Asian Barometer Survey Wave 4 2014-2016 TECHNICAL REPORT ()

By

Taylor Nelson Sofres Vietnam for Asian Barometer Survey Center for East Asia Democratic Studies National University

November 2015

Contact Information Taylor Nelson Sofres Vietnam 203 Minh Khai street, Hai Ba Trung District, Hanoi Tel: 84-4-39877030 Fax: 84-4-39877037 Email: [email protected]

Asian Barometer Survey No.1, Sec. 4, Roosevelt Road, Taipei 10617, Taiwan Center for East Asia Democratic Studies, College of Social Sciences National Taiwan University Tel: 886-2-3366-8456 Fax: 886-2-2365-7179 Email: [email protected]

1. BASIC INFORMATION

1.1 LOCATION The Asian Barometer 2015 Survey covered the entire Viet Nam, and had 6 major study areas: Northern Midlands and Mountains, Red River Delta, North and South Central Coast, Central Highlands, Southeast, Mekong River Delta.

1.2 POPULATION With an estimated 85.8 million inhabitants (Census 2009), it is the world’s 13th – most – populous country, and the eight-most-populous country in Asia. According to census data, during the last decade the proportion of people living in urban areas increased from 23.7% in 1999 to 29.6% in 2009. This translates into a total of 25.4 million urban residents out of a national population of 85.8 million in 2009. The main increase in urbanization took place in the already large urban areas with 200,000 residents or more. The total number of these urban centres has increased from 9 in 1999 to 15 in 2009.

1.3 GOVERNMENT The Socialist Republic of Vietnam is a single-party state. Its current state constitution, which replaced the 1975 constitution in April 1992, asserts the central role of the Communist Party of Vietnam in all organs of government, politics and society. The General Secretary of the Communist Party performs numerous key administrative and executive functions, controlling the party’s national organization and state appointments, as well as setting policy. Only political organizations affiliated with or endorsed by the Communist Party are permitted to contest elections in Vietnam. These include the Vietnamese Fatherland Front and worker and trade unionist parties. Although the state remains officially committed to socialism as its defining creed, its economic policies have grown increasingly capitalist, with The Economist characterizing its leadership as “ardently capitalist communists”.

1.4 ECONOMIC PERFORMANCE The economy of Vietnam is a developing planned and market economy. Since the mid-1980s, through the "Đổi Mới" reforms, Vietnam has made a shift from a highly centralized planned economy to a socialist-oriented market economy which uses both directive and indicative planning. Over that period, the economy has experienced rapid growth. Vietnam’s economic growth has been among the highest in the world since 2000, and according to Citigroup, such high growth is set to continue. Vietnam has the highest Global Growth Generators Index among 11 major economies. According to the International Monetary Fund (IMF) in 2012, the nominal GDP reached $138 billion, with nominal GDP per capita of $1527. According to a forecast in December 2005 by Goldman Sachs, Vietnamese economy will become the World’s 21st largest economy in the world with nominal GDP of $ 436 billion and nominal GDP per capita of 4,357 USD by 2025. According to a forecast by the PricewaterhouseCoopers in 2008, Vietnam may be the fastest growing of emerging economies by 2025, with a potential annual growth rate of about 10% in real dollar terms, which would increase the size of the economy to 70% of the size of the UK economy and would surpass the GDP of countries such as Norway, and Portugal by 2050.

Nowadays, Vietnam is in a period of being integrated into the global economy. Almost all Vietnamese enterprises are SMEs. Vietnam’s successful economic reforms resulted in it joining the World Trade Organization in 2007. Vietnam has become a leading agricultural exporter and served as an attractive destination for foreign investment in Southeast Asia. However, the country still suffers from relatively high levels of income inequality, disparities in healthcare provision, and poor gender equality.

The impressive growth has also thrown up issues such as rapid urbanization and environmental degradation in Vietnam. Achieving sustainable development with attention to ecological conservation presents a great challenge for Vietnam’s policy-makers, but many efforts have been made in the last decade.

It is believed that in spite of several laws aimed at environmental protection and sustainable growth, environmental management under these laws has proved rather ineffective. There is a lack of public awareness for the need for environmental protection. Conservation should begin at home, but in many areas, the people’s day-to-day activities unintentionally serve to destroy their natural environment.1

1.5 IMPORTANT POLITICAL AND SOCIAL EVENTS Vietnam has entered territorial disputes with in the South China Sea in 2014 after China set up an oil rig in an area to which both nations have claimed.

On 5 October 2015 Vietnam successfully concluded Trans- Pacific Partnership (TPP) negotiations with 11 countries which account for 40 % of global GDP.

1http://asiasociety.org/policy/environment/climate-change-and-energy/real-cost-developing-vietnam accessed on 6th February 2015 2. TIMETABLE

Activity Duration Preparation Questionnaire translation: Translate English 2015/09/8 – 15 Core Questionnaire into Vietnamese Pretest the questionnaire 2015/10/2 – 6 Pre-fieldwork Preparations: Logistics, 2015/10/8 – 19 Script the questionnaire Training for supervisors and interviewers 2015/10/19 Fieldwork Northern Midlands and Mountains 2015/10/27 – 11/10 Red River Delta 2015/10/20 – 11/5 North and South Central Coast 2015/10/24 – 11/5 Central Highlands 2015/10/23 – 11/1 Southeast 2015/10/20 – 11/5 Mekong River Delta 2015/10/23 – 11/1 Data Processing Data cleaning 2015/11/12 – 23 Submission of Vietnam data for checking 2015/12/1

3. RESPONDENTS

3.1 CRITERIA OF SELECTING RESPONDENTS Only one eligible person (a man or a woman) above the age of 18 years was interviewed from each sampled household, quotas or any other selection criteria (including gender quota) was not used to screen respondents. However, post stratification weights were applied if necessary during the data analysis stage.

In case more than one eligible respondent was present in the sampled household, the person to be interviewed was selected using KISH Grid method.

3.2 RESEARCH ETHNICAL REVIEW In Vietnam there is no need to obtain any permission for this type of research. For the ABS IV we did not need to go through the review process in Viet Nam. However, before the beginning of interview process, the interviewers informed all interviewees that ‘We will not record your name or any identifying information about you on the questionnaire. So no one will know who gave which answers to our questions. There is no risk to you in participating in the survey. There is also no benefit to you in participating in the survey. But if you answer our questions, you will help us understand how the public feels about issues facing the country today. Your participation in the survey is voluntary. If there are any questions you don't want to answer you don't have to answer them. And you can stop participating in the survey at any time’.

4. SAMPLING PROCESS AND METHODOLOGY

4.1 SAMPLING SIZES AND ERROR MARGINS The target population for the survey was people aged 18 years and above. The table below summarizes the population distribution by age group according to Vietnam Census 2009.

Age Population Proportion of total population 0-17 26,402,057 30.6 18-24 12,236,909 14.2 25-34 14,751,222 17.1 35-44 12,578,901 14.6 45-54 9,923,869 11.5 55-64 4,951,194 5.7 ≥ 65 5,545,016 6.4 Total 86,389,168 100.0

The population distribution in Vietnam is as mentioned:

Strata Vietnam population2 Proportion Urban 25,450,254 29.5 Rural 60,938,914 70.5 Total 86,389,168 100.0

As discussed, Vietnam is divided into 6 main geographic regions.3 We proposed to spread our sample across all the six geographic regions for a sample representative of

2 Census 2009 3 Census 2009 the universe. The distribution of provinces and population of the six major regions is as mentioned:

No. of No Geographic Region Population Provinces 1 Northern Midlands and Mountains 14 11,595,761 2 Red River Delta 11 19,584,287 3 North and South Central Coast 14 18,835,154 4 Central Highlands 5 5,115,135 5 Southeast 6 14,067,361 6 Mekong River Delta 13 17,191,470

Once the sample size is proportionately distributed among the six major geographic regions, the distributed sample size was distributed albeit proportionately by urbanity within each region. Hence, broadly there were twelve strata. Proportionate distribution by urbanity within each region is as below:

