Economics of charitable giving understanding the motivation of donation behaviour

VIETNAM – NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS ECONOMICS OF CHARITABLE GIVING: UNDERSTANDING THE MOTIVATION OF DONATION BEHAVIOR NGUYEN NGOC NU HO CHI MINH CITY, DECEMBER 2015 MASTER OF ARTS IN DEVELOPMENT ECONOMICS ABSTRACT This paper studies the individual determinants of charitable giving including in time, donation in gift, and money donation The research analyzes survey data in Ho Chi Minh City on charity giving in the last 12 months with 500 individuals We design a questionnaire to collect information on the donation behavior, individual characteristics, and perceptual and attitudinal factors A multivariate probit model is applied to analyze the three related behavior of in-kind, time and money donation We found that richer, religious and female respondents are more likely to give Perceptual and attitudinal factors are not significant Particularly we found that altruism, warm glow, prestige and reciprocity, and government not have a statistical significant impact on all kinds of donation The only exception is family influence, implying that the giving of the current generation would result in good giving habit of the future generations Key words: Economics of Charity Giving, altruism, pure altruism, impure altruism, warm glow, prestige, and reciprocity ACKNOWLEDGEMENT I would like to thank for special persons who supported me during the tough time of writing this thesis Specially, I would like to express my gratitude to all the lectures who provided evaluable knowledge and supported me to complete the thesis I am also indebted to Dr Pham Khanh Nam who positively encourage and assist me to finish the thesis process Additionally, thanks are also due to all the VNP staffs who create favorable conditions in my process of learning and researching: Ms Xuan Hong, Ms Man Thi, Mr Nhan Tam, Mr Quang Huy and many others For my family who created the conditions to me on the academic journey The completion of this dissertation is a valuable gift which assist me to go further in the studying process as well as my career in the future I recognized that although studying is really challenging, yet it is more interesting than I thought For this, I want to give a special thank to my mom who movetivated and supported me to finish the survey Completing the survey process is fraught with a lot of difficulties I acknowledge the contribution of all my friends, all the respondents who also enthusiastically supported me to fulfill the surveys Especially, I would like to say thank you, Le Viet Thanh, who guides me some necessary knowledge about running stability in working process My main research idea is about the charitable donation In the working process, I also receive the donated gift from many people That was the unforgettable journey with lots of enjoyable experiences Thank you, Nguyen Ngoc Nu Dec, 2015 TABLE OF CONTENTS Chapter 1: Introduction 1.1 Problem Statement 1.2 Research Objectives and scope of study 10 1.3 Structure of the thesis 11 2.1 Theoretical Models of Charity Giving .12 2.1.1 The Pure Altruism Model 12 2.1.2 The Impure Altruism Models 13 2.1.2.1 The Warm-glow Giving Model 13 2.1.2.2 The Conspicuous Giving Model .13 2.1.2.3 The Reciprocity Model .13 Chapter 2: Literature Review 2.2 A Review of empirical studies on charity giving 14 2.2.1 Empirical Studies .14 2.2.2 Experimental Studies 15 2.2.3 Psychology Studies 16 Chapter 3: Research Methodology 3.1 The model 18 3.2 Data collection method 19 3.3 the survey instruments 19 3.4 econometric models and Hypothesises 20 3.4.1 Logistic regression 20 3.4.2 Multivariate probit model 20 3.5 The independent variables .20 3.5.1 Individual Characteristics 20 3.5.1.1 Gender 20 3.5.1.2 Age 21 3.5.1.3 Education 21 3.5.1.4 Religion 21 3.5.1.5 Income 22 3.5.2 Altruism 22 3.5.3 Warm glow .23 3.5.4 Prestige .23 3.5.5 Reciprocity .23 3.5.6 Social influence 24 3.5.6.1 Government influence 24 3.5.6.2 Tax Incentive 24 3.5.6.3 Family Influence 24 Chapter 4: Analysis of Charity Giving Behavior 4.1 Respondents Profile 27 4.1.1 The Charity Giving 27 4.1.2 Gender 28 4.1.3 Education 28 4.1.4 Income 28 4.1.5 Religion 29 4.1.6 Age 29 4.1.7 Perception and attitude toward charity giving 29 4.2 Determinants of Donation: Econometric Analysis 32 4.2.1 Logit model 32 4.2.2 Multivariate Logistic Regression 34 4.3 Analysis of Donation Behavior 36 4.3.1 Information Channels .36 4.3.2 Kind of Donation 39 4.3.3 Kind of Money Donation 40 4.3.4 Donation Sectors 42 Chapter 5: Conclusions and Policy Implications 5.1 Research Objective Summary 46 5.2 The Regression Models Summary .46 5.3 The Data Collection Summary 46 5.4 The Main Finding 47 From Logit Model .47 From Multivariate Probit .47 From The Statistical Test .47 5.5 Policy Implication 48 5.6 Limitations .48 TABLE OF APPENDICES Appendix Pretest Questionnaire 52 Appendix The Final Questionnaires 67 Appendix Stata Analyses Output 71 Appendix 72 Appendix 73 Appendix 74 Appendix 75 Appendix 76 Appendix 77 10 Appendix 10 78 LIST OF TABLES Table Perceptual/Attitudinal questions used in this studies 17 Table Explanatory variables of the logit and multivariate probit models 24 Table Summary statistics .27 Table Percentage of donation by individual characteristics and Chi2 test .28 Table Respondents’ Age Descriptive Statistics .29 Table Summary of statistic results 30 Table Classification Predicted Table 32 Table Logistic Regression Results 33 Table Multivariate probit estimation results for in-kind gift, time, and money donation 35 Table 10 Percent of each cases of information channel 36 Table 11 Information Channel and Income .37 Table 12 Information Channel and Gender .37 Table 13 Information Channel and Religion .38 Table 14 Information Channel and Education 38 Table 15 Information Channel and Age 39 Table 16 Percent of Cases in Kind of Donation .39 Table 17 Donation Kind via Income, Gender, Religion, and Education 40 Table 18 Kind of Donation and Age 40 Table 19 Percent of cases in money donation kind 41 Table 20 Kind of Money via Income, Gender, Religion, and Education 41 Table 21 Methods of Money Transfer and Age .41 Table 22 Percent of Cases Donation Sectors 42 Table 23 Donation Sectors and Income 43 Table 24 Donation Sectors and Gender 43 Table 25 Donation Sectors and Religion 44 Table 26 Donation Sectors and Education 44 Table 27 Age and Donation Sectors 45 CHAPTER INTRODUCTION 1.1 PROBLEM STATEMENT In any society, there are people who are lagging behind the overall growth and development of the economy, including the poor, minor ethnic groups, people with illness and disability, and people suffering natural or personal/family disasters And these groups of people account for a considerable proportion According to The World Bank (2015), there is 12.7% of the world’s population in 2012, or 896 million people with daily income below $1.9 FAO (2014) reports that there are 794.6 million people with undernourishment around the world in 2014, which concentrated in the developing countries Climate change, global warming, and environmental disaster result in more than 300,000 deaths per year according to the World Hunger Education Service (2015) More than one million children in the world are living in the poverty situation with 22,000 children deaths per day in 2014 (UNICEF, 2015) In addition, the lives of more than million children are threatened by fatal but preventable diseases The Disabled World (2015) reported that people with disability accounts for 10% of global population, with 650 million people, and 20% of them is in the world poorest people in 2014 Leaving these people struggling themselves may result in a society with substantial unhappiness, inequality As a result, many economies while pursuing the