DSpace at VNU: The allocation of forestry land in Vietnam: did it cause the expansion of forests in the northwest?

Forest expansion coincided with a new forest policy that mandated the devolution of forest management authority.. The findings of three village studies suggest that the new forest policy

The allocation of forestry land in Vietnam: did it cause the

expansion of forests in the northwest?

Thomas SikorU,1

Center for Natural Resources and En¨ironmental Studies, Hanoi National Uni¨ersity, 167 Bui Thi Xuan, Hanoi, Vietnam

Received 5 June 2000; received in revised form 24 October 2000; accepted 7 November 2000

Abstract

Forests expanded rapidly in northwestern Vietnam in the 1990s Forest expansion coincided with a new forest policy that mandated the devolution of forest management authority A cornerstone of the new policy was the allocation of use rights for forestry land and trees to rural households This paper examines to what extent the new forest policy contributed to the observed forest expansion The findings of three village studies suggest that the new forest policy had minor effects on actual property rights, as villagers resisted its implementation Instead, forests expanded, mainly due to the liberalization of agricultural output markets and availability of new technology Changes

in markets and technology motivated farmers to intensify crop production, reducing agricultural pressure on land The research findings suggest the potential of market-based instruments and technology policy to facilitate forest regeneration They also demonstrate the benefits of in-depth village studies for forest policy analysis, as it provides

an integrated framework for assessing the relative effects of political, economic and technological changes on forests.

䊚 2001 Elsevier Science B.V All rights reserved.

Keywords: Land tenure; Devolution; Agricultural intensification; Research method

UHumboldt-Universitat zu Berlin, Institut fur Wirtschafts- und Sozialwissenschaften des Landbaus, Fachgebiet

omie, Luisenstr 56, D-10099 Berlin, Germany Tel.: q49-30-2093-6320; fax: q49-30-2093-6497.

E-mail address: thomas.sikor@rz.hu-berlin.de T Sikor

1 Vietnamese contact: Tel.: q84-4-978-0822; fax: q84-4-821-8934.

1389-9341 r01r$ - see front matter 䊚 2001 Elsevier Science B.V All rights reserved.

PII: S 1 3 8 9 - 9 3 4 1 0 0 0 0 0 4 1 - 1

1 Introduction

Forests rapidly expanded in northwestern

Viet-nam in the 1990s The People’s Committee of Son

La province reported that forest cover had risen

from 12% in 1990 to 22% in 1997, an increase of

80% in as little as 7 years At the same time,

changes in population and climate could not

ac-count for the increase Son La’s population had

continued to grow rapidly, and no change in

cli-mate had been recorded If not changes in

popu-lation or climate, what factors caused the

expan-sion of forest cover?

The People’s Committee attributed the

expan-sion of forests to successful implementation of

Vietnam’s new forest policy Beginning in 1993,

national policy had called for the devolution of

authority over forest management The policy

mandated that forest management be handed over

from state enterprises at the central government

and provincial levels to households, villages, and

communes The People’s Committee’s claim of

success resonates with a major theme in the

inter-national literature, which suggests that the

devo-lution of forest management bears significant

po-tential to improve forest management

This paper examines the causes underlying

forest expansion in northwestern Vietnam In

particular, it seeks to understand to what degree

the devolution of forest management authority

contributed to the increase in forest cover It

reports the findings of in-depth field research on

actual changes in property rights effected by the

new policy The research also compares the

ef-fects of the new policy with the influence of other

external changes, such as the liberalization of

agricultural output markets and the availability of

new technology

The paper is expected to make two

contribu-tions to the literature First, it hopes to improve

understanding of the conditions under which

de-volution can lead to better forest management

Its findings also call for attention to alternative

policy levers Policy modifying markets and

avail-able technology may have more potential to

facili-tate forest regeneration than devolution Second,

the findings of the paper highlight the benefits of

in-depth village studies Loclevel research

al-lows researchers to assess the relative effects of changes in property rights and other external factors on forests Its benefits stem from the fact that village land and tree tenure may be signifi-cantly different from national policy In Son La, village studies show that the change in authority over forest management, which appeared so radi-cal on paper, had little effect in practice

The paper begins with brief discussions of the literature and research design The following three sections report research results: the change in forest policy and its implementation in Son La province; the effects of policy implementation on property rights; and the effects of changing mar-kets and technologies on the broader dynamics of land use The paper concludes with a discussion

of implications for policy and research

2 The literature

This paper connects two themes in the litera-ture on forest policy First, there has been an explosion of studies on ‘local forest management’

ŽWollenberg, 1998 These studies document local people’s historical and contemporary forest-use practices, as well as associated property rights arrangements They also emphasize that local people’s involvement has the potential to improve forest management Noting that forest policy has

so far excluded local people from the forest, they call for the devolution of forest management

au-Ž thority to local levels of governance for example, see Poffenberger, 1990; Bartlett and Malla, 1992; Peluso, 1992a,b; Lynch and Talbott, 1995;