Geographic Number of Total Propor Total Proport Total Region provinces Urban tion Rural ion North Highlands & 14 11,595,761 1,822,021 16% 9,733,740 84% Mountains Red River 11 19,584,287 5,729,060 29% 13,855,227 71% Delta North and South 14 18,835,154 4,513,956 24% 14,321,198 76% Central Coast Central 5 5,115,135 1,416,122 28% 3,699,013 72% Highlands South East 6 14,067,361 8,043,806 57% 6,023,555 43% Mekong 13 17,191,470 3,925,289 23% 13,266,181 77% River Delta 63 86,389,168 25,450,254 29.5% 60,938,914 70.5%

The proportionate distribution would result in 354 interviews to be covered from the urban areas and 846 interviews from the rural areas. Thus, the proportionate distribution of the sample size was:

Total Proport Urban Proporti Rural Proporti Socio-Economic Populatio ionate Popula onate SS Popula onate SS Region n SS tion % Urban tion % Rural North Highlands 11,595,761 161 16% 25 84% 135 & Mountains Red River Delta 19,584,287 272 29% 80 71% 192 North and South 18,835,154 262 24% 63 76% 199 Central Coast Central 5,115,135 71 28% 20 72% 51 Highlands South East 14,067,361 195 57% 112 43% 83 Mekong River 17,191,470 239 23% 55 77% 184 Delta 86,389,168 1,200 29.5% 354 70.5% 846

Figure 1: Proposed design & Provincial spread

4.2 SAMPLING SCHEME

4.2.1 FIRST STAGE-SELECTION OF PSU PPS was applied at the region level to select ward/commune while taking into account the number of urban and rural PSUs assigned to each region. To do that, in each region, the complete lists of all wards and communes were compiled separately by descending order of the population.

The below tables present all selected PSUs for each region. In the northern midlands and mountains region, there are 13 rural PSUs and 3 urban PSUs covering the sample size of 160.

Northern Midlands and Mountains - PSU - URBAN (3) Province District Ward/Commune Urbanity Population Sơn La Huyện Mường La Thị trấn ít Ong Urban 16,878 Phú Thọ Thành phố Việt Trì Phường Thanh Urban 9,911 Miếu Phú Thọ Thị xã Phú Thọ Phường Hùng Urban 5,865 Vương

Northern Midlands and Mountains - PSU – RURAL (13) Province District Ward/Commune Urbanity Population Bắc Kạn Huyện Chợ Mới Xã Nông Thịnh Rural 1,687 Hà Giang Huyện Đồng Văn Xã Sủng Trái Rural 5,349 Huyện Bắc Quang Xã Bằng Hành Rural 4,389 Huyện Yên Minh Xã Đông Minh Rural 2,278 Phú Thọ Huyện Cẩm Khê Xã Phú Lạc Rural 3,582 Huyện Lâm Thao Xã Cao Xá Rural 9,154 Huyện Thanh Sơn Xã Văn Miếu Rural 6,653 Huyện Thanh Sơn Xã Thục Luyện Rural 4,728 Huyện Thanh Sơn Xã Giáp Lai Rural 3,158 Sơn La Huyện Mai Sơn Xã Chiềng Chăn Rural 5,921 Huyện Mộc Châu Xã Tân Xuân Rural 3,939 Huyện Sông Mã Xã Pú Pẩu Rural 2,791 Huyện Thuận Châu Xã Bó Mười Rural 7,592

In the Red River region, there are 19 rural PSUs and 8 urban PSU representing the sample size of 270.

RED RIVER - URBAN - PSU (8) Province District Ward/Commune Urbanity Population Hà Nội Quận Cầu Giấy Phường Nghĩa Đô Urban 28,929 Quận Ba Đình Phường Thành Urban 23,049 Công Thành phố Hà Phường Văn Quán Urban 19,932 Đông Huyện Mê Linh Thị Trấn Quang Urban 16,112 Minh Quận Long Biên Phường Phúc Đồng Urban 10,500 Hải Phòng Quận Ngô Quyền Phường Lê Lợi Urban 8,088 Huyện Thuỷ Thị trấn Núi Đèo Urban 4,260 Nguyên Thái Bình Thành phố Thái Phường Quang Urban 13,533 Bình Trung

RED RIVER - RURAL -PSU (19) Province District Ward/Commune Urbanity Population Hà Nội Huyện Từ Liêm Xã Trung Văn Rural 22,225 Huyện Đông Anh Xã Tiên Dương Rural 15,317 Huyện Mỹ Đức Xã Hợp Thanh Rural 11,789 Huyện Gia Lâm Xã Dương Rural 10,658 Quang Huyện Đông Anh Xã Đông Hội Rural 10,006 Huyện Thạch Thất Xã Thạch Hoà Rural 8,570 Huyện Thường Tín Xã Tân Minh Rural 7,615 Huyện Thường Tín Xã Lê Lợi Rural 7,179 Huyện Phúc Thọ Xã Vân Nam Rural 5,935 Huyện Ứng Hòa Xã Minh Đức Rural 4,990 Hải Phòng Huyện Thuỷ Nguyên Xã An Lư Rural 12,786 Huyện Tiên Lãng Xã Tiên Tiến Rural 2,713 Hưng Yên Huyện Phù Cừ Xã Minh Hoàng Rural 4,420 Thái Bình Huyện Kiến Xương Xã Quang Bình Rural 9,562 Huyện Tiền Hải Xã Bắc Hải Rural 8,053 Huyện Hưng Hà Xã Chi Lăng Rural 6,722 Huyện Đông Hưng Xã Đông Tân Rural 6,378 Huyện Hưng Hà Xã Hồng Lĩnh Rural 5,547

In the central coast region, 20 rural PSU and 6 urban PSU to cover the sample size of 260.

CENTRAL COAST-URBAN (6) Province District Ward/Commune Urbanity Population Bình Định Thành phố Qui Phường Ngô Mây Urban 22,690 Nhơn Thành phố Qui Phường Nhơn Bình Urban 18,387 Nhơn Hà Tĩnh Thị xã Hồng Lĩnh Phường Bắc Hồng Urban 8,036 Khánh Hòa Thành phố Nha Phường Tân Lập Urban 13,213 Trang Thành phố Nha Phường Xương Urban 10,778 Trang Huân Quảng Bình Thành Phố Đồng Phường Hải Thành Urban 5,079 Hới

CENTRAL COAST - RURAL-PSU (20) Province District Ward/Commune Urbanity Population Bình Định Huyện Hoài Nhơn Xã Tam Quan Bắc Rural 17,210 Huyện An Nhơn Xã Nhơn Khánh Rural 8,530 Huyện Phù Mỹ Xã Mỹ Lợi Rural 7,322 Hà Tĩnh Thành phố Hà Tĩnh Xã Thạch Trung Rural 7,881 Huyện Kỳ Anh Xã Kỳ Tiến Rural 5,711 Huyện Đức Thọ Xã Đức Thịnh Rural 3,673 Khánh Hòa Huyện Cam Lâm Xã Cam Thành Rural 13,333 Bắc Huyện Ninh Hòa Xã Ninh Thủy Rural 11,330 Thị xã Cam Ranh Xã Cam Thành Rural 4,504 Nam Quảng Bình Huyện Quảng Trạch Xã Quảng Tiên Rural 5,385 Thanh Hóa Huyện Hoằng Hóa Xã Hoằng Phụ Rural 10,177 Huyện Triệu Sơn Xã Dân Quyền Rural 9,045 Huyện Hà Trung Xã Hà Vinh Rural 6,912 Huyện Quảng Xã Quảng Chính Rural 6,387 Xương Huyện Thọ Xuân Xã Xuân Phú Rural 6,059 Huyện Vĩnh Lộc Xã Vĩnh Hưng Rural 5,119 Huyện Quảng Xã Quảng Văn Rural 4,827 Xương Huyện Nông Cống Xã Trung Thành Rural 4,086 Huyện Hậu Lộc Xã Hòa Lộc Rural 10,381 Huyện Như Xuân Xã Cát Tân Rural 2,498

In the central highlands and southeast region, 13 rural PSUs and 13 urban PSU were selected to cover the sample size of 260.

