The Vietnam Access to Resources Household Survey Supporting Evidence-based Policy through Data Collection, Capacity Building and Collaboration Professor Finn Tarp, University of Copenhagen and UNU- WIDER
Background Vietnam has made substantial socio-economic progress... Challenges remain: Completing the on-going structural transformation of the economy Distributing the benefits of economic growth, especially to rural areas... Much remains to be learnt about the pattern of changes in wealth and welfare in a socialist-oriented market economy 2
Why Survey Data? Surveying and re-surveying the same households over time is difficult and time-consuming So: why invest in panel data? How household welfare changes over time matters and we want to understand its determinants Macroeconomic growth means that choice and material welfare are indeed improving on average But: averages hide a lot of variation... and variation reveals a lot about underlying realities and point us to what to do about it 3
The VARHS Project The VARHS project supports policymakers with Vietnamese and international research expertise We produce detailed information about rural households, including Creating bi-annual panel data sets, building capacity to do research, producing original research Implemented by the University of Copenhagen, CIEM, ILSSA, and CAP-IPSARD 4
The VARHS Project (continued) VARHS is a unique panel survey which allows us to investigate whether: Some areas or groups are being left behind Households are not sharing equally in overall economic growth Some regions need additional resources or policy innovation It supplements and extends the VHLSS Supplements: repeated surveys of the same households (a unique panel dataset) Extends: asks questions about land, agriculture, income, spending, assets, investments, market linkages, and much more 5
VARHS Over Time The Vietnam Access to Resources Household Survey (VARHS) has been implemented since 2002, in 2006, 2008, 2010 and 2012 (with financial support from Danida since 2006) 2012 round includes updates And new questions about migration, climate change, and social welfare 6
VARHS Provinces Red River Delta: Ha Tay North East: Lao Cai and Phu Tho North West: Lai Chau and Dien Bien North Central Coast: Nghe An South Central Coast: Quang Nam and Khanh Hoa Central Highlands: Dak Lak, Dak Nong, and Lam Dong Mekong River Delta: Long An 7
Sample Size Has Grown over Time 2002: 932 households in four provinces 2006: 2,324 households in 12 provinces 2008: 3,223 households 2010: 3,202 households, of which 2,200 panel households 2012: 3,700 households surveyed Of which 2,741 are included in the 2012 report 2,197 panel households 8
The 2012 (Most Recent) Descriptive Report Presented in August at CIEM, available in print / online Produced collaboratively between partner institutions University of Copenhagen, ILSSA, CAP-IPSARD, and CIEM Reflects shared commitment to building research capacity, for example Research visit to the University of Copenhagen Applied economics course given in Ha Noi Continuous capacity building through collaboration 9
Relationship to VLSS / VHLSS General Statistics Office has implemented the Vietnam Household Living Standards Survey (VHLSS) since 1992 with support from the World Bank and others While VARHS follows households over time, VHLSS generates nationally representative data in each crosssection But does not track a large number of households across multiple survey rounds. The VARHS supplements and extends the VHLSS. 2,200 households surveyed in 2006 are included in each subsequent round Provides repeated data points about these households over time 10
Descriptive report is an extensive summary of the whole data set But many questions important to policy need indepth analysis Especially to identify causality: averages and trends along cannot do this Difficult and time-consuming But essential input to policymaking 11
Studies Based on 2008 Data The Effectiveness of Credit in Poverty Alleviation: An Application to Rural Vietnam Access to Land: Market- and non-market Land Transactions in Rural Vietnam 12
Studies Based on 2010 Data "Constraints to market participation in agriculture in Vietnam Income Shocks and Household Risk-Coping Strategies: The Role of Formal Insurance in Rural Vietnam Inter- and Intra-Farm Land Fragmentation in Vietnam Non-farm Income, Diversification, and Income: Evidence from Rural Vietnam The availability and effectiveness of credit in rural Vietnam: Evidence from the Vietnamese Access to Resources Household Survey 2006-2008-2010 Social Capital and Savings Behavior: The impact of group membership on household formal savings in rural Vietnam Welfare Dynamics in Rural Vietnam, 2006 to 2010 13
Studies Based on 2012 Data "Social and Economic Determinants of Happiness in Rural Vietnam Migration and the Impact of Remittances and Information on Sending Households: Evidence from Vietnam Diversity Among Rapid Transformation: Welfare Dynamics in Rural Vietnam, 2006 to 2012 Non-farm Income, Diversification, and Income: Evidence from Rural Vietnam 14
Broad Picture of Findings Firstly, agriculture remains very essential 15
Agriculture Remains Essential Most income for most rural households continues to come from farming and livestock, especially in North and Central Highlands (the six) Commercial sales of agriculture/livestock have grown in importance from 2008 to 2012 Policies to support the commercialisation of agriculture and link these households with markets still key to continued progress 16
Agriculture Remains Essential (cont.) Big provincial differences: commercialisation is chronically low in Lao Cai, Lai Chau, and Dien Bien...enabling markets for land sale and rental and land titling may increase land-related investment, which has decreased from 2012 to 2010 25% of surveyed households still benefit from Common Property Resources Implementing environmental policies is a short-term challenge Only 8% of CPR activities reported to be regulated 17
Secondly, progress is strikingly unequal 18
Unequal Progress The average story told by several rounds of the VARHS is positive Unambiguous and meaningful increase in overall prosperity according to different measures Implication for policy: Vietnam is in general on the right track But the average story is not the whole story Inequality in many dimensions of welfare still widespread In some cases, it is increasing (wage employment and in agriculture) The VARHS is useful because we can see variation at the HH level beyond the averages (and medians) This informs policy and is uniquely able to study changes over time 19
