSIZE Iteration Iteration Iteration Iteration Iteration Iteration Iteration log log log log log log log 0: 1: 2: 3: 4: 5: 6: HHEXP AVGPRICE REG2 REG3 likelihood likelihood likelihood likelihood likelihood likelihood likelihood -1861.1153 -1408.5391 -1266.8799 -1215.0995 -1208.8831 -1208.7783 -1208.7783 Number of obs LR chi2 (9) Prob > chi2 Pseudo R2 Logit estimates Log likelihood= -1208.7783 Y I Coef I I I I I I I 0115216 0787015 -.2165838 0002064 0005053 -.5928318 -.5986189 REG6 REG7 Std Err z P> I z I 2702 1304.67 0.0000 0.3505 [ 95% Conf Interval] -+ -AGE EDU HHSIZE HHEXP AVGPRICE REG2 REG3 0041538 0141328 0323496 000011 0002946 122877 2486709 2.77 5.57 -6.70 18.74 72 -4.82 -2.41 0.006 0.000 0.000 0.000 0.086 0.000 0.016 0033803 0510017 -.2799879 0001848 -.000072 -.8336662 -1.086005 019663 1064014 -.1531796 000228 0010827 -.3519974 -.1112329 59 REG6 REG7 cons -.S4102S -.7408369 -3.8S9201 1808S66 1792277 8239279 -2.99 -4.13 -4.68 -.89S4974 -1.092117 -S.47407 0.003 0.000 0.000 -.186SS2S -.389SS71 -2.244332 note: failures and 13 successes completely determined logit Y AGE EDU HHSIZE robust Iteration Iteration Iteration Iteration Iteration Iteration 0: 1: 2: 3: 4: S: log log log log log log lnHHEXP lnAVGPRICE REG2 REG3 likelihood likelihood likelihood likelihood likelihood likelihood -1861.11S3 -1240.3726 -11S7.8S63 -1149.0792 -1148 92SS -1148 92S4 Logit estimates Number of obs Wald chi2 {11) Prob > chi2 Pseudo R2 Log likelihood= -1148.92S4 y I I Coef I I I I I I I I I I I I 0130694 0600464 -.2782707 3.06986S 313388 -.681843 -.609S029 OS0697S - 27112SS -.6906737 -.9628886 -38.47161 REG4 REGS REG6 REG7, Robust Std Err z P>lzl 2702 697.S2 0.0000 0.3827 [9S% Conf Interval) -+ -AGE EDU HHSIZE lnHHEXP lnAVGPRICE REG2 REG3 REG4 REGS REG6 REG7 cons 0043284 014478 0338097 1447SS9 7S46027 1337166 2360764 3180018 224842S 1926S68 1926909 031113 3.02 4.1S -8.23 21.21 74 -S.10 -2.S8 0.16 -1.21 -3.S8 -s.oo -6.38 0.003 0.000 0.000 0.000 0.082 0.000 0.010 0.873 0.228 0.000 0.000 0.000 004S8S9 0316701 -.344S36S 2.786149 -.16S6062 -.9439226 -1.072204 -.S72S74S -.7118088 -1.068274 -1.340SS6 -S0.29237 021SS28 0884228 -.2120048 3.3S3S81 2.792382 -.4197633 -.1468017 6739696 169SS78 -.3130733 -.S8S2214 -26.6S084 - sw logit Y AGE EDU HHSIZE lnHHEXP lnAVGPRICE REG2 REG3 robust pr (0 1) begin with full model p 0.8733 >= 0.1000 removing REG4 p = 0.2133 >= 0.1000 removing REGS Logit estimates Number of obs Wald chi2(9) Prob > chi2 Pseudo R2 Log likelihood= -1149.72S7 y I I Coef REG4 REGS REG6 REG7, Robust Std Err z P>lzl 2702 697.92 0.0000 0.3822 [9S% Conf Interval) -+ -AGE EDU HHSIZE lnHHEXP lnAVGPRICE REG2 I I I I I 0130S36 061S784 -.2776072 3.041648 1.327134 -.6443188 00431S3 0144237 0337333 1412709 7S2S129 1264248 02 4.27 -8.23 21.S3 1.76 -S.10 002 0.000 0.000 0.000 0.078 0.000 004S9S7 0333084 -.3437233 2.764762 -.1477643 -.8921069 021S11S 0898484 -.2114912 3.318S33 2.802032 -.396S307 60 REG3 REG7 