Food Policy 94 (2020) 101813

Contents lists available at ScienceDirect

Food Policy
journal homepage: www.elsevier.com/locate/foodpol

The effects of trust and land administration on economic outcomes: T

Evidence from Vietnam
Duc Anh Danga,⁎, Kim Khoi Dangb, Vuong Anh Dangc, Thi Lan Vub
a
National Center for Socioeconomic Information and Forecast, Viet Nam
b
Institute of Policy and Strategy for Agriculture and Rural Development, Viet Nam
c
National Centre for Social and Economic Modelling, University of Canberra, Australia

ARTICLE INFO ABSTRACT

Keywords: This paper examines how the interaction of social trust and institutions, such as land administration, affects
Trust household economic decisions in Viet Nam. Using a panel dataset of rural households from 2008 to 2014, we
Land administration show that negative consequences of the duration of land administration on agricultural investment have been
Agricultural investment lower in rural households with higher level of trust. Higher-trust households, when facing higher barrier from
Viet Nam
land administration, tend to invest more on agricultural land compared to low-trust households where household
heads are male. We also find that high-trust households enhance land property rights when facing obstacles from
JEL classifications:
land administration.
Z13
O13
K4
O53

1. Introduction main result is that elections have very little effect in villages with low
social trust and a big effect in villages with high social trust. In another
Many empirical studies show that formal institutions, such as study, Nannicini et al. (2013) investigate the effects of civic behaviours2
property right protections, play important roles in determining eco- on political accountability. They documented that individuals living in
nomic outcomes (for example Acemoglu et al., 2001). Another strand of regions with higher social capital assign higher values to public goods.
literature stresses the vital effects of social capital, such as trust on Using data on Italian members of Parliament, they provided supporting
economic development (see Algan and Cahuc (2014) for a review). empirical evidence that elected leaders from districts with higher social
However, while much progress has been made in isolating the im- capital suffer more profound electoral punishment for misbehaviours.
portance of trust1 and institutions, we do not know much about how In above examples, formal institutions and social trust are complement.
they jointly affect development (Alesina and Giuliano, 2015). However, social trust may substitute formal institutions as weak legal
Recent empirical literature shows that formal institutions may enforcement could make people relying more on informal rules and
perform better in societies with high social trust. For example, Miguel local regulations and to develop particularized trust rather than gen-
et al. (2015) examine the joint interaction of political institutions and eralized trust (Algan and Cahuc, 2014). Therefore, depending on the
social trust in determining the provision of public goods in Chinese economic environment and initial conditions, trust and institutions
villages. The authors argue that elections in Chinese villages are more might complement each other, or might act as substitutes, in promoting
effective at choosing politicians, who provide more public goods in economic growth (Bisin and Verdier, 2017).
villages, where generalized trust (that is trust for unknown people) is The primary objective of this paper is to complement recent studies
high relative to personalized trust (that is trust for known people). The that try to understand the interaction effects of trust and institution on

Corresponding author.
⁎

E-mail address: [email protected] (D.A. Dang).

1
Fukuyama (1995) considered trust as a shared expectation of honest and cooperative behaviour in a community. The literature on social trust distinguishes
between generalized trust and particularized trust. Particularized trust refers to the situation in which individuals trust members of a narrow circle of persons. In this
paper, we focus on generalized trust, which is trust for unknown people (Delhey et al., 2011).
2
The authors define civic capital is “those persistent and shared beliefs and values that help a group overcome the free rider problem in the pursuit of socially
valuable activities.”

https://doi.org/10.1016/j.foodpol.2019.101813
Received 11 November 2019; Accepted 11 December 2019
Available online 24 December 2019
0306-9192/ © 2020 UNU-WIDER. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
D.A. Dang, et al. Food Policy 94 (2020) 101813