Unequal Progress (continued) Households are generally better educated and have access to more resources However, those resources are unequally shared Lao Cai, Lai Chau, and Dien Bien lag behind in many important factors such as poverty In Lao Cai, food security has dropped significantly from 2010 to 2012 20
Finally, fundamental changes in wealth and income underway 21
Wealth Creation is Changing Agriculture is still very relevant to how households earn income and accumulate wealth......but income from household enterprises and wages is quickly becoming important (but in different ways and importance) Average household in surveyed provinces earns between 15% (Lam Dong) and 40% (Khanh Hoa) of net household income from wages 22
Wealth Creation is Changing (continued) Household enterprises contribute less to household income than agriculture (12.5% compared to 32.3% from agriculture) but are increasingly important and absorbing larger shares of household labour Richer households are much less likely to own agricultural land, graduating from agriculture Migration / remittances is an important feature of the rural economy: between 8% (Lao Cai) and 46% (Nghe An) report a migrant 23
Main Points There is a large, persistent gap in development outcomes between upland and lowland areas And between members of the Kinh majority and other ethnic groups Households in upland areas are significantly poorer Have much lower access to essential goods such as safe drinking water and high-quality housing Much less connected to markets for land, labour, and agricultural inputs These gaps should be a major focus for policymakers 24
Policy Relevant? 25
From Evidence to Policy Economic development in upland areas should be a priority Big payoffs to investments in human capital and physical infrastructure Enabling upland communities to add value through access to markets for goods, labour, and capital Internal migration may be one part of the answer 26
Increasingly Successful Mainstreaming Legislation Decree no. 61 on Investment Decision 80, Ministry of Agriculture and Rural Development New Rural Village Program Resolution N0 26/NQ-TW on agriculture Land Law revision Decree 41/2010/ NĐ-CP Details Revised legislation on agricultural investment Updating legislation governing contract farming arrangements Provided references for the implementation process of the cross- Government, national New Rural Village Program pilot program Introduced new national target programs in agricultural training, climate change adaptation, and construction of New Rural Areas Providing data and research results on land fragmentation as an input to the ongoing revision of Vietnam's Land Law Introducing credit schemes targeting small-scale agricultural producers 27
Increasingly Successful Mainstreaming Risk Analysis: Risks Racing Vietnamese Agricultural Households Situation Analysis and Proposed Solutions for Improving Efficiency of Vietnamese Agricultural Extension Services Promoting the Concentration of Land for Agricultural Production National Strategy for Agriculture and Rural Development to 2020, vision to 2030 Policy evaluation of pro-poor programs 30a, 80a, and National Program for Poverty Reduction Government research programs Take advantage of VARHS data on reported income shocks to rural households used to study risks facing small-scale family-run farms Uses reported uptake of Government extension services used to study their efficacy and penetration Exploits VARHS plot-level data used to study land concentration and agricultural income Uses survey data and analysis on rural residents livelihood, agricultural productivity, welfare and household risk assessments References VARHS data on coverage of, and households' access to, programs for 'Speedy and Sustainable Poverty Reduction' Assessment of the impacts of financial crisis on rural households in 2009 VARHS panel data used for comparisons of households across time in conjunction with aggregated Commune- and Province-level data from the VARHS and other sources 28
Increasingly Successful Mainstreaming Government Publications Ministry of Planning and Investment: State Management of District Peoples Committees: Problems and Solutions Ministry of Agriculture and Rural Development: Social impacts of implementing the land law 2003 on rural households Enhancing the market participatory of farmer through agricultural value chain The roles and solutions to strengthening the cooperative s role to its members in rural area of Viet Nam Ministry of Labor, Invalids, and Social Affairs: Labor demand in emerging processing industries (also published by ILO as Labor and Social Trends in Vietnam ) A thematic report on access to household access to resources from a gender-based perspective 29
What s Next? VARHS will continue to monitor development in rural Vietnam and collect much-needed data VARHS will continue to collaborate on rigorous research and build capacity to do this research Policies based on economic evidence can improve outcomes for rural households Especially in areas that have not yet shared equally in Vietnam s exceptional growth 30
What s Next? (Continued) A further round of VARHS in 2014 would be very desirable Appears that this is feasible However: support for research and policy outputs remains a priority Mainstreaming research is an area where donor partners can be proactive CAP/IPSARD provides several examples of how successful this approach has been so far 31
On-Going Research: Example from the VARHS 2012 round 32
Economic Development and Subjective Well-Being An in-depth study based on VARHS 2012
Introduction Aim: Understand how the many dimensions of economic development affect happiness/life satisfaction in rural Vietnam Specific focus on the effects of income and occupation, but other factors also considered. Standard economic analyses focus on objective measures of welfare, such as consumption or fulfillment of basic needs. We supplement these analyses by using a subjective measure of welfare, namely self-reported life satisfaction, or happiness.