REG6 cons -.5759294 -.9141049 -.634742 -38.37077 2325836 1842337 1814856 6.010673 logistic Y AGE EDU HHSIZE -2.48 -4.96 -3.50 -6.38 o.ooo -1.031785 -1.275196 -.9904472 -50.15147 -.1200739 -.5530135 -.2790368 -26.59007 lnHHEXP lnAVGPRICE REG2 REG3 REG6 REG7, robust Logit estimates Number of obs Wald chi2(9) Prob > chi2 Pseudo R2 Log likelihood= -1149.7257 Y I Odds Ratio 0.013 0.000 0.000 Robust Std Err z P> I z I 2702 697.92 0.0000 0.3822 [ 95% Conf Interval] -+ -AGE EDU HHSIZE lnHHEXP lnAVGPRICE REG2 REG3 REG6 REG7 I I I I I I I I I 1.013139 1.063514 7575943 20.93971 3.770222 52502 5621821 5300722 4008753 004372 0153399 0255562 2.958172 2.83714 0663756 1307543 0962005 0738547 3.02 4.27 -8.23 21.53 1.76 -5.10 -2.48 -3.50 -4.96 0.002 0.000 0.000 0.000 0.078 0.000 0.013 0.000 0.000 1.004606 1.033869 7091251 15.87526 8626344 4097914 3563703 3714106 2793761 1.021744 1.094008 8093764 27.61981 16.47809 6726496 8868549 7565121 5752138 lfit Logistic model for Y, goodness-of-fit test number of observations number of covariate patterns Pearson chi2(2659) Prob > chi2 2702 2669 2633.87 0.6317 lroc Logistic model for Y number of observations area under ROC curve 2702 0.8839 61 Area under ROC curve= 0.8839 1.00 0.75 ~ ·;;: :;:::; "iii c: 0.50 Q) en 0.25 0.00 0.00 0.25 0.50 - Specificity 0.75 1.00 0.00 0.25 0.50 Probability cutoff 0.75 1.00 lsens 1.00 0.75 ~ "(3 q:: "(3 Q) c en > :!: 0.50 > :;:::; "iii c: Q) en 0.25 0.00 62 lstat Logistic model for Y True -Total Classified I D -D -+ + + I 1201 264 I 1465 I 276 961 I 1237 -+ + Total 1477 1225 2702 Classified + if predicted Pr(D) >= True D defined as Y -= Sensitivity Specificity Positive predictive value Negative predictive value Pr( +I D) Pr( -1-D) Pr ( Dl +) Pr(-DI -) 81.31% 78.45% 81.98% 77.69% False + rate for true -D False - rate for true D False + rate for classified + False - rate for classified - Pr( +I-D) Pr( -I D) Pr(-DI +) Pr( Dl -) 21.55% 18.69% 18.02% 22.31% Correctly classified 80.01% 63 APPENDIX2 Regions VLSS97-98 Code Red River delta HaNoi , Hai Phong, Ha Tay, Hai Duong Hung Yen, Ha Nam, Nam Dinh, Thai Binh, Ninh Binh Northeast Ha Giang, Cao Bang, Lao Cai, Bac Can, Lang Son, Tuyen Quang, Yen Bai, Thai Nguyen, Phu Tho, Vinh Phuc, Bac Giang, BacNinh, QuangNinh Northwest Lai Chau, Son La, Hoa Binh North central coast Thanh Hoa, Nghe An, Ha Tinh, Quang Binh, Quang Tri, Thua Thien- Hue South central coast Da Nang, Quang Nam, Quang Ngai, Binh Dinh, Phu Yen, Khanh Hoa Central highlands Kon Tum, Gia Lai, Dac Lac Southeast TP Ho Chi Minh, Lam Dong, Ninh Thuan, Binh Phuoc, Tay Ninh, Binh Duong, Dong Nai, Binh Thuan, Ba Ria- Vung Tau Mekong delta Long An, Dong Thap, An Giang, Tien Giang, Vinh Long, Ben Tre, Kien Giang, Can Tho, Tra Vinh, Soc Trang, Bac Iieu,Ca Mau 64 REFERENCES Amaniya T (1981), Qualitative Response Model: A Survey, Journal of Economic Literature, Volume 19, Issue (dec., 1981) Anh, Tran Hoang (200 1), Determinants