economic outcomes in a developing country. Viet Nam offers an at- 2. Characteristics of trust and land administration in rural Viet
tractive setting to study these effects. Viet Nam has experienced ex- Nam
ceptional per capita income growth in the last three decades without
large-scale social or political upheavals. However, this impressive 2.1. Land administration in Viet Nam
economic growth cannot be explained by the quality of formal in-
stitutions as Viet Nam has low position on international ranking tables There have been many changes in land policy over the last 30 years
such as Polity IV and the Governance Indicator. One possible ex- in Vietnam. Resolution 10 of the Party in 1988 is a milestone policy of
planation is that weak formal institutions are likely to be substituted by the agriculture reform which expanded the implementation of the
strong informal institutions (Dang, 2012). For instance, the World “package-contract” scheme in which all means of production (land and
Value Surveys show that the Vietnamese national level of social trust is equipment) were handed back to farmers instead of cooperatives. The
higher than some other East Asian nations at the same stage of eco- Land Law enacted in 1993 enables farmers to entitle to long-term
nomic development (Dalton and Ong, 2005). agricultural land assignment (20 years) through the issuance of land-
We examine the hypothesis that social trust substitutes the low ef- use right certificates (LURCs). This law allowed farmers to not only
fectiveness of land administration, which is considered as one of the cultivate on their assigned land but also trade, transfer, rent or inherit
main factors affecting economic decisions by rural households. To test the land use right. In 2003, the Land Law was revised another time with
this hypothesis, we use panel data from the Viet Nam Access to its focus on regulating land area limits and land use terms of households
Resources Household Survey in 2008–14 to investigate whether trust and setting the government land price frame. Most recently, in 2013
and land administrative procedures have joint effects on household’s this Law was amended to increase area limits of land trading and land
agricultural land investment, land-use right certificates ownership and use terms per household. Therefore, land rights, including the right to
credit access. sell, rent, mortgage, exchange, and bequest a plot of land, were guar-
There are difficulties in empirically determining whether land ad- anteed through LURCs and have been gradually reinforced and refined
ministration and social trust are complements or substitutes. Trust is through various amendments to the Land Law. These developments are
correlated with other factors that could influence how land adminis- often recognized as an important determinant of rural economic growth
tration can affect household decisions. To address this, we document in Viet Nam (Pingali and Xuan, 1992; Rozelle and Swinnen, 2004;
the correlates of land administration (such as political connections) and Deininger and Jin, 2008; Do and Iyer, 2008; Kompas et al., 2012;
directly control for the interaction of each correlate and land admin- Newman et al., 2015).
istration in the baseline specification. We also provide a large set of Although many legislation reforms have been implemented, some
robustness tests to rule out the potentially confounding influence from studies show that households cannot fulfil all their land property rights
other factors such as households head characteristics and the demo- and the rights are not always well protected (Markussen, 2017). For
graphic composition of communes. example, Markussen et al. (2011) show that many households face
The results reveal that for households with high trust, the effects of some restrictions on crop choice. Rice households cannot change their
land administration have less negative impacts on their investment in land use purpose to other crops because food security reasons reduce
agricultural land, especially soil investment, compared to those with their production productivity (Giesecke et al., 2013). Luu et al. (2013)
lower trust. High-trust households tend to invest more on agricultural also indicate that while the efficiency of land market transactions has
land compared to low-trust households where household head are male. been improved they are still immature in many regions of Viet Nam.
Compared to low-trust households, they also have higher probability to The government system in Viet Nam consists of both the central
own land-use right certification when facing weak land administration. government and local governments, which comprise of three levels:
The results also confirm the hypothesis that social capital such as trust provincial, district and communal. Accompanying legislation reform,
and institutions such as land administration are substitutes: trust helps the power of land management has been decentralized to lower levels
to improve agricultural land investment where the quality of land ad- of government. This has classified clearer power and mandates at dif-
ministration is weak. In other words, they provide empirical evidence ferent government levels, helping local governments to better manage
that trust can play an important role in determining the success of land. Specifically, the land use designation (lap quy hoach) and planning
improvements in formal institutions. (lap ke hoach) is prepared at three levels: national, provincial, and
This study makes several contributions to the literature. To the best district. Land use plans, developed every 5 years for national and pro-
of our knowledge, this paper is one of the few studies that provide vincial level and every year for district level, are issued by the Ministry
rigorous empirical evidence on the interaction effect of trust and formal of Natural Resources and Environment (MONRE) after the approval of
institutions on household economic decisions such as agricultural land the Government and the National Assembly based on a 10-year land use
investment, ownership of land-use right certificates and access to designation, 5-year socio-economic development plan, the land demand
formal credit. This study complements the findings from recent studies of different sectors, and the performance of the land use plan im-
that have shown the joint effects of trust and institutions on economic plementation in the previous term. Following the approval of land-use
development (such as Nannicini et al., 2013; Miguel et al., 2015). Our planning and plans, the communal People’s Committees decide limits
study contributes to the existing evidence by directly examining land for land allocation, leasing and reclamation and to issue a land price
administration and its effect on household outcomes. We also add to frame in accordance with general regulations by the central govern-
studies that investigate the impacts of generalized trust on economic ments (Le et al., 2015). As the lowest level of the administrative system,
outcomes (for example, Aghion et al., 2010; Alesina and La Ferrara, communes are responsible to manage, use and exploit public lands, and
2002; Algan and Cahuc, 2010). to update and store information relating to any changes in their terri-
The paper is organized as follows. We begin in Section 2 by de- tory. Land users can exploit their land by their own decision but are not
scribing the characteristics of trust and land administration in Viet allowed to change the land use purpose/land use categories regulated
Nam. Section 3 illustrates the conceptual framework. Section 4 dis- in land use plans without the approval of MONRE and provincial au-
cusses our data, along with descriptive analyses of trends in trust and thorities who are responsible for the monitoring and evaluation of the
land administration variables. In Section 5, we present the empirical land use plan implementation in practice.
strategy. Section 6 presents estimation results. Section 7 summarizes As such, commune governments play a crucial role in the security of
the key findings and concludes. property rights. This is most evident in their role in the State issuance of
LURCs. While the process of issuing LURCs to millions of land users
progressed with impressive speed and without obvious signs of wide-
spread abuse by local authorities in the 1990s, current management of