Introduction Happiness has recently attracted a lot of attention from economists, but most studies focus on Western countries. This is the first, systematic study of happiness in Vietnam.
Literature on happiness Effect of income on happiness is controversial Deaton (2008) finds that life satisfaction is higher in rich countries than in poor. On the other hand, Easterlin (1974, 2003) and Layard (2006) point out that although income has increased strongly in e.g. the U.S., selfreported happiness is stable. => indicates that relative rather than absolute levels of income matter.
Literature on happiness Effects of occupation have been studied less Exception: many studies find negative effect of unemployment. We study effects of wage work vs. self-employment in agriculture or non-agriculture. Expected effect ambiguous: Wage work reduces uncertainty but also leads to loss of autonomy and possibly status. Other studies find that health, age, marital status, shocks and social capital are also important determinants of happiness/life satisfaction.
Data VARHS 2012: 2,740 observations. Drawback: only one respondent per household, typically the household head. Advantages: - Cluster-sampling => we can estimate average income in local neighborhood => we can distinguish between effects of absolute and relative income. - Previous survey rounds => we can estimate effect of changes vs. levels of income. - Very information-rich survey => many potential determinants of happiness can be investigated.
Happiness "Taking all things together, would you say you are.." Not at all pleased with your life 6% Very pleased with your life 7% Not very pleased with your life 42% Rather pleased with your life 45%
0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 Happiness and income 1 2 3 4 5 6 7 8 9 10 Income decile Share "very" or "rather" pleased with life
Happiness and main occupation 0.65 Share "very" or "rather" pleased with life 0.6 0.55 0.5 0.45 0.4 0.35 0.3 Own farm Wage labor HH enterprise CPR collection No occupation
Happiness and children Share "very" or "rather" pleased with life 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0 1 2 3 4 or more Number of children below 15 in hh
Happiness and marital status 0.6 Share "very" or "rather" pleased with life 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 Married Single, never married Widowed Divorced/separated
0.6 0.55 0.5 0.45 0.4 0.35 0.3 Happiness and ethnicity Kinh Non-Kinh Share "very" or "rather" pleased with life
Happiness and formal networks Share "very" or "rather" pleased with life 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 Communist Party Mass organization Other formal group No formal groups Respondent member of...