CENTRAL HIGHLANDS - URBAN-PSU (13) Province District Ward/Commune Urbanity Population Đắk Lắk Huyện Krông Pắc Thị trấn Phước An Urban 17,400 Đồng Nai Thành phố Biên Hòa Phường Long Bình Urban 75,894 Thành phố Biên Hòa Phường Bửu Long Urban 25,765 Hồ Chí Minh Quận Bình Tân Phường Tân Tạo A Urban 52,553 Quận 12 Phường Tân Chánh Urban 45,519 Hiệp Quận Tân Phú Phường Phú Trung Urban 40,014 Quận Bình Thạnh Phường 26 Urban 33,166 Quận Thủ Đức Phường Linh Chiểu Urban 28,882 Quận 3 Phường 11 Urban 23,294 Quận 1 Phường Phạm Ngũ Urban 19,699 Lão Quận Phú Nhuận Phường 11 Urban 9,640 Kon Tum Thị xã Kon Tum Phường Duy Tân Urban 12,777 Lâm Đồng Thành phố Đà Lạt Phường 10 Urban 15,123

CENTRAL HIGHLANDS - RURAL-PSU (13) Province District Ward/Commune Urbanity Population Đắk Lắk Huyện Krông Bông Xã Cư Pui Rural 11,139 Huyện Krông Năng Xã Ea Tân Rural 9,522 Đồng Nai Huyện Long Thành Xã An Phước Rural 22,422 Huyện Long Thành Xã Long An Rural 14,217 Huyện Định Quán Xã Phú Túc Rural 12,232 Huyện Cẩm Mỹ Xã Thừa Đức Rural 8,134 Huyện Tân Phú Xã Phú Lập Rural 6,693 Hồ Chí Minh Huyện Bình Chánh Xã Vĩnh Lộc A Rural 49,946 Huyện Bình Chánh Xã Lê Minh Xuân Rural 31,348 Huyện Hóc Môn Xã Tân Xuân Rural 18,678 Kon Tum Thị xã Kon Tum Xã Vinh Quang Rural 10,384 Huyện Ngọc Hồi Xã Đắk Xú Rural 5,018 Huyện Sa Thầy Xã Sa Nhơn Rural 2,999 Lâm Đồng Thị xã Bảo Lộc Xã Lộc Châu Rural 16,273

In the Mekong river region, there are 18 rural PSUs and 6 urban PSUs to cover the sample size of 240.

SOUTHEAST-URBAN-PSU (6) Province District Ward/Commune Urbanity Population Bến Tre Thị xã Bến Tre Phường Phú Urban 17,367 Khương Thị xã Bến Tre Phường 6 Urban 8,132 Đồng Tháp Thành phố Cao Phường 6 Urban 21,894 Lãnh Huyện Tam Nông Thị trấn Tràm Chim Urban 10,167 Kiên Giang Huyện Kiên Lương Thị trấn Kiên Lương Urban 32,837 Vĩnh Long Thị xã Vĩnh Long Phường 8 Urban 13,583

SOUTHEAST –RURAL-PSU (18) Province District Ward/Commune Urbanity Population Bến Tre Huyện Giồng Xã Bình Hoà Rural 9,077 Trôm Huyện Giồng Xã Châu Bình Rural 7,926 Trôm Huyện Chợ Lách Xã Vĩnh Hòa Rural 6,517 Huyện Giồng Xã Hưng Phong Rural 5,066 Trôm Đồng Tháp Huyện Châu Xã Tân Phú Trung Rural 18,244 Thành Huyện Lấp Vò Xã Bình Thành Rural 16,783 Huyện Lấp Vò Xã Mỹ An Hưng B Rural 15,728 Huyện Hồng Ngự Xã Thường Phước 2 Rural 10,128 Hậu Giang Huyện Phụng Xã Tân Bình Rural 20,086 Hiệp Thị xã Ngã Bảy Xã Đại Thành Rural 11,074 Huyện Châu Xã Thạnh Xuân Rural 10,624 Thành A Kiên Giang Huyện Châu Xã Mong Thọ B Rural 14,748 Thành Huyện Tân Hiệp Xã Tân Hội Rural 13,700 Huyện Giồng Xã Bàn Tân Định Rural 12,581 Riềng Huyện Gò Quao Xã Thủy Liễu Rural 11,758 Huyện An Minh Xã Đông Hưng Rural 9,556 Vĩnh Long Huyện Vũng Xã Quới An Rural 8,637 Liêm Thị xã Vĩnh Long Xã Trường An Rural 7,413

4.2.2 SECOND STAGE-SELECTION OF SSU In each selected PSU (ward in urban / commune in rural), 2 sampling points were selected randomly by the field supervisors. Residential areas within the sampled wards and villages within the sampled communes were the sampling points. We completed 5 interviews from each sampling point.

4.2.3 THIRD STAGE-SELECTION OF HOUSEHOLDS Taking into account that the full and updated list of households was not available, the starting points were selected based on some fixed landmarks such as the ward/commune people committee’s building; ward/commune health station’s building; the starting point of the main road etc.

Multi-story Buildings: Since the PSU’s in the urban areas were wards and the sample to be covered from a ward in any urban area was not more than 10, only one interview was conducted in any multi-story building (if they get selected into the sample) to ensure better spread of the sample.

Direction: From the starting point onwards, interviewers followed the Right Hand Rule to choose the households for interviewing.

There were no more than 5 interviews per sampling point. Hence the total number of sampling points to be sampled was 240. Skip patterns: Skip 2 households after every successful interview (Interval =3). o In some rural villages especially in the mountains the households were spread out. In such cases we followed an interval of 2. o In case of refusal or nobody was eligible at the selected household interviewers went the next household. o The interviewers were not allowed to change the person selected by using the Kish Grid. If the selected person refused to participate, the interviewers went to the next household following the skip pattern. o Two attempts were made in total to complete an interview within the selected respondent before substituting another household. In case of a refusal or nobody eligible, interviewers went to the next household. If nobody was at home or the respondent was not available at the time of the visit, interviewers re-visited once before going to the next household. o This process continued until the required numbers of respondents were interviewed in the sampled PSU.

Sampling Points to be No. of PSUs to be Sampled Geographic Region Sampled # # Urban # Rural Residential # Villages Wards Communes Areas North Highlands & 3 13 6 26 Mountains Red River Delta 8 19 16 38 North, South, Central 7 20 14 40 Coast Central Highlands & 13 13 26 26 Southeast Mekong River Delta 5 19 10 38 Sub - Total 36 84 72 168 Total PSUs & SPs 120 240

Figure 2: Example of skip patterns and right hand rule

SP1

Ward A HH1 Street 1

Street 2

HH2

Street 3

4.2.4 FOURTH STAGE-SELECTION OF RESPONDENT Only one eligible person (a man or a woman) above the age of 18 years was interviewed from each sampled household, quotas or any other selection criteria (including gender quota) was not used to screen respondents. However, post stratification weights were applied if necessary during the data analysis stage. In case more than one eligible respondent was present in the sampled household, the person to be interviewed was selected using KISH Grid method.

Step 1: Interviewers listed down all the eligible members who were in the target age group in the sampled households

Step 2: Horizontal line was drawn from left to right across the last person who was in the target group. Doing this ensures that the random number sequence to be used was up to the number of eligible respondents. In the example below it was 5.

Step 3: Then a vertical line was drawn in the column that matches with the last digit of the questionnaire serial number. Since serial numbers of questionnaires were numbered sequentially (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11… etc.) this ensured that all the vertical columns had an equal probability of being sampled. Suppose the serial number of the questionnaire was 2013. The last digit was 3. Hence vertical line was drawn as below.

Step 4: The number at the intersection of the two lines was the serial number of the sampled person to be interviewed for the survey. This was 2 in the example in the table. Hence the interviewer had to interview the person who was listed as No. 2 and no one else in the household. This ensured randomness of selection within household.