2
D.A. Dang, et al. Food Policy 94 (2020) 101813

LURC issuance is widely perceived to be highly affected by corruption administration, trust can help access to credit faster. In case slow land
(World Bank, 2010; Anderson and Davidsen, 2011). Land administra- administrative procedures, trust can substitute land-use rights as a form
tion in Viet Nam has been regarded as incomplete and unclear in- of collateral that helps households access better to credit.
formation about administrative procedures to the public and the pro-
cesses for issuing property rights and certificates were complicated and 4. Data sources and description
expensive. Therefore, as Markussen and Tarp (2014) point out there is a
risk of land reclamation by local government and household’s land in- This study exploits a four-wave household panel dataset that was
vestment depends on their informal relationship with local government collected in the Vietnam Access to Resources Household Survey
officials. (VARHS) from 2008 to 2014. The VARHS is a panel survey, conducted
in the rural areas of 12 provinces in Viet Nam every second year. The
2.2. Social trust in Viet Nam VARHS re–interviewed all rural households sampled for the income and
expenditure modules of the 2004 Vietnam Household Living Standards
Trust is an important factor in both individual and household social Survey (VHLSS). Attrition in the VARHS is fairly low with an overall
capital. For example, trust can enable people’s engagement in poten- attrition rate from 2008 to 2014 of seven per cent. A common reason for
tially profitable business with (trusted) strangers. Trust within a com- attrition is migration. Based on the responses from local authorities,
munity may facilitate economic cooperation with their partners that two third of migrating households are believed to have migrants per-
benefits all in the long term. Individuals that live in societies with a manently, whereas one third is believed to have migrated temporarily
high level of trust are more likely to divert fewer resources to pur- (Brandt and Tarp, 2017).
chasing protection such as paying bribes (Knack and Keefer, 1997). In The VARHSs collect a broad range of detailed information about
countries that may lack formal institutions, generalized trust could economic and social aspects of the lives of households in rural areas,
substitute for such institutions as a second-best solution. such as rural employment, on- and off-farm income generating activ-
Dalton and Ong (2005) show that Vietnamese levels of social trust ities, rural enterprises, property rights, savings, investment, insurance,
are higher than some other East Asian nations at the same stage of participation in formal and informal social networks, and land invest-
economic development. Rural households in Viet Nam living in a ment. They also include a commune questionnaire that asks informa-
high–risk environment have evolved a series of institutions which serve tion on the general situation of the commune and demographic in-
to reduce individual insecurity. This is accomplished by spreading risk- formation. The commune questionnaire includes one module that asks
taking over a group larger than the nuclear family, such as the extended about administrative procedures related to land-use rights.
family and the corporate community. By choosing to cooperate with
other members, trust amongst village members increases (Dang, 2012).
4.1. Trust variables
Saint-Macary and Zeller (2012) also show that high trust makes people
better access to credit with lower lending rate in Northern provinces in
The surveys ask two standard questions about self-reported trust.
Vietnam.
The exact wording of the questions is as follows: ‘Please tell me whether
in general you agree or disagree with the following statements: Most
3. Conceptual framework
people are generally honest and can be trusted, and: In this commune
one has to be careful, there are people you cannot trust?’3 Respondents
There are several potential mechanisms through which land ad-
could either agree or disagree. Since respondents’ answers to the trust
ministration and trust might influence investment in agriculture land,
questions are binary, we construct a measure of trust that takes on the
LURC ownership and access to credit by rural households.
binary value of 0 and 1, where 0 corresponds to the response ‘Disagree’
First, transaction costs play an important role in land market de-
and 1 to the response ‘Agree’. As shown in Table 2, there is a negative
velopment. High transaction costs in land markets originate from in-
link between two trust scores.
secure land rights and low levels of trust. Better land administration
The distributions of responses for each question are reported in
might make investors feel more secure in investing in agricultural land
Table 1. Based on the data in Table 1, in 2008, 91 per cent of the in-
as it is an important factor in reducing the cost of transactions. Longer
terviewed households agreed that most people are generally honest and
processing time and slower issuance of land-use rights may exacerbate
can be trusted; however, this fell to 87 per cent by 2014, a decrease of
household fears of expropriation or loss of control of land on which
four percentage points since 2008. This trend indicates that rural
investments would be made, potentially deterring such investment. In
households generally have a high level of trust, but they also recognise
societies with efficient land administration, trust could reinforce and
that social interactions could bring more risk (CIEM et al., 2009).
make land administration perform better. Therefore, trust complements
For the second measure of social trust, even though people were
quality of land administration. On the contrary, where land regulations
becoming less confident in others, only 48 per cent of households in
are insufficient, trust also helps reduce transaction costs in land markets
2014 agreed that there are some people who cannot be trusted. This
as the costs of acquiring information, and negotiating and enforcing
ratio was significantly lower than the level of 65 per cent in 2008. This
contracts tend to be much lower. In these cases, high trust may sub-
tendency seems to contradict with the trend of the first measure of trust.
stitute weak land administration.
One potential explanation is that rural households may trust other
Second, shorter time for land administration makes land transaction
people more as they have more social interaction. As Ermisch and
and households receiving LURC faster. Incomplete and unclear in-
Gambetta (2010) suggest, interacting more with other peoples can lead
formation about land administrative procedures to the public makes the
to more “outward exposure”, and improve their ability to trust other
processes for issuing property right certificates complicated and ex-
people by (1) estimating more accurately the probability of trust-
pensive. Therefore, we expect that when facing this situation, high-trust
worthiness; or (2) reading the signs of untrustworthiness more pre-
individuals are more willing to take part in this process but low-trust
cisely.
ones may not want to do so.
Another important point is that the trust questions are answered by
Third, access to credit might be hindered if land-use rights are not
household heads and the household heads could change over time.
sufficiently defined for land to serve as collateral for loans. The longer
time for land administration processes, the lower probability that
households can get a loan. At the same time, social trust plays an im- 3
These trust questions may not fully reflect individual trust attitudes and
portant role in formal financial markets, where it helps to reduce the does not explicitly specify the object of the respondent’s trust (Durante, 2009).
risks of asymmetric information. Therefore, for better land They also could be induced by aversion against risk (Algan and Cahuc, 2014)

3
D.A. Dang, et al. Food Policy 94 (2020) 101813

Table 1 administration systems are unreliable, not transparent and insecure,

Descriptive statistics. land transactions will require more time to be completed. Then, the
VARIABLES 2008 2010 2012 2014 number of processing days is likely to be a sensible proxy for quality of
land administration. Table 1 indicates that the days for land adminis-
Age of household heads 48.9 50.4 51.85 53.5 tration have been shortened recently, decreasing from 13.8 days in
(14.0) (13.47) (13.60) (13.3)
2010 to 10.9 days in 2014.
Gender (Male:=1) 0.83 0.82 0.82 0.80
(0.38) (0.38) (0.38) (0.40)
Year of schooling of household heads 8.07 8.19 8.24 8.76 4.3. Agricultural land investment
(3.59) (3.52) (3.55) (3.30)
Days for land administration at commune 12.26 13.79 12.33 10.9 Table 1 presents a summary on total value of cash investment in
levels
(12.74) (17.86) (11.57) (8.44)
household agricultural land and three different types of land-related
Most people can be trusted (Trust:=1) 0.91 0.89 0.90 0.87 investment, including investments in soil and water conservation,
(0.28) (0.32) (0.30) (0.34) structures for aquaculture (mainly ponds) and other structures, such as
There are people you cannot trust 0.65 0.60 0.53 0.48 farm buildings, fences and animal sheds. The figures indicate that on
(Careful:=1)
average total investment in agricultural land in 2014 has declined and
(0.48) (0.49) (0.50) (0.50)
Ln (1 + Land investment) 1.46 1.12 1.20 0.79 is lower than the past years mainly due to lower levels of investment in
(3.06) (2.64) (2.82) (2.33) permanent structure. However, the soil investment shows an improve-
Ln (1 + Soil investment) 1.43 1.57 1.64 3.09 ment in 2014 compared to those in the past years.
(2.84) (2.89) (2.88) (3.33)
Ln (1 + Aquaculture investment) 0.49 0.71 0.87 0.68
(1.95) (2.28) (2.45) (2.21)
4.4. Formal credit
Ln (1 + Permanent investment) 6.43 0.00 1.26 1.05
(3.54) (0.00) (3.13) (3.08) Table 1 also shows the summary of access to formal credit by rural
Share of household’s land with land-use 0.71 0.63 0.71 0.71 households, which are loans from banks and unions. Overall, formal
right certificates
credit access has decreased with fewer households with loans. Of
(0.40) (0.44) (0.40) (0.40)
Share of borrowing households with formal 0.34 0.38 0.29 0.25 which, there are 34 per cent of households access to formal credit in
credit 2008 but down to 25 per cent in 2014.
(0.47) (0.49) (0.45) (0.43)
Number of observations 1923 1890 1825 1917
5. Empirical strategies
Note: Prices are adjusted for inflation. Standard errors are in parentheses.
Source: Author’s calculation from VARHS 2008–14. Our empirical strategy can be summarized by the following equa-
tion:
yijt = 1 Trustijt + 2 Land _adjt + 3 (Trustijt × Land _adjt ) + Xijt + ijt + + + ijt
Consequently, the difference of the responses may reflect changes in the i t