Pof- fenberger and McGean, 1996 The second theme is the recent emergence of studies that examine the effects of macroeconomic

Ž forces on forests for example, see Kaimowitz et al., 1998; Ndoye and Kaimowitz, 1998; Sunderlin and Pokam, 1998; Sunderlin, 1999; Colchester,

1999; Minde et al., 1999 These studies view forests as part of livelihood strategies and busi-ness plans that respond to macroeconomic policy changes The changes that have received most attention include those effected by structural ad-justment programs, economic crisis, the liberal-ization of agricultural output markets, and trade

policy Correspondingly, the studies advocate for

intervention at the macroeconomic level to

achieve better forest management

Three disjunctures separate these bodies of

literature: a theoretical, a conceptual and a

methodological one The theoretical disjuncture

lies in the explanatory significance attributed to

the factors affecting forest management The

lit-erature on local forest management emphasizes

property rights as the key factor affecting forest

management It gives much attention to the

ques-tion of how to distribute property rights between

the state and villages, and among villagers Other

factors are assumed to be irrelevant, in most

cases implicitly, but sometimes this assumption is

also stated explicitly Richards, 1993 In stark

contrast, property rights do not figure

promi-nently in the equation of studies on the effects of

macroeconomic factors Strengthening local

peo-ple’s rights to land and trees appears a feeble

attempt in the wake of the unstoppable force of

macroeconomic factors

The theoretical disjuncture is rooted in a more

fundamental conceptual one, as I argue following

Romm 1986 Studies of ‘local forest

manage-ment’ typically consider local people and forests

in isolation from larger economic forces They

tend to ignore those forces, despite their effects

on the livelihood and income of those people who

are expected to manage forests sustainably once

they are given the respective authority A rare

exception are studies of the role of non-timber

forest products in local economies that pay

atten-Ž tion to national and global markets for example,

see Raintree and Francisco, 1994; Godoy et al.,

1995; Wilkie and Godoy, 1996 In contrast,

stud-ies on the impacts of macroeconomic forces

as-sume that those forces determine the fate of

people and forests They consider resource-use

practices to express the combination of those

forces Putting forest management in the hands of

local people, private companies, or the state would

not make much difference, as macroeconomic

factors would determine the actions of all three

The conceptual differences are tied to different

research methods Studies of ‘local forest

man-agement’ are typically located at the local level

They emphasize the diversity and complexity of

local property regimes If they aim at larger scales

of relevance, they typically combine the findings

of several case studies with attention to national legislation on property rights In contrast, studies

on the impacts of large-scale forces tend to re-main at the macro level They pay little attention

to the concrete local expressions and effects of large-scale forces

The challenge is to find an analytical approach that allows weighing the relative significance of macroeconomic forces versus local property rights

as factors modifying forests Such an approach would need to be sensitive to the local specificity

of macro forces, as well as to macro influences on local conditions It would yield insights into the conditions that influence the relative influence of the different factors This paper is intended to demonstrate such an approach

3 Research design and background

The remainder of this paper presents an empir-ical analysis of the causes underlying the increase

in forest cover in northwestern Vietnam The research took a broad look at village land-use practices and the strategies that guide villagers’ allocation of labor to different opportunities It examined the concrete effects of policy reform on property rights and villagers’ allocation of labor, and it contrasted them with the effects of changes

in markets and technologies The research thus combined in-depth understanding of local change with attention to the effects of large-scale forces The research included the collection of so-cioeconomic data in three villages In each vil-lage, semi-structured interviews with a 20% ran-dom sample of households yielded data on cur-rent land-use practices, changes in land and tree tenure over the past 15 years and plot histories Direct observation and key informant interviews deepened the understanding of policy implemen-tation and village tenure Review of government documents helped to describe formal tenure Interpretation of remote imagery yielded data

on changes in land use and forest cover The interpretation used SPOT images from 1989, 1993 and 1996r1997, as well as observation on the

Fig 1 Chieng Dong in northern Vietnam.

ground Land use was classified into five

cate-gories: wet-rice fields; upland fields; scrub land;

open-canopy forest; and closed-canopy forest

Partial cloud cover on the 1989 image, however,

hampered easy comparison with the other images

The following discussions will therefore use the

data from 1993 and 1996r1997 only This does

not affect the general validity of the reported

land-use trend Comparison with the 1989 image

suggests that the reported increase in forest cover

began before 1993

The research took place in Chieng Dong

mune of Son La province Fig 1 Almost all of

the commune’s inhabitants and a majority of the

provincial population are Black Thai Black Thai

inhabited the region that is today northwestern

Vietnam for centuries Yet the integration of the

previously quite autonomous Black Thai villages

into the Vietnamese state happened only in the

course of the 20th century After 1954, the year of

independence from French rule, the Vietnamese

state extended its administrative structure to Son

La The state recruited Black Thai mostly to staff

communes the lowest administrative unit ,

dis-tricts and the provincial administration of Son La

Until today, almost all Black Thai live off the

land They typically work wet-rice fields in the

valleys dissecting the northwestern mountains They supplement wet-rice cultivation with dry rice, corn and cassava fields, as well as animal hus-bandry on the surrounding slopes The land re-quirements of rice swiddening, in particular, are extensive, as villagers work fields for 2 or 3 years only and then fallow the land for several years By the late 1980s, villagers had expanded rice swid-dens far up the slopes As a result, forests had receded to mountain tops, rocky outcrops, and other land unsuitable for cultivation