target of efficiency also try to provide help and assistance to people who are in need Great efforts have been devoted to relieve their pains, to help them to respond and recover from disasters, as well as to improve their ability to reside According to the US Government Spending (2015), the amount of expenditure for beneficiaries of the federal US government is $888 billion Moreover, the spending for old age survivor issue was $ 440 per year per person and for disability issue was $90 billion per year in 2005 Eurostat Statistics (2015), presented the pension beneficiaries across 21 countries among 28 EU members in 2014, that there was a total of EUR 1717 billion spent for unemployment, survivors, disability, and old age in 2012 The spending was 17.5% of GDP in Greece; 15% of GDP in Italy, France, and Austria; 7.9% in Estonia; 7.3% in Ireland; and 7.7% in Lithuania Funding for these relief, recovery and resilience activities are certainly from taxation And it is well known that taxes distort markets Goods and services taxes result in inefficient allocation of resources and lead to deadweight loss Income taxes weaken the incentive to work and thus result in a reduction in overall welfare Charity giving is an alternative sources of funding for the activities of helping people in need Charity donation is believed to be more efficient than taxation as it does not distort I want to be recognized The tax benefit of giving is the main motive Because I will receive local prestige In some case, I want to give back the benefit 5 for me to perform charity from charity giving I join in charity donation because of government suggestion Part B: Respondents Profile In this part, please tell me about yourself 10 Gender  Male ��Female 11 Year of birth: … 12 Education:  High school � Upper high school 13 Monthly Income:  million VND ��More than million VND 14 Religion :  Yes ��No 70 Appendix Stata Analyses Output Logistic regression Number of obs LR chi2(15) Prob > chi2 Pseudo R2 Log likelihood = -124.37603 charity Coef Std Err gender age education religion income q1 q2 q3 q4 q5 q6 q7 q8 q9 q10 _cons -1.288058 0749029 5896281 3.219022 2.034262 1.223715 3164695 -.079889 -1.0079 1.829458 4608119 -2.143921 2589032 -.9143583 1.834079 -6.501922 3687952 0164282 4234195 3886855 4721801 5944152 4419429 4607389 4987486 5023488 6246343 8448009 6555101 6602706 4782714 8850331 z -3.49 4.56 1.39 8.28 4.31 2.06 0.72 -0.17 -2.02 3.64 0.74 -2.54 0.39 -1.38 3.83 -7.35 P>|z| 0.000 0.000 0.164 0.000 0.000 0.040 0.474 0.862 0.043 0.000 0.461 0.011 0.693 0.166 0.000 0.000 = = = = 453 330.66 0.0000 0.5707 [95% Conf Interval] -2.010883 0427042 -.2402588 2.457213 1.108806 0586829 -.5497226 -.9829206 -1.985429 8448729 -.7634489 -3.7997 -1.025873 -2.208465 8966847 -8.236555 -.5652323 1071016 1.419515 3.980832 2.959718 2.388748 1.182662 8231426 -.0303705 2.814044 1.685073 -.4881417 1.543679 3797483 2.771474 -4.767289 71 Appendix Variable VIF 1/VIF q1 q3 11.41 8.06 0.087662 0.124112 q2 age 6.84 6.70 0.146147 0.149230 income q4 q5 5.46 5.24 4.69 0.183250 0.190941 0.213243 religion q8 4.44 2.54 0.225110 0.392972 q9 gender 2.19 2.12 0.456218 0.471062 q7 q10 2.05 2.01 0.488151 0.498243 education q6 1.94 1.57 0.515451 0.637302 Mean VIF 4.48 Appendix Logistic regression Number of obs LR chi2(14) Prob > chi2 Pseudo R2 Log likelihood = -124.39109 charity Coef Std Err gender age education religion income q1 q2 q4 q5 q6 q7 q8 q9 q10 _cons -1.28021 0748126 5923099 3.224222 2.025981 1.180826 3147099 -1.012214 1.817148 4548754 -2.124598 2531122 -.9125041 1.832437 -6.517427 3659292 0164294 4233591 3875836 4695702 540731 4415193 4987805 4970994 622788 8377261 6543757 6595566 4778916 8805403 z -3.50 4.55 1.40 8.32 4.31 2.18 0.71 -2.03 3.66 0.73 -2.54 0.39 -1.38 3.83 -7.40 P>|z| 0.000 0.000 0.162 0.000 0.000 0.029 0.476 0.042 0.000 0.465 0.011 0.699 0.167 0.000 0.000 = = = = 453 330.63 0.0000 0.5706 [95% Conf Interval] -1.997418 0426116 -.2374587 2.464572 1.10564 1210131 -.550652 -1.989806 8428514 -.7657666 -3.766511 -1.029441 -2.205211 8957869 -8.243254 -.5630019 1070137 1.422078 3.983872 2.946322 2.24064 1.180072 -.0346222 2.791445 1.675517 -.4826848 1.535665 380203 2.769088 -4.7916 Appendix Variable VIF 1/VIF q1 q2 age income q4 q5 religion q8 q9 gender q7 q10 education q6 8.65 6.79 6.57 5.38 5.22 4.51 4.41 2.54 2.19 2.09 2.02 2.01 1.94 1.57 0.115660 0.147289 0.152204 0.185704 0.191606 0.221914 0.226678 0.394049 0.456298 0.478630 0.495646 0.498745 0.515575 0.637324 Mean VIF 3.99 Appendix Logistic regression Number of obs Wald chi2(14) Prob > chi2 Pseudo R2 Log pseudolikelihood = -124.39109 charity Coef gender age education religion income q1 q2 q4 q5 q6 q7 q8 q9 q10 _cons -1.28021 0748126 5923099 3.224222 2.025981 1.180826 3147099 -1.012214 1.817148 4548754 -2.124598 2531122 -.9125041 1.832437 -6.517427 Robust Std Err .3852036 0131486 4403632 4001478 3988551 523134 4400608 4661325 4610101 6522805 8613866 8141501 6877711 5018802 7691307 z -3.32 5.69 1.35 8.06 5.08 2.26 0.72 -2.17 3.94 0.70 -2.47 0.31 -1.33 3.65 -8.47 P>|z| 0.001 0.000 0.179 0.000 0.000 0.024 0.475 0.030 0.000 0.486 0.014 0.756 0.185 0.000 0.000 = = = = 453 132.53 0.0000 0.5706 [95% Conf Interval] -2.035195 0490419 -.270786 2.439947 1.244239 1555026 -.5477934 -1.925817 9135851 -.8235709 -3.812884 -1.342593 -2.260511 8487703 -8.024895 -.5252247 1005833 1.455406 4.008498 2.807723 2.20615 1.177213 -.0986111 2.720711 1.733322 -.436311 1.848817 4355024 2.816104 -5.009958 Appendix Marginal effects after logit y = Pr(charity) (predict) = 8027271 variable dy/dx Std Err gender* age educat~n* religion* income* q1* q2* q3* q4* q5* q6* q7* q8* q9* q10* -.2085868 0118614 0889839 6071716 4160612 238153 0527601 -.0124831 -.1502946 2958511 0651341 -.4699614 0385764 -.1730602 2510282 06446 00227 0611 06198 0836 11687 07738 05687 06705 0744 08036 19012 11276 14587 05401 z -3.24 5.23 1.46 9.80 4.98 2.04 0.68 -0.22 -2.24 3.98 0.81 -2.47 0.34 -1.19 4.65 P>|z| [ 0.001 0.000 0.145 0.000 0.000 0.042 0.495 0.826 0.025 0.000 0.418 0.013 0.732 0.235 0.000 95% C.I -.334922 007414 -.030764 485699 252206 009096 -.098896 -.123951 -.281712 150034 -.092369 -.842588 -.182434 -.458959 145178 ] -.082251 016309 208732 728644 579917 46721 204416 098985 -.018877 441668 222638 -.097334 259587 112838 356879 (*) dy/dx is for discrete change of dummy variable from to X 472406 32.7108 364238 682119 803532 83223 774834 774834 602649 518764 10596 064018 119205 134658 357616 Appendix Multivariate probit (SML, # draws = 500) Log pseudolikelihood = -601.23971 Number of obs = 453 Wald chi2(45) = 264.08 Prob > chi2 = 0.0000 Robust Coef Std Err z P>|z| [95% Conf Interval] giftinkind gender -.0523802 1411424 -0.37 0.711 -.3290143 2242539 age 0073961 0059216 1.25 0.212 -.00421 0190022 education 2658538 1550097 1.72 0.086 -.0379597 5696673 religion 7771537 1724273 4.51 0.000 4392024 1.115105 income 4526875 1922394 2.35 0.019 0759052 8294699 q1 -.0890494 2541126 -0.35 0.726 -.5871009 409002 q2 3851184 2144501 1.80 0.073 -.0351962 8054329 q3 -.0439215 2308314 -0.19 0.849 -.4963427 4084998 q4 -.172364 2029283 -0.85 0.396 -.5700961 2253681 q5 3690725 2021084 1.83 0.068 -.0270527 7651977 q6 1588256 2807233 0.57 0.572 -.391382 7090333 q7 -.719733 4321988 -1.67 0.096 -1.566827 127361 q8 0176964 3407005 0.05 0.959 -.6500643 685457 q9 3580664 2728608 1.31 0.189 -.1767308 8928637 q10 1913057 1471274 1.30 0.194 -.0970587 47967 _cons -2.261066 3291613 -6.87 0.000 -2.906211 -1.615922 gender -.038803 1625932 -0.24 0.811 -.3574798 2798737 age -.0037099 0066474 -0.56 0.577 -.0167386 0093188 education 1173318 1742109 0.67 0.501 -.2241153 4587789 religion 7217603 2256694 3.20 0.001 2794565 1.164064 income 4825038 2439849 1.98 0.048 0043021 9607055 q1 1385695 3458843 0.40 0.689 -.5393512 8164902 q2 -.3627593 2496138 -1.45 0.146 -.8519934 