The example of Kish Grid table

The last digit of household ID Name of eligible respondents Age 1 2 3 4 5 6 7 8 9 0 1. 1 1 1 1 1 1 1 1 1 1 2. 2 1 2 1 2 1 2 1 2 1 3. 2 3 1 2 3 1 2 3 1 2 4. 3 4 1 2 3 4 1 2 3 4 5. 5 1 2 3 4 5 1 2 3 4 6. 5 6 1 2 3 4 5 6 1 2 7. 3 4 5 6 7 1 2 3 4 5 8. 6 7 8 1 2 3 4 5 6 7 9. 8 9 1 2 3 4 5 6 7 8 More than 10 9 10 1 2 3 4 5 6 7 8

If the selected eligible respondent was not in the household after 2 re-visits, such households were replaced through the same interval that was followed. This process continued until the required numbers of respondents are interviewed in the sampled PSU.

4.3 NUMBER OF CALLS AND SUBSTITUTION/ALTERNATE SAMPLES In the initial sampling list, 120 selected PSUs were spread across 24 provinces. After looking closely at each PSU for their accessibility, 2 PSUs had to be replaced owing to their inaccessibility. The replacement PSUs were selected following a similar sampling strategy from those provinces which had similar socio economic characteristics. Hence, in the final execution 120 PSUs were spread across 24 provinces.

Replacements

Reason Provinc Provinc Reasons why the PSU e e For PSU Urban / No replace Replace Sample Replace Replace Selected Rural ment d d d ment selected Mountai Same nous Only socio-ec Xã province Xã Đức 1 Bac Kan PSU onomic Nông Rural ; Vân replaced characte Thịnh difficult ristics access Mountai Same nous Only socio-ec Thanh province Xã Lộc Xã Hòa 2 PSU onomic Rural Hoa ; Thịnh Lộc replaced characte difficult ristics access

4.4 SURVEY STATISTICS A total of 4905 households were contacted for the survey, of which 1200 were completed interviews.

Frequency Percent Interviewed 1200 24.5 Completed 1200 24.5 Eligible, No Interview (Non-response) Nobody is home/Unknown if HH has an eligible respondent 840 17.1 HH-level refusal 2453 50.0 Partial interview 168 3.4 HH-level language problem & health problem 19 0.4 Not Eligible 244 5.0 Total 4905 100.0

The refusal rate for this survey is 74.3%. TNS computed the refusal rate (RR) following the definition that refusal rate is one minus the number of complete interviews (I) with reporting cases divided by the number of eligible cases in the sample:

1

Where I = Complete interview P = Partial interview R = Refusal and break-off NC = Non-contact O= Physically or Mentally unable/incompetent

5 RESEARCH METHODOLOGY

5.1 PREPARATION 5.1.1 QUESTIONNAIRE The survey instrument was provided by Asian Barometer Survey. The questionnaire was translated into the local language and further would be back translated into English to assure that the translations are accurate. All measures were adopted to prevent data loss in the process of questionnaire translations.

The first translation from original English version to Vietnamese was done by a translator (from our production team) who has not seen the original English version of the questionnaire. The research team reviewed the translation and made changes if required before put in the pre-testing.

5.1.2 PRE-TESTING AND FINALIZING QUESTIONNAIRE Once the tools were approved by the client a pilot testing of the tools was carried out in Hanoi and Ho Chi Minh City to ensure appropriateness of the tools to be used for collection of data and also the response of the respondents to the survey. A total of 20 interviews were conducted in Hanoi and HCMC during the pre-test exercise from those PSUs which were not part of the main survey.

Besides the importance of pre-testing of the questionnaire during the pilots, the field test measured the following quality control indicators:

Team members understand their roles Team members understand, and correctly follow interviewing protocols

Early identification of field related problems/difficulties

5.1.3 TRAINING The training for interviewers and field supervisors and quality control supervisors took place in four key cities where TNS has its offices.

The following people were trained: o Interviewers o Field supervisors and QC supervisors

The training included: o Background and the objectives of the study o Study tools including the questionnaire o Other supporting tools such as prompt cards, show card and picture sets for questions (if any) o Field movement (Right Hand Rule, using interval to choose households, Kish Grid etc.) o Mock interviews: The training will include an interview practice session to ensure interviewers completely understand and skilled in interviewing using questionnaire and data collection tools.

5.1.3.1 SUPERVISOR The Operations Director, Fieldwork Director, Fieldwork Director and QC Director are staff of TNS. The Supervisor and QC teams had at least 2 years of working experience in TNS.

Overall FW Managers managed all field supervisors in surveyed provinces and reported to TNS P&S team and Field Director about the project’s progress. The Field Supervisors supervised closely interviewers in the field and provided them support whenever required. Besides that, quality control supervisors also worked independently in the field to ensure that all requirements related to the project had been followed by interviewers.

5.1.3.2 INTERVIEWER The selection of interviewers was carried out from the pool of talented resources who have been working with TNS for a long period of time. The recruitment of these people was done by TNS project team based on the criteria below:

Has had at least 1 year of work experience Have worked on social and opinion survey research projects

Preferably have worked on studies related to the survey topic

TNS conducted the fieldwork in adherence to our quality standards. We mobilized several fieldwork teams for data collection. Teams were recruited locally from each of the cities to ensure that the interviewers were familiar with the local dialect and sensitivities as each of these cities were in different geographical regions of Vietnam. Each team of interviewers was led by a senior field supervisor. Each team travelled to the sampled districts and remained there until field work was completed in that district.

5.2 FIELDWORK 5.2.1 WORKERS ON HAND For this project, a total of 46 interviewers were deployed. The field interviewers were from the pool of TNS interviewers.

City Number of interviewers Hanoi 17 Da Nang 6 HCMC 15 Can Tho 8 Total 46

5.2.2 SUPERVISION In the field, the quality control was ensured by field supervisor and quality control supervisor who were assigned to each region. However, due to the fact that the surveying places were quite spread out, field supervisor and quality control supervisor could not work together at the same place like other surveys where there is a much larger sample size and more concentration at several places. They had to divide their supervision work and each of them accompanied a team of interviewers which consisted of 2 or 3 interviewers.

5.2.2.1 EVALUATION OF INTERVIEWER’S WORK The quality control team used a validation form with some key questions to check the interviewers’ accuracy in coding the respondent’s answer. They also checked the length of the interview as well as the attitude of the interviewers (as feedback from the respondent). In this validation form, the rules to warn the interviewers or cancel the result were also included. It was instructed that if the rate of cancelled interview of one interviewer was too high, that interviewer will be fired. To control this rate, quality control team also used a report form in which the result of quality control process was shown clearly by each interviewer, city and in total.

5.2.2.2 SPOT-CHEKING In addition to back check and observation in the field, we randomly checked completed interviews by listen to recordings to ensure interviews were conducted in a correct manner.

5.2.2.3 RE-TESTING There was no re-testing for this survey. However, the supervisors ensured the quality of the questionnaires on spot.

5.3 FIELD EDITING The completed questionnaires, once, uploaded in the server were downloaded by TNS research team to scrutinize the results and random check the quality of certain questions. Once found out mistakes (which were mainly due to typo when entering answers to open-ended questions), the questionnaires with ID of interviewers were sent to field supervisors to back check with interviewers and correct information at data cleaning stage.

5.3.1 DATA PROCESSING Completed questionnaires in electronic format were sent to the company Server under the supervision of the Scripting team.

Related to coding process, for closed-end or pre-coded questions, TNS coding team use the codes specified on the final questionnaire for all items.

For open-end questions, TNS referred to the code frame included in the questionnaire by the ABS.

5.3.2 DATA ENTERING AND CLEANING. When scripting the questionnaire, as per the request from research team, the scripting team programmed logic-checks so that all completed questionnaires were automatically checked at those points.

During the field work, field supervisors and QC supervisors in charge can get access to the raw data by using export tools provided by Nfield. Thus, they can check whether the questions were wrongly coded or not. In case of mistakes, they will ask interviewers for revisits (for filling up of the gaps or necessary correction etc.).

After the field work, the data processing team downloaded all completed questionnaires and cleaned the data before putting it in a required format.