identity of the respondent rather than trust behaviour and this could be (1)
related to investment and other outcome variables. However, as shown where yijt is the level of outcome variables of household i in commune j
in Table 1, the average age of household heads increases over time during year t (which are agricultural land investment, land registration
following survey timing. The share of male of household heads is almost and access to credit).Trustijt is the generalized trust by household i in
unchanged over time. Therefore, we expect that the head of households commune j during year t. Land _adjt is the time for land administrative
have not changed substantially throughout this period. procedures in commune j at time t. Trustijt × Land _adjt is the interaction
effect of a dummy variable for trust at household levels, Trustijt , and the
time for land administrative procedures at commune levels,Land _adjt .
4.2. Land administrative procedures Xijt are interaction of days for land administration and household’s
political connections. ijt are household and commune characteristics. i ,
We use the number of days that the Commune Government takes to t are household and year fixed effects, respectively. Standard errors are
process an application for land transactions in the commune, such as clustered by commune. 2 captures the effect of land administration
sale, rental, exchange, or other type of transaction of land-use rights as procedures for households with low trust, Trustijt = 0 , which is expected
a measure of quality of land administrative procedures. Of course, the to have negative impacts on household outcomes ( 2 < 0). 2 + 3
number of days may only capture a part of quality of land adminis- captures the effect of land administration procedures for households
tration systems. World Bank (2015) proposed a full set of indicators on with high trust, Trustijt = 1. Our main variable α3 measures the different
reliability, transparency, coverage and dispute resolution to measure effect of land administration on high-trust households compared to low-
the overall quality of land administration. However, if the land trust ones. We expect that trust and institutions, which are proxied by

Table 2
Pairwise correlation among main variables.
(1) (2) (3) (4) (5) (6) (7) (8) (9)

1. Ln (1 + Land investment) 1
2. Share of HH land with LURC −0.04* 1
3. Whether HHs have formal loans 0.004 −0.02 1
4. Days for land administration 0.004 −0.002 0.02* 1
5. Most people can be trusted −0.02 −0.03* 0.002 −0.009 1
6. There are people you cannot trust −0.006 0.01 0.04* −0.03* −0.21* 1
7. HH members hold positions 0.06* −0.01 −0.03* 0.003 0.01 0.01 1
8. HH relatives hold positions 0.04* 0.03* 0.02 −0.02* 0.02 0.05* 0.11* 1
9. HH friends hold positions 0.04* 0.03* 0.00 −0.01 −0.01 0.05* 0.21* 0.26* 1

Source: Author’s calculation from VARHS 2008–14. *Significant at 5 percent level.

4
D.A. Dang, et al. Food Policy 94 (2020) 101813

Table 3
Effects of trust and land administration on household’s agricultural land investment.
Variables (1) (2) (3) (4) (5) (6) (7) (8)

Dependent variable: Ln(Land investment + 1)

FE RE FE RE FE RE FE RE

Trust X Days for land administration 0.019** 0.012** 0.011 0.007

(0.008) (0.006) (0.008) (0.006)
Trust:=1; w/o = 0 −0.67*** −0.47*** −0.59*** −0.41***
(0.176) (0.149) (0.174) (0.148)
Careful X Days for land administration −0.014*** −0.006 −0.014*** −0.006
(0.005) (0.004) (0.005) (0.004)
Careful:=1; w/o = 0 0.311*** 0.042 0.313*** 0.053
(0.114) (0.085) (0.114) (0.086)
Days for land administration −0.017** −0.012* −0.008 −0.005 0.012** 0.004 0.014*** 0.006
(0.008) (0.006) (0.009) (0.007) (0.005) (0.004) (0.005) (0.005)

Observations 8706 8706 8706 8706 7689 7689 7689 7689

R-squared 0.018 0.018 0.02 0.02 0.019 0.019 0.021 0.021
Number of HH 2708 2708 2708 2708 2697 2697 2697 2697

Other controls Yes Yes Yes Yes Yes Yes Yes Yes
Other interacts No No Yes Yes No No Yes Yes
Year dummies Yes Yes Yes Yes Yes Yes Yes Yes
Household FE Yes No Yes No Yes No Yes No

Note: Standard errors, clustered at commune level, are in parentheses. Trust means “Most people can be trusted” and Careful means “There are people you cannot
trust”. Other control variables are age of household head, gender, year of schooling, an indicator for household has a relative who holds a position of public or
bureaucratic responsibility, an indicator for household has a relative who holds a public position, an indicator for household has a member who holds a public
position, an indicator for household has a friend who holds a public position, number of households in communes, number of poor households in communes. Other
interacts includes interaction of days for land administration and indicator for household has a friend who holds a public position, interaction of days for land
administration and indicator for household has a member who holds a public position, interaction of days for land administration and indicator for household has a
relative who holds a public position. *** p < 0.01, ** p < 0.05, * p < 0.1.
Source: Author’s calculation from VARHS 2008–14.

days for land administration procedures, are substituted, then the in- invariant. To check the potential effects of omitted varying variables, a
teraction effect will be positive, α3 > 0. number of control variables are included. In column (1) and (2), we
The challenge in estimating the Eq. (1) is that there is potential include a few characteristics of the household head, namely age, gender
endogeneity between outcomes variables and the interaction between and schooling as controls. In columns (3) and (4), we control for a
trust and land administration. The source of this endogeneity could be larger set of variables. Table 2 shows that political connections are
due to omitted variable bias. It is possible that unobserved character- related to household land investment. At the same time, households
istics of the household or communes which are correlated with trust with better connections with officials may be more confident with in-
and land administration may affect household decisions. Using house- vesting (Markussen and Tarp, 2014). Therefore, we include both poli-
hold-level fixed effects eliminates the potential for any time-invariant tical connections and interaction of political connection and land ad-
characteristics of households and communes to act as confounding ministration in the regression in column (3) and (4). We also control for
factors in our analysis. Moreover, we also introduce the interaction of other commune characteristics such as number of households and
land administration variable with a vector of other social connections number of poor households in communes. Moreover, year-fixed effects
such as political connections, Xijt. By controlling for the interaction of are included to take account of changes over time in the economic
these potential correlates with land administration, we mitigate the environment. The control variables are not presented for ease of ex-
possibility that the coefficient of interest is contaminated by the influ- position. Models in columns (5)–(8) have characteristics identical to
ence that other correlates may have on the effect of the land adminis- columns (1)–(4), respectively, except that social trust is measured as an
tration. alternative indicator of trust.
Models in columns (1) and (2) show positive joint effects of trust
6. Results and land administration on land-related investment. They are con-
sistent with the hypothesis that trust is substitute to land administra-
6.1. Land investment tion. The results in columns (3) and (4) are not statistically significant
but the magnitude of coefficients is almost similar. The effects are also
We now turn to estimating the joint effects of trust and land ad- similar with the second measures of self-reported trust in columns (5)
ministration on agricultural land investment. Table 3 presents estimates and (7). For the magnitude of the coefficients, an one–day increase in
of Eq. (1) with both fixed and random effects. To implement the log- duration of land administration results in low-trust households in-
linear version of Eq. (1) without dropping observations with zero-va- vesting in agricultural land 0.14 per cent4 less than high-trust house-
lues on the dependent variables, we use ln(Land investment + 1) as our holds. In other words, given that the average days of land administra-
dependent variable. The values of investment are inflation-adjusted to tion is around 12.3, 10 per cent increase in the days of land
reflect changes in prices over time. A number of unobservable house-
hold characteristics, such as entrepreneurial spirit, cognitive abilities
and risk preferences are likely to affect both investment decisions and 4
Given low variation of the first measure of social trust over time, we rely on
the trust and therefore may be a source of endogeneity bias. Household the second measure of trust to evaluate the joint impacts of trust and land
fixed effects account for these factors, to the extent that they are time- administration on household outcomes.