of income of household income in rural Vietnam, Vietnam-Dutch programme for MA in Development Economics Begg, D., Fisher S., and Dornbusch R (1991), Economics, 3rd edn, London: McGraw Hill Chon, LeVan (2001), Determinants of enrollment in Vietnam's secondary education, Vietnam-Dutch programme forMA in Development Economics Cramer J S., (1991), The Logit Model, London: Edward Arnold Deaton, A (1997), The Analysis of Households Surveys: A Microeconometric Approach to Development Policy, Baltimore and London: The John Hopkins University Press Deaton A and J Muellbauer, (1980), Economics and Consumer Behavior, Cambridge: Cambridge University Press GSO (1997), Vietnam Living Standard Survey Questionnaire GSO, (1999), Vietnam Living Standard Survey 1997- 1998 GSO, (1999), Vietnam Living Standard Survey 1997-1998 CD-ROM Gujarati, D N (1995), Basic econometrics, 3rd edn, New York: McGraw Hil Hamilton L C (1998), Statistics with Stata 5, Pacific Grove: Duxbury Press Haugton, D., J Haugton, Bales, S., Chuyen, T T K and Nga, N N (1999), Health and Wealth in Vietnam: An Analyze of Household Living Standards, Pasir Panjang: Seng Lee Press Pte Ltd Himmelwei~ S., A Trigg, N Costello, et al., (1998), Understanding economic Behavior: Households, Firms and Markets, 1st Edn, Glasgow: Bath Press 65 JICA and National Economics University (2002) Industrial and Trade Policies of Vietnam under International Integration, Seminar Document Joint Report of the Government Donor, NGO Working Group, (2000), Vietnam Development Report 2000: Attacking poverty, Draft fur discussion Kooreman, P and S Wunderink, (1997) The Economics of Household Behavior, 1st edn, London: Macmillan Press, Limited Pindyck, R and D L Rubinfeld, (1992), Microeconomics, New York: Macmillan Phuoc, Le Huu (2000) Thuc trang nganh cong nghiep dien tu VietNam va mot so bien phap phat trien nganh dien tu, Dai hoc kinh te Tp.HCM Sadoulet E and A de Janvry (1995), Quantitative Development Policy Analysis, London: John Hopkins Press Ud., Train, K E (1986), Quantitative Choice Analysis: Theory, Econometrics, and an application to Automobile Demand, Cambridge: MA: The MIT Press The Vietnamese Electrmics Industry, (2002-2003), GC.Comm Tuan, Nguyen Anh (2002) Hydropower in Vietnam: Updates on Current and future developments, Electricity of Vietnam 66 ... Household owning TV in Vietnam in the year 1992 and 1998 39 Table 3.8 Share of Household owning TV in Vietnam by expenditure quintile in 1998 40 Table 3.9 Share of Household owning TV in Vietnam. .. of Economics Ho Chi Minh City Institute of Social Studies The Hague Vietnam The Netherlands VIETNAM- THE NETHERLANDS PROGRAMME FOR MASTER OF ARTS IN DEVELOPMENT ECONOMICS DEMAND ANALYSIS OF TELEVISIONS. .. it is expected that demand of TVs will increase more than percent when income of consumer increased by percent Demand for durable goods: Before going into demand analysis of durable goods, it