5.4 PROBLEMS AND RECOMMENDATIONS Open ended questions: For some respondents it was not easy for them to answer questions when there were no support tools such as show cards. For example, for the question 99, many respondents just gave a general answers such as health care, education, economic issues when being asked about “what are the most important problems facing this country that government should address” instead of giving more detailed answer even though interviewers tried to probe more.

There were some errors when interviewers typed numbers for monthly household income. Thus, we suggested that pre-codes for this question in next survey would help to minimize this issue.

Due to sensitivity of some questions in Vietnamese context, to ensure safety of interviewers when travelling to rural areas and conducting interviews with local people, we proposed to revise some questions after the pre-test in Hanoi and HCMC. These revisions were approved by the ABS representative before the implementation of the main field work.

Q# Remarks 7-19 Revise the response categories into “a great deal of trust”, “quite a lot of trust”, “some trust”, “hard to say”, and “don’t know or decline to answer”. 53-54 Deleted. Officially, respondents in Vietnam know Communist Party only and other’s if any are non-existent 76-77 Deleted. Demonstration, protest or use of force/violence for a political cause is not allowed in Vietnam. Rarely respondents dare to admit that they have done one of these actions. 98 Deleted. Direct reference to the President. 117-118 Rephrase the response categories into: “All of them are clean and not corrupt”, “most of them are clean and not corrupt”, “quite a few of them are clean and corrupt”, “hard to say”. 120 Deleted. 130-133 Deleted. Statements are too strong to administer referring to political party – elections – Parliament

Overall the questionnaire took around 60 minutes to go through. It was lengthy considering that some respondents may lose their attention at the middle and tended to stop interview if they had other commitments to deal with. Thus, interviewers were instructed to inform respondents at the beginning about the length of interview and asked them when it was most appropriate timing for the interview. Then interviewers would come back on the agreed time to avoid any interruptions.

6 RELIABILITY ANALYSIS

As part of a full review of the survey, questions were subjected to a reliability analysis. Reliability was measured using Cronbach’s alpha, α. It is a common rule of thumb that a Cronbach α value of .6 to .7 is an acceptable value. This Technical Report uses commonly accepted rule of thumb in interpreting Cronbach’s α values:

α ≥ 0.9 Excellent 0.7 ≤ α < 0.9 Good 0.6 ≤ α < 0.7 Acceptable 0.5 ≤ α < 0.6 Below Standard α < 0.5 Poor

6.2 ECONOMIC EVALUTIONS The Cronbach’s α value for the six items is 0.638, which is considered acceptable. Each of the six items is positively correlated with each of the other questionnaires in this battery. Q4 (0.315) and Q6 (0.323) have the lowest corrected item-total correlation. Deleting any of the items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α = 0.638

Corrected Item-total Cronbach's α if item Cronbach's α = 0.638 correlation deleted Q1. How would you rate the overall economic condition of our country 0.418 0.593 today? Q2. How would you describe the change in the economic condition of 0.429 0.588 our country over the last few years? Q3. What do you think will be the state of our country’s economic 0.391 0.604 condition a few years from now? Q4. As for your own family, how do you rate the economic situation of 0.315 0.629 your family today? Q5. How would you compare the current economic condition of your 0.409 0.598 family with what it was a few years ago? Q6. What do you think the economic situation of your family will be a few 0.323 0.630 years from now?

6.2 TRUST IN POLITICAL INSTITUTIONS The Cronbach’s α for the 13 items is 0.910, which is considered Excellent. Each of the 13 items is positively correlated with each of the other questions in this battery. Q17 (0.516) has the lowest corrected item-total correlation. Deleting any of the items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α = 0.910.

Corrected Item-total Cronbach's α if item Cronbach's α = 0.910 correlation deleted Q7. The Prime Minister .706 .901 Q8. The courts .707 .900 Q9. The national government [in .730 .899 capital city] Q10. Political parties [not any .597 .905 specific party] Q11. Parliament .784 .896 Q12. Civil service .590 .906 Q13. The military(or armed forces) .578 .906 Q14. The police .603 .905 Q15. Local government .579 .906 Q16. Newspapers .604 .905 Q17. Television .516 .908 Q18. Election Commission .712 .900 Q19. NGOs .572 .908

6.3 SOCIAL CAPITAL 6.3.1 FAMILY TRUST The Cronbach’s α value for the three items is 0.668, which is considered acceptable. Each of 3 items is positively correlated with each of the questions in this battery. Q28 (0.432) has the lowest corrected item-total correlation. Deleting any of the items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α = 0.668.

Corrected Item-total Cronbach's α if item Cronbach's α = 0.668 correlation deleted Q26. Your relatives .437 .627 Q27. Your neighbors .594 .432 Q28. Other people you interact with .432 .657

6.3.2 SOCIAL NETWORK The Cronbach’s α value for the three items is 0.331, which is considered unacceptable. Each of the three items is positively correlated with each of the other questions in this battery, though the correlations are low. Q32 (0.058) has the lowest corrected item-total correlation. Deleting Q32 from this battery would have a moderate, because it would increase the overall reliability from 0.311 to 0.605, which is above that minimum threshold (α = 0.6) for it to be acceptable. Deleting any of Q30 or Q31 from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.311.

Corrected Item-total Cronbach's α if Cronbach's α = 0.311 correlation item deleted Q30. If you have a difficult problem to manage, are there people outside your .267 .064 household you can ask for help? Q31. When people outside your household have problems, do they come .271 .105 to you for help? Q32. If you had friends or co-workers whose opinions on politics differed from .058 .605 yours, would you have a hard time conversing with them?

6.4 ACCESS TO PUBLIC SERVICE The Cronbach’s α value for the four items is 0.594, which is considered below standard. Each of the four items is positively correlated with each of the other questions in this battery. Q41 (0.398) has the lowest corrected item-total correlation. Deleting any of the items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α = 0.594.

Corrected Item-total Cronbach's α if Cronbach's α = 0.594 correlation item deleted Q39. An identity document (such as a .288 .591 birth certificate or passport) Q40. A place in a public primary school .421 .507 for a child Q41. Medical treatment at a nearby .398 .518 clinic Q42. Help from the police when you .469 .441 need it

6.5 PSYCHOLOGICAL INVOLVEMENT The Cronbach’s α value for the three items is 0.679, which is considered acceptable. Each of the three items is positively correlated with each of the other questions in this battery. Q46 (0.421) has the lowest corrected item-total correlation. Deleting Q46 from this battery would have a moderate, because it would increase the overall reliability from 0.679 to 0.722, which is above that “good” threshold (α = 0.7). Deleting any of Q44 or Q45 from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.679.

Corrected Item-total Cronbach's α if Cronbach's α = 0.679 correlation item deleted Q44. How interested would you say you .624 .421 are in politics? Q45. How often do you follow news .618 .498 about politics and government? Q46. When you get together with your family members or friends, how often .421 .722 do you discuss political matters?

6.6 INTERNET AND SOCIAL MEDIA The Cronbach’s α value for the six items is 0.513, which is considered below standard. Each of the three items is positively correlated with each of the other questions in this battery, though the correlations are low. Q47 (0.210) has the lowest corrected item-total correlation. Deleting any of six items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.513

Corrected Cronbach's α Cronbach's α = 513 Item-total If Item Deleted Correlation Q47. Do you have Internet access at .210 .506 home? Q48. Do you have Internet access on a .243 .504 mobile phone? Q49. How often do you use the .268 .468 internet? Q50. Do you currently use any of .267 .500 the following social media networks? Q51. How often do you use the .434 .386 Internet including social media networks to find information about politics and government? Q52. How often do you use the Internet including social media .435 .362 networks to express your opinion about politics and government?

6.7 TRADITIONALISM The Cronbach’s α for the 14 items is 0.783, which is considered good. Each of the 14 items is positively correlated with each of the other questions in this battery. Q66 (0.180) and Q67 (0.186) have the lowest corrected item-total correlation. Deleting Q66 or Q67 from this battery would have a moderate but inconsequential effect on the overall Cronbach’s α value. That is, if it were to be deleted, it would increase the overall reliability from 0.783 to 0.792 (Q66) and to 0.790 (Q67), which is not significant change.