5
D.A. Dang, et al. Food Policy 94 (2020) 101813

Table 4
Effects of trust and land administration on household’s decision on agricultural land investment.
Variables (1) (2) (3) (4) (5) (6) (7) (8)

Dependent variable: Household’s decision to invest in agricultural land

FE RE FE RE FE RE FE RE

Trust X Days for land administration 0.002 0.002 0.002* 0.002

(0.002) (0.002) (0.001) (0.001)
Trust:=1; w/o = 0 −0.12*** −0.10*** −0.13*** −0.1***
(0.032) (0.030) (0.028) (0.025)
Careful X Days for land administration −0.002* −0.003 −0.002* −0.002
(0.001) (0.002) (0.001) (0.001)
Careful:=1; w/o = 0 0.074*** 0.073*** 0.076*** 0.075***
(0.024) (0.023) (0.024) (0.023)
Days for land administration −0.000 −0.000 −0.001 −0.001 0.003*** 0.003** 0.003** 0.003**
(0.002) (0.002) (0.001) (0.001) (0.001) (0.002) (0.001) (0.002)

Observations 8706 8706 8706 8706 7689 7689 7689 7689

R-squared 0.144 0.144 0.145 0.145 0.151 0.151 0.152 0.152
Number of HH 2708 2708 2708 2708 2697 2697 2697 2697

Other controls Yes Yes Yes Yes Yes Yes Yes Yes
Other interacts No No Yes Yes No No Yes Yes
Year dummies Yes Yes Yes Yes Yes Yes Yes Yes
Household FE Yes No Yes No Yes No Yes No

Note: Standard errors, clustered at commune level, in parentheses. Trust means “Most people can be trusted” and Careful means “There are people you cannot trust”.
Other control variables are age of household head, gender, year of schooling, an indicator for household has a relative who holds a position of public or bureaucratic
responsibility, an indicator for household has a relative who holds a public position, an indicator for household has a member who holds a public position, an
indicator for household has a friend who holds a public position, number of households in communes, number of poor households in communes Other interacts
includes interaction of days for land administration and indicator for household has a friend who holds a public position, interaction of days for land administration
and indicator for household has a member who holds a public position, interaction of days for land administration and indicator for household has a relative who
holds a public position. *** p < 0.01, ** p < 0.05, * p < 0.1.
Source: Author’s calculation from VARHS 2008–14.

administration results in an average of 1.7 per cent higher in land in- significant in column (4). The estimate in the last column shows that
vestment of high-trust people compared to low-trust ones.5 male household heads with higher trust tend to invest more than those
The results in Table 3 also reveal that trust has significant effects on with lower trust when facing more land administration.
agricultural investment. However, it has negative effects on land in- We now turn to the examination of how interaction of trust and land
vestment6. These results seem to contradict with the standard argu- administration affects land investment by different ethnic groups7. The
ments about the effects of trust on investment. There is one possible fixed effect estimates are reported in Table 6. According to estimates
interpretation for this situation. As mentioned above, the trust ques- from column (1), as land administration increases, other ethnic
tions do not explicitly specify the object of the respondent’s trust. Their households with higher trust invest more than lower trust ones for the
responses may reflect and be influenced by historical experiences, such first measure of trust. However, the result in column (4) shows that
as experiences with collectivized agriculture. Rural households may among Kinh households, higher trust households invest more on agri-
have had bad experiences from collective farming where incentives to cultural land than lower trust ones as facing an increase in duration of
invest in agricultural land are reduced because of free rider problems land administration.
(CIEM et al., 2009). In addition, although many legislation reforms We next turn to a closer examination of the impact of joint effects of
have been implemented, there is still a risk of land reclamation by local trust and land administration on different types of agricultural land
governments (Markussen and Tarp, 2014). All of these could affect investment. The fixed effect estimates are reported in Table 7. As shown
more profound on high-trust households and their decision to invest in in columns (1) and (4), the effects of joint trust and land administration
agricultural land. are mainly attributed to investment in soil and irrigation systems. These
Along with examining the joint effect of social trust and land ad- results are not surprising because crop production still plays an im-
ministration on agricultural land investment, we also investigate its portant role in the majority of households in rural Viet Nam.
impact on household’s decision to invest. The results are reported in
Table 4. Rather than the amount of land investment, the dependent 6.2. Land property rights
variable is now a dummy variable, which takes value of 1 if households
invest and 0 otherwise. Similar to the findings in Table 3, the key A key channel through which land administration and trust may
variables in columns (1)–(4) show that with an increase in duration of affect investment is through property rights. In general, land plots with
land administration, households with high trust have a higher prob- LURC have received more investment than plots without LURC
ability to invest than ones with lower trust, although the estimates are (VARHS, 2014). However, the process of issuing LURCs is complex and
only statistically significant in some estimations. time-consuming and we expect that high trust individuals are more
Table 5 presents the fixed effect estimation of trust and land ad- willing to take part in the process. Table 8 reports the joint effect of
ministration on household’s land investment by gender. The main trust and land administration on LURC ownership by rural households.
coefficients in all models have expected signs but are only statistically The control variables in Table 8 are generally the same as in Table 3.
We present results both with a limited set of exogenous characteristics
of household heads, and with a larger set of controls. The results in
5
The magnitude is calculated as 1.23 * 0.014 * 100 = 1.7%.
6
The effect of trust for a household with an average of days for land ad-
7
ministration in column (1) in Table 3 is -0.67 + 0.019 * 12.3 = −0.43. The non-Kinh group accounts for 38 per cent of our sample.