Happiness and informal networks Share "very" or "rather" pleased with life 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0-9 10-19 20 or more Number of weddings attended in other households
Happiness and shocks Share "very" or "rather" pleased with life 0.6 0.55 0.5 0.45 0.4 0.35 0.3 No shocks Natural disaster Pest infection, crop disease or avian flu Economic shock (price change, unemployment, investment failure or land loss) Household hit by Death, serious illness or injury Other shock
Table 1: Determinants of happiness Dependent variable: Happiness (four categories) Income per hh member, log 0.433** 0.392*** 0.476*** 0.474*** [0.038] [0.040] [0.048] [0.077] Median commune income per capita, log -0.125* -0.099-0.141* -0.341** << [0.069] [0.073] [0.077] [0.168] Change in log income per cap, 2010-2012 -0.156*** -0.192*** [0.036] [0.058] Main occupation Wage worker -0.119** -0.323*** -0.367*** -0.345*** [0.052] [0.060] [0.067] [0.096] Non-farm enterprise 0.082-0.217*** -0.292*** -0.321*** [0.075] [0.083] [0.092] [0.119] CPR collection -0.186-0.186-0.220-0.219 [0.179] [0.167] [0.204] [0.317] None 0.017-0.096-0.111-0.086 [0.072] [0.079] [0.086] [0.130] Control variables No No Yes Yes Yes
Table 1: Determinants of happiness, continued Landless 0.095 0.064 0.124 [0.084] [0.105] [0.153] Female -0.031-0.066 0.009 [0.073] [0.081] [0.123] Age in year -0.030*** -0.036*** -0.019 [0.011] [0.013] [0.021] Age squared/1000 0.338*** 0.394*** 0.235 [0.098] [0.120] [0.188] Years of schooling, ln(x+1) 0.112*** 0.121*** 0.252*** [0.042] [0.044] [0.095] Children below 15, ln(x+1) 0.005 0.079 0.113 [0.050] [0.057] [0.091] Marital status Never married -0.199-0.248* -0.363* [0.123] [0.144] [0.197] Widowed -0.265*** -0.255*** -0.282** [0.087] [0.094] [0.134] Divorced or separated -0.623*** -0.653*** -0.457 [0.207] [0.246] [0.440] Kinh -0.084 0.003 0.426** [0.112] [0.118] [0.195]
Table 1 : Determinants of happiness, continued Member of Communist Party 0.528*** 0.487*** 0.576*** [0.099] [0.117] [0.214] Member of Mass Organization 0.174*** 0.167*** 0.147 [0.057] [0.064] [0.119] Member of group other than party, mass org 0.148* 0.098 0.121 [0.077] [0.083] [0.121] Weddings attended in other hh, log(x+1) 0.138*** 0.117** 0.078 [0.040] [0.046] [0.073] Shocks to hh in last two years Natural disaster 0.000 0.022-0.170 [0.084] [0.089] [0.160] Pest infection, crop disease or avian flu -0.064-0.057-0.060 [0.057] [0.062] [0.102] Economic (unemployment, loss of land etc.) -0.235** -0.242** -0.403** [0.097] [0.105] [0.185] Illness -0.340*** -0.344*** -0.207 [0.083] [0.092] [0.141] Other shock 0.078-0.048-0.111 [0.209] [0.249] [0.521]
Table 1 : Determinants of happiness, continued Days unable to work due to illness in last year, log(x+1) -0.060*** -0.056*** -0.003 [0.019] [0.021] [0.032] Head born in commune 0.109* 0.141** 0.130 [0.059] [0.064] [0.106] Hh member migrated -0.068-0.065-0.125 [0.072] [0.077] [0.129] Hh member migrated 0.149* 0.157* 0.234* [0.079] [0.085] [0.135] Hh head -0.051-0.08-0.136 [0.088] [0.098] [0.135] Province dummies Yes Yes Yes Yes Yes Observations 2,594 2,680 2,534 2,058 858 Note: Ordered probit regressions. Standard errors adjusted for clustering at the village level. Regression 4 and 5 include only hh interviewed in both 2010 and 2012. Regression 5 includes only observations from communes with at least 10 observations.* significant at 10%; ** significant at 5%; *** significant at 1%
Table 2: Occupation and happiness Dependent variable: Happiness (four categories) Age<49 Age>=49 Main occupation Wage worker -0.254*** -0.399*** [0.085] [0.106] Unskilled wage worker -0.324*** [0.072] Skilled wage worker -0.321*** [0.084] Private sector wage worker -0.361*** [0.064] Public sector wage worker -0.225* [0.118] SOE wage worker -0.182 [0.250] Wage worker in: Agriculture -0.304** [0.135] Mining -0.484 [0.341] Manufacturing -0.361*** [0.123] Construction -0.285*** [0.093] Services -0.341*** [0.091] Province dummies Yes Yes Yes Yes Yes Control variables As in Tab. As in Tab. As in Tab. As in Tab. As in Tab.
Conclusions Income a strong determinant of happiness. But relative income may be more important than absolute. Controlling for level of income, recent income growth has a negative effect on happiness. Holding income constant, workers on own farms are more happy than wage workers and non-farm enterprise operators. Party members are much happier than others. Age, marital status, networks and shocks are also important determinants of happiness.
Conclusions Results are remarkably similar to those from rich, Western countries, for example for income, age, health, schooling, marital status and social networks. Implication: The values of farmers in rural Vietnam are not radically different from those of people in the streets of Copenhagen or New York. Our core values are not Western or Eastern, traditional or modern, but universal.
Thank you! ILSSA for data collection and collaboration on the draft report CIEM and CAP-IPSARD for joint research Danida and the Embassy of Denmark in Viet Nam for support Participants, commentators and numerous colleagues 55