Corrected Item-total Cronbach's α if item Cronbach's α = 0.783 correlation deleted Q55.For the sake of the family, the individual should put his personal .425 .769 interests second. Q56.In a group, we should sacrifice our individual interest for the sake of .508 .762 the group’s collective interest. Q57.For the sake of national interest, .521 .758 individual interest could be sacrificed. Q58.When dealing with others, developing a long-term relationship is .495 .762 more important than securing one’s immediate interest. Q59.When dealing with others, one should not only focus on immediate .520 .759 interest but also plan for future. Q60.Even if parents’ demands are .401 .770 unreasonable, children still should do what they ask. Q61.When a mother-in-law and a daughter-in-law come into conflict, even if the mother-in-law is in the .414 .769 wrong, the husband should still persuade his wife to obey his mother. Q62.Being a student, one should not .315 .778 question the authority of their teacher. Q63.In a group, we should avoid open quarrel to preserve the harmony of the .436 .768 group. Q64.Even if there is some disagreement with others, one should .338 .775 avoid the conflict. Q65.A person should not insist on his own opinion if his co-workers disagree .468 .763 with him. Q66.Wealth and poverty, success and .180 .792 failure are all determined by fate. Q67.If one could have only one child, it is more preferable to have a boy than .186 .790 a girl. Q68.When dealing with others, one should not be preoccupied with .482 .763 temporary gains and losses.

6.8 POLITICAL PARTICIPATION 6.8.1 EXPRESSION OF IDEAS The Cronbach’s α for the five items is 0.809, which is considered good. Each of the five items is positively correlated with each of the other questions in this battery. Q73 (0.506) has the lowest corrected item-total correlation. Deleting any of five items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.809. Corrected Item-total Cronbach's α if item Cronbach's α = 0.809 correlation deleted Q69. Contacted elected officials or .550 .785 legislative representatives at any level. Q70. Contacted officials at higher .641 .762 level. Q71. Contacted traditional .683 .745 leaders/community leaders Q72. Contacted other influential people .629 .762 outside the government. Q73. Contacted news media. .506 .803

6.8.2 CIVIC ACTION The Cronbach’s α value for the three items is 0.581, which is considered under standard. Each of the three items is positively correlated with each of the other questions in this battery. Q78 (0.171) has the lowest corrected item-total correlation. Deleting Q78 from this battery would have a moderate, because it would increase the overall reliability from 0.679 to 0.790, which is above that “good” threshold (α = 0.7). Deleting any of Q74 or Q75 from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.581

Corrected Item-total Cronbach's α if item Cronbach's α = 581 correlation deleted Q74. Got together with others to try to .568 .195 resolve local problems. Q75. Got together with others to raise .483 .336 an issue or sign a petition. Q78. Have you voted in every election, voted in most elections, voted in some .171 .790 elections or hardly ever voted?

6.9 REGIME PREFERENCE 6.9.1 IDEAL SYSTEM OF DEMOCRATIC GOVERNMENT This battery consists of four main questions, with each question having a follow-up question. The main question and the follow-up question were grouped together, then re-coded as “Statement 1: Strongly agree”, “Statement 1: Agree,” “Statement 2: Strongly agree,” “Statement 2: Agree”. For example, Q79 and Q79a were grouped, and re-coded as: 1. Statement 1. Government leaders implement what voters want: Strongly agree 2. Statement 1. Government leaders implement what voters want: Agree 3. Statement 2. Government leaders do what they think is best for the people: Strongly agree 4. Statement 2. Government leaders do what they think is best for the people: Agree The Cronbach’s α value for the four items is 0.499, which is considered poor. Each of the four items is positively correlated with each of the other questions in this battery, but the correlations are low. Q81+Q81a (.141) has the lowest corrected item-total correlation. Deleting Q81+Q81a from this battery would have a moderate but inconsequential effect on the overall Cronbach’s α value. That is, if it were to be deleted, it would increase the overall reliability 0.499 to 0.549, which is still below that minimum threshold (α = 0.6) for it to be acceptable. Deleting any other items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.499.

Corrected Item-total Cronbach's α if Cronbach's α = 0.499 correlation item deleted Q79+Q79a. Regime Preference 1: Statement 1. Government leaders implement what voters want. .382 .337 Statement 2. Government leaders do what they think is best for the people. Q80+Q80a. Regime Preference 2: Statement 1. Government is our employee, the people should tell government what needs to be .363 .358 done. Statement 2. The government is like parent, it should decide what is good for us. Q81+Q81a. Regime Preference 3: .141 .549 Statement 1. The media should have the right to publish news and ideas without government control. Statement 2. The government should have the right to prevent the media from publishing things that might be politically destabilizing. Q82+Q82a. Regime Preference 4: Statement 1. Political leaders are chosen by the people through open and competitive elections. .292 .429 Statement 2. Political leaders are chosen on the basis on their virtue and capability even without election.

6.9.2 OPERATION OF CURRENT GOVERNMENT INSTITUTIONS The Cronbach’s α for the four items is 0.791, which is considered good. Each of the four items is positively correlated with each of the other questions in this battery. Q86 (0.556) has the lowest corrected item-total correlation. Deleting any of four items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.791.

Corrected Item-total Cronbach's α if Cronbach's α = 0.791 correlation item deleted Q83. Over the long run, our system of government is capable of solving the .628 .724 problems our country faces. Q84. Thinking in general, I am proud of our .593 .748 system of government. Q85. A system like ours, even if it runs into .652 .714 problems, deserves the people's support. Q86. I would rather live under our system of government than any other that I can .556 .770 think of.

6.10 MEANING OF DEMOCRACY The Cronbach’s α for the four items is 0.738, which is considered good. Each of the four items is positively correlated with each of the other questions in this battery. Q88 (0.495) has the lowest corrected item-total correlation. Deleting any of four items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.738.

Corrected Cronbach's α Cronbach's α = 0.738 Item-total If Item Deleted Correlation Q88. 1. Government narrows the gap between the rich and the poor. 2. People choose the government leaders in free and fair election. .495 .700 3. Government does not waste any public money. 4. People are free to express their political views openly. Q89. 1. The legislature has oversight over the government. 2. Basic necessities, like food, clothes and shelter, are provided for all. .527 .681 3. People are free to organize political groups. 4. Government provides people with quality public services. Q90. 1. Government ensures law and order. 2. Media is free to criticize the things government does. .568 .657 3. Government ensures job opportunities for all. 4. Multiple parties compete fairly in the election. Q91. 1. People have the freedom to take part in protests and demonstrations. .534 .678 2. Politics is clean and free of corruption. 3. The court protects the ordinary people from the abuse of government power. 4. People receive state aid if they are unemployed.

6.11 QUALITY OF GOVERNANCE The Cronbach’s α for the nine items is 0.830, which is considered good. Each of the nine items is positively correlated with each of the other questions in this battery. Q106 (0.490) has the lowest corrected item-total correlation. Deleting any of nine items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.830.

Corrected Item-total Cronbach's α if Cronbach's α = 0.830 correlation item deleted Q101.People has the power to change a .500 .817 government they don’t like. Q102.Political parties or candidates in our country have equal access to the mass .570 .809 media during the election period. Q103.Between elections, the people have no way of holding the government .555 .812 responsible for its actions. Q104.When government leaders break the .494 .819 laws, there is nothing the court can do. Q105.All citizens from different ethnic communities in [Country X] are treated .561 .810 equally by the government. Q106.Rich and poor people are treated .490 .818 equally by the government. Q107.People has basic necessities like .527 .816 food, clothes, and shelter. Q108.People is free to speak what they .573 .809 think without fear. Q109.People can join any organization .601 .805 they like without fear.