1264748 q3 -.1334505 2728226 -0.49 0.625 -.6681731 4012721 q4 3445398 231622 1.49 0.137 -.109431 7985106 q5 2532597 2229493 1.14 0.256 -.1837129 6902324 q6 -.2813875 2879148 -0.98 0.328 -.8456901 282915 q7 3972831 4006326 0.99 0.321 -.3879425 1.182509 giftintime q8 1708776 349766 0.49 0.625 -.5146512 8564064 q9 -.3279584 2921439 -1.12 0.262 -.9005499 2446332 q10 2034038 1596413 1.27 0.203 -.1094875 516295 _cons -2.060114 3462353 -5.95 0.000 -2.738723 -1.381506 age -.4146986 027377 1516726 0065585 -2.73 4.17 0.006 0.000 -.7119714 0145226 -.1174258 0402314 education 3026534 1698641 1.78 0.075 -.0302741 635581 religion 1.346427 1746986 7.71 0.000 1.004024 1.68883 income 635186 1728773 3.67 0.000 2963527 9740193 q1 3783863 2511945 1.51 0.132 -.1139458 8707185 q2 0865247 2051161 0.42 0.673 -.3154955 4885449 q3 2554746 2241028 1.14 0.254 -.1837589 6947081 q4 -.2702772 2276444 -1.19 0.235 -.7164519 1758975 q5 4631999 2147075 2.16 0.031 0423809 884019 q6 -.0357675 3076036 -0.12 0.907 -.6386595 5671245 q7 -1.311663 4672054 -2.81 0.005 -2.227368 -.395957 q8 -.3715744 3444368 -1.08 0.281 -1.046658 3035092 q9 2377875 2929474 0.81 0.417 -.3363788 8119538 q10 1597776 1757561 0.91 0.363 -.184698 5042532 _cons -2.687164 3342485 -8.04 0.000 -3.342279 -2.032049 moneydonation gender /atrho21 182347 0983484 1.85 0.064 -.0104123 3751063 /atrho31 0952477 0943318 1.01 0.313 -.0896393 2801346 /atrho32 264262 1150985 2.30 0.022 038673 489851 rho21 1803525 0951494 1.90 0.058 -.0104119 35845 rho31 0949607 0934812 1.02 0.310 -.0894 2730297 rho32 2582776 1074206 2.40 0.016 0386538 4540981 Likelihood ratio test of chi2(3) = rho21 = rho31 = rho32 = 0: 9.69212 Prob > chi2 = 0.0214 10 Appendix 10 Charity Organization Sector: (E1-E5, Y1-Y4) E1: The organization provides me the way how money will be used clearly Valid Frequency Percent 0.6 17 10.7 12 7.5 82 51.6 47 29.6 159 100.0 Frequency E1 90 80 70 60 50 40 30 20 10 E2: The organization has clearly operation purpose Valid Frequency 11 16 71 59 159 Percent 1.3 6.9 10.1 44.7 37.1 100.0 Frequency E2 80 71 70 59 60 50 40 30 20 10 16 11 E3: The organization uses money effectively Valid Frequency 17 16 79 43 5 Percent 2.5 10.7 10.1 49.7 27.0 159 100.0 Frequency E3 90 80 70 60 50 40 30 20 10 79 43 17 16 E4: The organization keep in touch frequently Valid Frequency Percent 2.5 17 10.7 21 13.2 91 57.2 26 16.4 159 100.0 Frequency E4 100 91 90 80 70 60 50 40 30 17 20 26 21 10 E5: The organization make me feel like I’m helping to solve problem Valid Frequency Percent 1.9 19 11.9 28 17.6 83 52.2 26 16.4 159 100.0 Frequency E5 90 83 80 70 60 50 40 30 28 20 10 26 19 3 80 Y1: I volunteer my time to this organization Valid Frequency 17 30 81 31 Percent 10.7 18.9 50.9 19.5 159 100.0 Frequency Y1 90 81 80 70 60 50 30 17 20 10 31 30 40 Y2: I will be a loyal supporter to this organization Valid Frequency Percent 5.0 20 12.6 21 13.2 87 54.7 23 14.5 159 100.0 Frequency Y2 100 87 90 80 70 60 50 40 30 20 10 20 21 23 81 Y3: I care about long-term success of this organization Valid Frequency Percent 11 6.9 18 11.3 20 12.6 85 53.5 25 15.7 159 100.0 Frequency 90 85 80 70 60 50 40 30 20 10 18 20 25 11 82 Y4: I continue to donate next year Valid Frequency Percent 19 11.9 11 6.9 17 10.7 83 52.2 29 18.2 159 100.0 Frequency 90 83 80 70 60 50 40 30 29 19 20 17 11 10 ... 70 with the average of age of 33 (See Table 5) In terms of charity donation, there the average of age of the donating group is 35, while that of the non-donating group is 28 To test for the difference... which stated that there is no difference between the mean of donating sample of 35 and the mean of observed groups On the other hand, the other kind of money donation have the P-valueless than10%... the percentage of donation of 75.1% and the other 52.8% The pvalue equal 0.000 Hence, there is difference between the proportions of importance group and the other in charity donation (See Table