6
D.A. Dang, et al. Food Policy 94 (2020) 101813

Table 5 Table 6
Effects of trust and land administration on household’s land investment by Effects of trust and land administration on household’s land investment by
gender. ethnicity.
Variables (1) (2) (3) (4) Variables (1) (2) (3) (4)

Dependent variable: Ln(Land investment + 1) Dependent variable: Ln(Land investment + 1)

Female Male Female Male Other ethnic Kinh Other Kinh

groups ethnic
Trust X Days for land 0.018 0.013 groups
administration
(0.014) (0.009) Trust X Days for land 0.023*** 0.010
Trust:=1; w/o = 0 −0.392 −0.689*** administration
(0.280) (0.188) (0.007) (0.012)
Careful X Days for land −0.004 −0.009* Trust:=1; w/o = 0 −0.993*** −0.514***
administration (0.286) (0.195)
(0.012) (0.005) Careful X Days for land 0.002 −0.021**
Careful:=1; w/o = 0 0.006 0.200 administration
(0.173) (0.132) (0.007) (0.009)
Days for land −0.005 −0.011 0.008 0.009* Careful:=1; w/o = 0 −0.169 0.324**
administration (0.186) (0.137)
(0.015) (0.009) (0.011) (0.005) Days for land −0.022*** 0.003 −0.001 0.023**
administration
Observations 1578 7107 1413 6262
(0.007) (0.012) (0.007) (0.009)
R-squared 0.037 0.037 0.032 0.038
Number of HH 571 2307 562 2287 Observations 2986 5699 2618 5057
R-squared 0.042 0.051 0.038 0.055
Other controls Yes Yes Yes Yes
Number of HH 1034 1710 1023 1707
Other interacts Yes Yes Yes Yes
Year dummies Yes Yes Yes Yes Other controls Yes Yes Yes Yes
Household FE Yes Yes Yes Yes Other interacts Yes Yes Yes Yes
Year dummies Yes Yes Yes Yes
Note: Standard errors, clustered at commune level, in parentheses. Trust means Household FE Yes Yes Yes Yes
“Most people can be trusted” and Careful means “There are people you cannot
trust”. Other control variables are age of household head, gender, year of Note: Standard errors, clustered at commune level, in parentheses. Trust means
schooling, an indicator for household has a relative who holds a position of “Most people can be trusted” and Careful means “There are people you cannot
public or bureaucratic responsibility, an indicator for household has a relative trust”. Other control variables are age of household head, gender, year of
who holds a public position, an indicator for household has a member who schooling, an indicator for household has a relative who holds a position of
holds a public position, an indicator for household has a friend who holds a public or bureaucratic responsibility, an indicator for household has a relative
public position, number of households in communes, number of poor house- who holds a public position, an indicator for household has a member who
holds in communes. Other interacts includes interaction of days for land ad- holds a public position, an indicator for household has a friend who holds a
ministration and indicator for household has a friend who holds a public po- public position, number of households in communes, number of poor house-
sition, interaction of days for land administration and indicator for household holds in communes. Other interacts includes interaction of days for land ad-
has a member who holds a public position, interaction of days for land ad- ministration and indicator for household has a friend who holds a public po-
ministration and indicator for household has a relative who holds a public sition, interaction of days for land administration and indicator for household
position. *** p < 0.01, ** p < 0.05, * p < 0.1. has a member who holds a public position, interaction of days for land ad-
Source: Author’s calculation from VARHS 2008–14. ministration and indicator for household has a relative who holds a public
position. *** p < 0.01, ** p < 0.05, * p < 0.1.
Source: Author’s calculation from VARHS 2008–14.
Table 8 show that the interaction of trust and land administration has
positive and statistically significant effects on the ownership of LURC
by rural households. There are no statistically significant effects of the interaction of trust
and land administration on access to credit. The results in columns
(5)–(8) for the second measure of trust suggest that access to credit is
6.3. Credit access
improved for rural households with high trust when facing the burden
from land administration although the coefficients are not statistically
Land administration and LURC play an important role in the allo-
significant.
cation of loans from lending institution. Shorter time for land admin-
istration makes land transaction and issuing LURC faster. LURC is the
main collateral that may improve access to formal loans. Therefore, we 7. Conclusion
expect that the longer time for land administration processing, the
lower probability that households can get a loan. At the same time, In this study, we examine the combined effects of trust and land
social trust plays an important role in formal financial markets, where it administrative procedures on investment in agricultural land, land-use
helps to reduce the risks of asymmetric information. If trust is a sub- right certificates and formal borrowings in rural communities in Viet
stitute to land administration, we expect that households with high Nam. The findings suggest that when facing longer duration of land
trust may have higher chances to access formal loans. administration households with higher trust tend to increase their levels
Table 9 presents regressions for whether households have taken of land-related investment, especially soil investment, compared to
loans from formal lenders. The set of control variables is similar to the those with lower trust. Higher-trust households, when confronting with
set used in Table 3. Again, we present results both with a limited set of higher barrier from land administration, are likely to invest more on
exogenous characteristics of the household head, and a set of controls agricultural land compared to low-trust households where household
which are the interaction of days for land administration with political heads are male. We also find that high-trust households strengthen de
connections and commune characteristics. facto land property rights when facing obstacles from land adminis-
The results in columns (1)–(4) represent the differential effects of tration. These results highlight the economic importance of substitution
land administration on households with high trust relative to house- between trust and institutions, particularly in environments where
holds with low levels of trust for the first measure of self-reported trust. property rights institutions are not efficient. On the broader sense, the

7
D.A. Dang, et al. Food Policy 94 (2020) 101813

Table 7
Effects of trust and land administration on types of household’s agricultural land investment.
Variables (1) (2) (3) (4) (5) (6)

Ln(1 + Soil Ln(1 + Aqua Ln(1 + Permanent Ln(1 + Soil Ln(1 + Aqua Ln(1 + Permanent
investment) investment investment) investment) investment investment)

Trust X Days for land 0.020** 0.005 0.085

administration
(0.009) (0.008) (0.069)
Trust:=1; w/o = 0 −1.138*** 0.256 −1.402
(0.301) (0.203) (1.038)
Careful X Days for land −0.011* −0.002 0.008
administration
(0.006) (0.005) (0.044)
Careful:=1; w/o = 0 0.176 −0.133 1.220
(0.178) (0.165) (0.743)
Days for land administration −0.019** −0.003 −0.102 0.011* 0.004 −0.023
(0.009) (0.008) (0.069) (0.006) (0.004) (0.035)