6.12 PREFERENCE FOR DEMOCRACY The Cronbach’s α for the five items is 0.706, which is considered good. Each of the five items is positively correlated with each of the other questions in this battery. Q127 (0.352) has the lowest corrected item-total correlation. Deleting any of five items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.706

Corrected Cronbach's α if Cronbach's α = 0.706 Item-total item deleted correlation Q125. Which of the following statements comes closest to your own opinion? (1) Democracy is always preferable to any other kind of government (2) Under some circumstances, an .479 .651 authoritarian government can be preferable to a democratic one (3) For people like me, it does not matter whether we have a democratic or a nondemocratic regime Q126. Which of the following statements comes closer to your own view? (1) Democracy is capable of solving the .563 .613 problems of our society (2) Democracy cannot solve our society’s problems Q127. If you had to choose between democracy and economic development, .352 .701 which would you say is more important? Q128. If you had to choose between reducing economic inequality and .459 .661 protecting political freedom, which would you say is more important? Q129. Do you agree or disagree with the following statement: “Democracy may .491 .656 have its problems, but it is still the best form of government.”

6.13 LEGITMACY OF DEMOCRACY/ DETACHMENT FROM AUTHORITARIANISM

This battery was not included in the survey due to political sensibility.

Corrected Cronbach's α Cronbach's α = Item-total If Item Deleted Correlation Q130. We should get rid of parliament and elections and have a strong leader - - decide things Q131. Only one political party should be allowed to stand for election and - - hold office Q132. The army (military) should - - come in to govern the country Q133. We should get rid of elections and parliaments and have experts - - make decisions on behalf of the people

6.14 CITIZEN EMPOWERMENT AND POLITICAL SUPPORT The Cronbach’s α for the five items is 0.743, which is considered good. Each of the five items is positively correlated with each of the other questions in this battery. Q138 (0.430) has the lowest corrected item-total correlation. Deleting any of four items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.743

Corrected Item-total Cronbach's α Cronbach's α = 743 Correlation If Item Deleted Q134. I think I have the ability to .445 .720 participate in politics Q135. Sometimes politics and government seems so complicated .566 .674 that a person like me can’t really understand what is going on Q136. People like me don’t have any influence over what the government .530 .689 does Q137. You can generally trust the people who run our government to .565 .676 do what is right Q138. A citizen should always remain loyal only to his country, no .430 .724 matter how imperfect it is or what wrong it has done

6.15 AUTHORITARIAN/DEMOCRATIC VALUES The Cronbach’s α value for the eleven items is 0.896, which is considered good. Each of the three items is positively correlated with each of the other questions in this battery. Q139 (0.247) has the lowest corrected item-total correlation. Deleting Q139 from this battery would have a moderate, because it would increase the overall reliability from 0.896 to 0.904, which is above that “excellent” threshold (α = 0.9). Deleting any of other items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.679

Corrected Cronbach's α if Cronbach's α = 0.896 Item-total item deleted correlation Q139.Women should not be involved in politics .247 .904 as much as men. Q140.The government should consult religious .630 .887 authorities when interpreting the laws. Q141.People with little or no education should have as much say in politics as highly-educated .549 .891 people. Q142.Government leaders are like the head of a .703 .883 family; we should all follow their decisions. Q143.The government should decide whether certain ideas should be allowed to be discussed .721 .882 in society. Q144.Harmony of the community will be .613 .888 disrupted if people organize lots of groups. Q145.When judges decide important cases, they .710 .882 should accept the view of the executive branch. Q146.If the government is constantly checked [i.e. monitored and supervised] by the .721 .881 legislature, it cannot possibly accomplish great things. Q147.If we have political leaders who are morally upright, we can let them decide .648 .886 everything. Q148.If people have too many different ways of .593 .889 thinking, society will be chaotic. Q149.When the country is facing a difficult situation, it is ok for the government to .700 .883 disregard the law in order to deal with the situation.

6.16 GLOBALIZATION The Cronbach’s α value for the four items is 0.454, which is considered under standard. Each of the four items is positively correlated with each of the other questions in this battery, though the correlations are low. Q150 (0.170) has the lowest corrected item-total correlation. Deleting Q150 from this battery would have a moderate, because it would increase the overall reliability from 0.454 to 0.458, which is above that minimum threshold (α = 0.6) for it to be acceptable. Deleting any of other items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.454.

Corrected Cronbach's α if Cronbach's α = 0.454 Item-total item deleted correlation Q150.How closely do you follow major events .170 .458 in foreign countries / the world? Q151.Our country should defend our way of life instead of becoming more and more like other .313 .335 countries. Q152.Do you agree or disagree with the following statement: “We should protect our .300 .365 farmers and workers by limiting the import of foreign goods.” Q153. Do you think the government should increase or decrease the inflow of foreign .299 .365 immigrants into the country?

6.17 REDISTRIBUTION The Cronbach’s α value for the four items is 0.507, which is considered under standard. Each of the four items is positively correlated with each of the other questions in this battery, though the correlations are low. Q158 (0.088) has the lowest corrected item-total correlation. Deleting Q158 from this battery would have a moderate, because it would increase the overall reliability from 0.507 to 0.535, which is above that minimum threshold (α = 0.6) for it to be acceptable. Deleting any of other items from this battery would not have significant effect on the overall Cronbach’s α value, i.e., none of the items would increase the reliability if they were deleted because all values are less than the overall reliability α 0.507.

Corrected Item-total Cronbach's α Cronbach's α = 0.507 Correlation If Item Deleted Q155. How fair do you think income .338 .421 distribution is in [Country]? Q156. Do you agree or disagree with the following statement: It is the responsibility of the government to .314 .430 reduce the differences between people with high income and those with low incomes. Q157. How concerned are you about the loss of your or your family's .149 .516 major source of income within the next 12 months? Q158. The following is a hypothetical question: If you were unfortunate enough to lose your .088 .535 main source of income, how serious would it be for you and your family? Q159. Considering all the effort that you and your family members have made in the past, do you think the .309 .433 income that your family currently receives is FAIR or NOT FAIR? Q160. As compared to my parent's generation, my generation has more or fewer opportunities to improve .390 .409 one's standard of living or social status?

7 EXAMINATION OF REPRESENTATIVENESS OF SAMPLE SET AND WEIGHTING

There are few steps to get to the final weights, where firstly we weight by age groups at the regional level. Then another weighting factor is calculated using gender distribution at the regional level as well. The final weighting factor, then, is the multiplication of the two weighting factors.

7.1 GOODNESS-OF-FIT TESTS

Gender

Gender Sample Population Result Frequency Percent Percent

Male 600 50.00 48.46 Chi square= 0.095 Female 600 50.00 51.54 P>.05 (P=0.758)

Total 1200 100.00 100.00 Consistent with the population

Age Group

Age Group Sample Population Result Frequency Percent Percent

18 – 24 182 15.2 20.4

25 – 34 461 38.4 24.6 Chi square= 14.616 35 – 44 249 20.8 21.0 P<.05 (P=0.012) 45 – 54 202 16.8 16.5

55 – 64 81 6.8 8.3 Not consistent with

>=65 25 2.1 9.2 the population

Total 1200 100.00 100.00

Location (6 regions)

Sample Population Result Location Frequency Percent Percent

Northern and 160 13.3 13.4 Chi square= 0.851 mountains P>.05 (P=0.974) Red River 270 22.5 22.7 Consistent with the Delta population Central Coast 260 21.7 21.8

Central 90 7.5 5.9 Highlands

Southeast 180 15.0 16.3

Mekong River 240 20.0 19.9 Delta

Total 1200 100.0 100.0

7.2 SAMPLE REPRESENTATIVENESS

Gender Before weighting After weighting Gender Frequency Percent Frequency Percent

Male 600 50.00 582 48.46

Female 600 50.00 618 51.54

Total 1200 100.0 1200 100.0 Chi square=.0007, p-value>.05, consistent with the population after weighting.

Age Group Before weighting After weighting Age Group Frequency Percent Frequency Percent

18 – 24 182 15.2 245 20.4

25 – 34 461 38.4 295 24.56

35 – 44 249 20.8 252 20.98

45 – 54 202 16.8 199 16.57

55 – 64 81 6.8 99 8.24

>=65 25 2.1 111 9.24

Total 1200 100.00 1201 100.0 Chi square=.012, p-value>.05, consistent with the population after weighting.

Location (6 Regions)

Before weighting After weighting Location Frequency Percent Frequency Percent

Northern and 160 13.3 155 12.91 mountains

Red River 270 22.5 282 23.48 Delta

Central Coast 260 21.7 252 20.98

Central 90 7.5 62 5.16 Highlands

Southeast 180 15.0 207 17.24

Mekong 240 20.0 243 20.23 River Delta

Total 1200 100.0 1201 100.0 Chi square=2.731, p-value>.05, consistent with the population after weighting.

Appendix

1. List of PSU

PSU Region District Ward/Commune Huyện Ngân Sơn Xã Đức Vân Huyện Đồng Văn Xã Sủng Trái Huyện Bắc Quang Xã Bằng Hành Huyện Yên Minh Xã Đông Minh Huyện Cẩm Khê Xã Phú Lạc Huyện Lâm Thao Xã Cao Xá Huyện Thanh Sơn Xã Văn Miếu Northern and Huyện Thanh Sơn Xã Thục Luyện mountains Huyện Thanh Sơn Xã Giáp Lai Thành phố Việt Trì Phường Thanh Miếu Thị xã Phú Thọ Phường Hùng Vương Huyện Mai Sơn Xã Chiềng Chăn Huyện Mộc Châu Xã Tân Xuân Huyện Mường La Thị trấn ít Ong Huyện Sông Mã Xã Pú Pẩu Huyện Thuận Châu Xã Bó Mười Huyện Lý Nhân Xã Phú Phúc Huyện Đông Anh Xã Tiên Dương Huyện Đông Anh Xã Đông Hội Huyện Gia Lâm Xã Dương Quang Huyện Mê Linh Thị Trấn Quang Minh Huyện Mỹ Đức Xã Hợp Thanh Huyện Phúc Thọ Xã Vân Nam Red River Delta Huyện Thạch Thất Xã Thạch Hoà Huyện Thường Tín Xã Tân Minh Huyện Thường Tín Xã Lê Lợi Huyện Từ Liêm Xã Trung Văn Huyện Ứng Hòa Xã Minh Đức Quận Ba Đình Phường Thành Công Quận Cầu Giấy Phường Nghĩa Đô Quận Long Biên Phường Phúc Đồng Thành phố Hà Đông Phường Văn Quán Huyện Thuỷ Nguyên Thị trấn Núi Đèo Huyện Thuỷ Nguyên Xã An Lư Huyện Tiên Lãng Xã Tiên Tiến Quận Ngô Quyền Phường Lê Lợi Huyện Phù Cừ Xã Minh Hoàng Huyện Đông Hưng Xã Đông Tân Huyện Hưng Hà Xã Chi Lăng Huyện Hưng Hà Xã Hồng Lĩnh Huyện Kiến Xương Xã Quang Bình Huyện Tiền Hải Xã Bắc Hải Thành phố Thái Bình Phường Quang Trung Huyện An Nhơn Xã Nhơn Khánh Huyện Hoài Nhơn Xã Tam Quan Bắc Huyện Phù Mỹ Xã Mỹ Lợi Thành phố Qui Nhơn Phường Ngô Mây Thành phố Qui Nhơn Phường Nhơn Bình Huyện Đức Thọ Xã Đức Thịnh Huyện Kỳ Anh Xã Kỳ Tiến Thành phố Hà Tĩnh Xã Thạch Trung Thị xã Hồng Lĩnh Phường Bắc Hồng Huyện Cam Lâm Xã Cam Thành Bắc Huyện Ninh Hòa Xã Ninh Thủy Thành phố Nha Trang Phường Tân Lập Central Coast Thành phố Nha Trang Phường Xương Huân Thị xã Cam Ranh Xã Cam Thành Nam Huyện Quảng Trạch Xã Quảng Tiên Thành Phố Đồng Hới Phường Hải Thành Huyện Hà Trung Xã Hà Vinh Huyện Hoằng Hóa Xã Hoằng Phụ Huyện Ngọc Lặc Xã Lộc Thịnh Huyện Như Xuân Xã Cát Tân Huyện Nông Cống Xã Trung Thành Huyện Quảng Xương Xã Quảng Chính Huyện Quảng Xương Xã Quảng Văn Huyện Thọ Xuân Xã Xuân Phú Huyện Triệu Sơn Xã Dân Quyền Huyện Vĩnh Lộc Xã Vĩnh Hưng Huyện Ea H'leo Xã Ea Khal Huyện Krông Năng Xã Ea Tóh Huyện Krông Năng Xã Ea Hồ Huyện Krông Pắc Xã Ea Kly Huyện Lắk Xã Đắk Phơi Thành phố Buôn Ma Phường Ea Tam Thuột Huyện Cẩm Mỹ Xã Sông Ray Huyện Thống Nhất Xã Gia Tân 1 Huyện Trảng Bom Xã Cây Gáo Huyện Xuân Lộc Xã Lang Minh Huyện Củ Chi Xã Tân Thạnh Đông Huyện Hóc Môn Xã Thới Tam Thôn Central Huyện Hóc Môn Xã Xuân Thới Đông Highlands & Quận 10 Phường 09 Southeast Quận 4 Phường 02 Quận 5 Phường 08 Quận 7 Phường Phú Mỹ Quận 8 Phường 02 Quận 9 Phường Tăng Nhơn Phú A Quận 9 Phường Long Thạnh Mỹ Quận Gò Vấp Phường 17 Quận Tân Phú Phường Phú Thạnh Quận Tân Phú Phường Tân Thới Hòa Quận Thủ Đức Phường Hiệp Bình Chánh Quận Thủ Đức Phường Linh Xuân Huyện Cát Tiên Xã Phù Mỹ Huyện Di Linh Xã Đinh Lạc Huyện Châu Thành Xã Quới Thành Huyện Chợ Lách Xã Hưng Khánh Trung Huyện Mỏ Cày Xã Bình Khánh Đông Mekong River Thị xã Bến Tre Phường Phú Khương Delta Thị xã Bến Tre Phường 6 Huyện Cao Lãnh Xã Mỹ Hiệp Huyện Châu Thành Xã Tân Nhuận Đông Huyện Hồng Ngự Xã Phú Thuận A Huyện Lai Vung Xã Tân Hòa Huyện Tam Nông Thị trấn Tràm Chim Huyện Tam Nông Xã Phú Thọ Thành phố Cao Lãnh Phường 6 Thành phố Cao Lãnh Xã Tân Thuận Tây Huyện An Biên Xã Nam Thái A Huyện Châu Thành Xã Giục Tượng Huyện Kiên Lương Thị trấn Kiên Lương Huyện Phú Quốc Xã Hàm Ninh Huyện Tân Hiệp Xã Tân Hiệp B Huyện Bình Tân Xã Tân Bình Huyện Long Hồ Xã Phú Đức Huyện Long Hồ Xã Hòa Ninh Huyện Trà Ôn Xã Vĩnh Xuân Huyện Trà Ôn Xã Hựu Thành Thị xã Vĩnh Long Phường 8

2. Other supplement information 2.1. Sex Distribution of the Population (Census 2009) Total N (in 2009) % Male 29,068,759 48.46 Female 30,918,352 51.54 Total 59,987,111 100.00

2.2. Distribution of Population (Census 2009) AGE COUNTRY GROUP Total % Male % Female % 18-24 12,236,909 20.40% 6,190,636 50.59% 6,046,273 49.41%

25-34 14,751,222 24.59% 7,415,507 50.27% 7,335,715 49.73%

35-44 12,578,901 20.97% 6,306,952 50.14% 6,271,949 49.86%

45-54 9,923,869 16.54% 4,754,170 47.91% 5,169,699 52.09%

55-64 4,951,194 8.25% 2,239,204 45.23% 2,711,990 54.77% ≥ 65 5,545,016 9.24% 2,162,290 39.00% 3,382,726 61.00%

Total 59,987,111 100.00% 29,068,759 30,918,352