Observations 3663 3663 1364 3209 3209 1173

R-squared 0.055 0.007 0.460 0.054 0.007 0.504
Number of HH 2018 2018 1157 1902 1902 1015

Other controls Yes Yes Yes Yes Yes Yes

Other interacts Yes Yes Yes Yes Yes Yes
Year dummies Yes Yes Yes Yes Yes Yes
Household FE Yes Yes Yes Yes Yes Yes

Note: Standard errors, clustered at commune level, in parentheses. Trust means “Most people can be trusted” and Careful means “There are people you cannot trust”.
Other control variables are age of household head, gender, year of schooling, an indicator for household has a relative who holds a position of public or bureaucratic
responsibility, an indicator for household has a relative who holds a public position, an indicator for household has a member who holds a public position, an
indicator for household has a friend who holds a public position, number of households in communes, number of poor households in communes. Other interacts
includes interaction of days for land administration and indicator for household has a friend who holds a public position, interaction of days for land administration
and indicator for household has a member who holds a public position, interaction of days for land administration and indicator for household has a relative who
holds a public position. *** p < 0.01, ** p < 0.05, * p < 0.1.
Source: Author’s calculation from VARHS 2008–14.

Table 8
Effects of trust and land administration on share of household plots with land-use right certificates.
Variables (1) (2) (3) (4) (5) (6) (7) (8)

Dependent variable: Share of household plots with land-use right certificates

FE RE FE RE FE RE FE RE

Trust X Days for land administration 0.002 0.002 0.002* 0.002

(0.002) (0.002) (0.001) (0.001)
Trust:=1; w/o = 0 −0.12*** −0.10*** −0.13*** −0.1***
(0.032) (0.030) (0.028) (0.025)
Careful X Days for land administration −0.002* −0.003 −0.002* −0.002
(0.001) (0.002) (0.001) (0.001)
Careful:=1; w/o = 0 0.074*** 0.073*** 0.076*** 0.075***
(0.024) (0.023) (0.024) (0.023)
Days for land administration −0.000 −0.000 −0.001 −0.001 0.003*** 0.003** 0.003** 0.003**
(0.002) (0.002) (0.001) (0.001) (0.001) (0.002) (0.001) (0.002)

Observations 8706 8706 8706 8706 7689 7689 7689 7689

R-squared 0.144 0.144 0.145 0.145 0.151 0.151 0.152 0.152
Number of HH 2708 2708 2708 2708 2697 2697 2697 2697

Other controls Yes Yes Yes Yes Yes Yes Yes Yes
Other interacts No No Yes Yes No No Yes Yes
Year dummies Yes Yes Yes Yes Yes Yes Yes Yes
Household FE Yes No Yes No Yes No Yes No

Note: Standard errors, clustered at commune level, in parentheses. Trust means “Most people can be trusted” and Careful means “There are people you cannot trust”.
Other control variables are age of household head, gender, year of schooling, an indicator for household has a relative who holds a position of public or bureaucratic
responsibility, an indicator for household has a relative who holds a public position, an indicator for household has a member who holds a public position, an
indicator for household has a friend who holds a public position, number of households in communes, number of poor households in communes. Other interacts
includes interaction of days for land administration and indicator for household has a friend who holds a public position, interaction of days for land administration
and indicator for household has a member who holds a public position, interaction of days for land administration and indicator for household has a relative who
holds a public position. *** p < 0.01, ** p < 0.05, * p < 0.1.
Source: Author’s calculation from VARHS 2008–14.

8
D.A. Dang, et al. Food Policy 94 (2020) 101813

Table 9
Effects of trust and land administration on household’s formal borrowing.
Variables (1) (2) (3) (4) (5) (6) (7) (8)

Dependent variable: Households have formal borrowing

FE RE FE RE FE RE FE RE

Trust X Days for land administration −0.001 −0.001 −0.002 −0.001

(0.001) (0.001) (0.001) (0.001)
Trust:=1; w/o = 0 0.017 0.018 0.021 0.018
(0.020) (0.018) (0.020) (0.019)
Careful X Days for land administration −0.025 −0.023 −0.025 −0.023
(0.017) (0.014) (0.017) (0.014)
Careful:=1; w/o = 0 0.000 0.000 0.000 0.000
(0.001) (0.001) (0.001) (0.001)
Days for land administration 0.001 0.000 0.001 0.000 −0.001 −0.001 −0.001 −0.001
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)

Observations 8706 8706 8706 8706 7689 7689 7689 7689

R-squared 0.024 0.024 0.024 0.024 0.025 0.025 0.025 0.025
Number of HH 2708 2708 2708 2708 2697 2697 2697 2697

Other controls Yes Yes Yes Yes Yes Yes Yes Yes
Other interacts No No Yes Yes No No Yes Yes
Year dummies Yes Yes Yes Yes Yes Yes Yes Yes

Note: Standard errors, clustered at commune level, in parentheses. Trust means “Most people can be trusted” and Careful means “There are people you cannot trust”.
Other control variables are age of household head, gender, year of schooling, an indicator for household has a relative who holds a position of public or bureaucratic
responsibility, an indicator for household has a relative who holds a public position, an indicator for household has a member who holds a public position, an
indicator for household has a friend who holds a public position, number of households in communes, number of poor households in communes. Other interacts
includes interaction of days for land administration and indicator for household has a friend who holds a public position, interaction of days for land administration
and indicator for household has a member who holds a public position, interaction of days for land administration and indicator for household has a relative who
holds a public position. *** p < 0.01, ** p < 0.05, * p < 0.1.
Source: Author’s calculation from VARHS 2008–14.

results also provide evidence showing why effects of institutional Working Paper No. 23375. National Bureau of Economic Research, Cambridge, MA.
changes on economic development are more profound in one region but Brandt, K., Tarp, F., 2017. Characteristics of the VARHS data and other data sources. In:
Tarp, F. (Ed.), Growth, Structural Transformation, and Rural Change in Viet Nam: A
not in the others, which depend on different cultural traits of people in Rising Dragon on the Move. Oxford University Press. UNU-WIDER Studies in
the regions. Development Economics, pp. 26–50. https://doi.org/10.1093/acprof:oso/
The findings also show some evidence that faster economic devel- 9780198796961.003.0002.
CIEM, DOE, ILSSA, IPSARD, 2009. Characteristics of the Vietnamese rural economy:
opment may be promoted if land administrative procedures can be Evidence from a 2008 rural household survey in 12 provinces in Vietnam. Statistical
simplified, which enhances investment of households who have Publishing House.
stronger trust on others. This would help increase agricultural invest- Dalton, R., Ong, N., 2005. Civil Society and Social Capital in Vietnam. Centre for the
Study of Democracy, University of California, Irvine.
ment and the agricultural sector, which are good for improved food
Dang, A., 2012. Cooperation makes beliefs: weather variation and sources of social trust
security in developing countries, especially in rural areas, where pov- in Vietnam. In: ANU Working Papers in Economics and Econometrics. Australian
erty is more pronounced. National University, Canberra, pp. 2012–2596.
Deininger, K., Jin, S., 2008. Land sales and rental markets in transition: evidence from
rural Vietnam. Oxford Bull. Econ. Stat. 70 (1), 67–101.
Acknowledgements Delhey, J., Newton, K., Welzel, C., 2011. How general is trust in most people'? Solving the
radius of trust problem. Am. Sociol. Rev. 76 (5), 786–807.
This paper is based on research carried out for UNU-WIDER’s pro- Do, Q.T., Iyer, L., 2008. Land titling and rural transition in Vietnam. Econ. Dev. Cult.
Change 56 (3), 531–579.
ject on Rural households in Viet Nam. This research is also supported by Durante, R., 2009. Risk, Cooperation and the Economic Origins of Social Trust: An
Vietnam National Foundation for Science and Technology Development Empirical Investigation. Brown University, Mimeo.
(NAFOSTED), Vietnam under grant number 502.01-2015. We would Ermisch, J., Gambetta, D., 2010. Do strong family ties inhibit trust? J. Econ. Behav. Org.
75 (3), 365–376.
like to thank Saurabh Singhal, Vladimir Betov, and two anonymous Fukuyama, F., 1995. Trust: The Social Virtues and the Creation of Prosperity. Free Press,
referees for comments that substantially improved the paper. We also New York.
thank seminar participants at UNU-WIDER conference for valuable Giesecke, J.A., Tran, N.H., Corong, E.L., Jaffee, S., 2013. Rice land designation policy in
Vietnam and the implications of policy reform for food security and economic wel-
comments. fare. J. Dev. Stud. 49 (9), 1202–1218.
Knack, S., Keefer, P., 1997. Does social capital have an economic payoff? A cross-country
References investigation. Quart. J. Econ. 112 (4), 125–188.
Kompas, T., Nhu Che, T., Nguyen, H.T.M., Nguyen, H.Q., 2012. Productivity, net returns,
and efficiency: land and market reform in Vietnamese rice production. Land Econ. 88
Acemoglu, D., Johnson, S., Robinson, J.A., 2001. The colonial origins of comparative (3), 478–495.
development: an empirical investigation. Am. Econ. Rev. 91 (5), 1369–1401. Le, Q. Trung, Phuong, Vu T., Yang, A., Hai, Vo D., 2015. The Distribution of powers and
Aghion, P., Algan, Y., Cahuc, P., Shleifer, A., 2010. Regulation and distrust. Quart. J. responsibilities affecting forest, land use, and REDD+ across levels and sectors in
Econ. 125 (3), 1015–1049. Vietnam: a legal study. Occasional Paper 137. Centre for International Forestry
Alesina, A., Giuliano, P., 2015. Culture and institutions. J. Econ. Literat. 53 (4), 898–944. Research, Bogor.
Alesina, A., La Ferrara, E., 2002. Who trusts others? J. Public Econ. 85 (2), 207–234. Luu, D.K., Markussen, T., McCoy, S., Tarp, F., 2013. Access to land: market and non-
Algan, Y., Cahuc, P., 2010. Inherited trust and growth. Am. Econ. Rev. 100 (5), market land transactions in rural Vietnam. In: Holden, Stein, Otsuka, Keijiro,
2060–2092. Deininger, Klaus (Eds.), Land Tenure Reforms in Asia and Africa: Assessing Impacts
Algan, Y., Cahuc, P., 2014. Trust, well-being and growth: new evidence and policy im- on Poverty and Natural Resource Management. Palgrave Macmillan, London, pp.
plications. In: Handbook of Economic Growth, pp. 49–120. 162–186 Chapter 7.
Anderson, J., Davidsen, S. (Eds.), 2011. Recognizing and Reducing Corruption Risks in Markussen, T., Tarp, F., Van den Broeck, K., 2011. The forgotten property rights: evidence
Land Management in Vietnam. National Political Publishing House (Su That), Hanoi. on land use rights in Vietnam. World Dev. 39 (5), 839–850.
Bisin, A., Verdier, T., 2017. On the Joint Evolution of Culture and Institutions. NBER Markussen, T., Tarp, F., 2014. Political connections and land-related investment in rural

9
D.A. Dang, et al. Food Policy 94 (2020) 101813

Vietnam. J. Dev. Econ. 110, 291–302. of joint land titling in Vietnam. Land Economics 91 (1), 91–105.
Markussen, T., 2017. Land and land markets. In: Tarp, F. (Ed.), Growth, Structural Pingali, P., Xuan, V.T., 1992. Vietnam: decollectivization and rice productivity growth.
Transformation, and Rural Change in Viet Nam: A Rising Dragon on the Move. Oxford Econ. Dev. Cult. Change 40 (4), 697–718.
University Press. https://doi.org/10.1093/acprof: oso/9780198796961.003.0002. Rozelle, S., Swinnen, J.F.M., 2004. Success and failure of reform: insights from the
UNU-WIDER Studies in Development Economics. transition of agriculture. J. Econ. Literat. 42 (2), 404–456.
Miquel, G., Padró, I., Qian, N., Xu, Y., Yao, Y., 2015. Making democracy work: culture, Saint-Macary, C., Zeller, M., 2012. Rural Credit Policy in Mountains of Northern Vietnam:
social capital and elections in China. NBER Working Paper No. 21058. National Sustainability, Outreach and Impact. Conference paper.
Bureau of Economic Research, Cambridge, MA. World Bank, Embassy of Denmark and Embassy of Sweden, 2010. Assessing Risk Factors
Nannicini, T., Stella, A., Tabellini, G., Troiano, U., 2013. Social capital and political ac- for Corruption in Land Management. Hanoi, Vietnam.
countability. Am. Econ. J.: Econ. Policy 5 (2), 222–250. World Bank, 2015. Doing Business 2015: Going Beyond Efficiency. Washington DC.
Newman, C., Tarp, F., van den Broeck, K., 2015. Property rights and productivity: the case

10