4 Forest policy and its implementation in Son La

The Land Law, passed by Vietnam’s National Assembly in 1993, mandated the state to formally allocate land to households under long-term lease arrangements In the case of forestry land, alloca-tion entitled land users to use the land for 50 years, enjoy its benefits, transfer use rights to other entities, and use allocated rights as collat-eral The Land Law was a cornerstone in a quite radical plan to devolve authority over forest man-agement to the local level While state forest enterprises located at the central and provincial levels had previously managed most land, policy

now wanted communes, villages and households

to take their place

Yet the Land Law and subsequent

implement-ing decrees limited the rights given to the local

level First, the state reserved the right of land

classification to the district, provincial, and

cen-tral government levels Once land was classified

as forestry land, its use was solely restricted to

forestry Second, subsequent decrees mandated

that natural forests and forestry land in important

watersheds could not be allocated to households

Instead, the land would be allocated to

com-munes, districts and State Forest Enterprises,

which in turn would sign protection contracts with

households The contracts included small

pay-ments to households for their protection services,

but restricted their use of the forest to minor

forest products

The People’s Committee of Son La province

initiated a campaign for forest protection and

land allocation in 1994 The People’s Committee

called upon district administrations to form task

forces for each commune It also set out strict

guidelines for land classification: the task force

had to zone all existing forest, rocky outcrops, the

top third of each hill and mountain, and 80-m

wide belts on each side of the national road as

forestry land

In Chieng Dong, district cadre, commune

offi-cials, and village leaders formed a task force to

implement the campaign The task force

demar-cated forestry land following the provincial

crite-ria, communicated the concrete boundaries to

villagers, made them commit to abandon fields on

the protected land, and allocated the forestry

land to each village Much of it was scrub land, a

consequence of extensive rice swiddening in

pre-vious years The task force also requested

vil-lagers to organize forest protection groups and

promised payments in the case of successful

pro-tection In addition, it signed up villagers to plant

trees, providing seedlings and cash payments

The district administration initiated a new

round of devolution in 1995, when the district was

selected as a pilot site for the implementation of

the German-funded Social Forestry Development

Project Song Da District cadres returned to the

villages to assign specific parcels of forestry land

to households for protection They concluded protection contracts with individual households and groups of households At the same time, the district cadres sought ways to increase villagers’ commitment to forest protection They negotiated forest protection codes with villages They also fined approximately 90 households that had cont-inued cultivating fields on land zoned as forestry land in the previous year

The number of reported violations against forest protection significantly subsided in the fol-lowing years In 1996 and 1997, the district’s Forest Protection Unit fined only two households

In addition, the Unit expressed satisfaction with villagers’ protection efforts and disbursed all promised payments Villagers’ requests for tree seedlings exceeded nursery capacity in many vil-lages They planted approximately 400 ha between

1995 and 1997 At the same time, forest cover increased from 1993 to 1996r1997, as it did in the

whole of Son La province Fig 2 Had forest land allocation and protection contracts caused vil-lagers to protect forests, and hence led to the expansion of forests?

5 The effects of land allocation and protection contracts on the forest

Land allocation had the goal of providing in-centives for forestry in the uplands The rationale was that villagers would be more interested in forest protection and management if they got formal rights for the benefits from the forest In particular, it was assumed that by allocating rights

to specific parcels to individual villagers, they would be motivated to keep other villagers out Under the previous policy, which reserved control over forestry land to State Forest Enterprises, villagers had not enjoyed such right In addition, the new policy included financial incentives, small cash payments for protection services and rela-tively significant subsidies for tree planting The Vietnamese state hoped that legal and financial incentives would be sufficiently strong to promote forestry

The incentives proved too weak to exclude agri-cultural uses on the land classified as forestry

Fig 2 Forest expansion in Son La and Chieng Dong.

land In the years following allocation, villagers

expanded their fields up the slopes into the

sup-posedly protected top parts Some villagers even

opened up new fields on protected land The

upland area under cultivation in Chieng Dong

1996r1997, as it did in the whole province of Son

La Fig 3

The forest protection contracts did not modify

the ways in which villagers gained access to

forestry land and resources in Chieng Dong

Households with contracts did not prevent others

from using the assigned land to graze their

live-stock, extract wood, collect other forest products,

or expand adjacent upland fields They did not

even stop fellow villagers from opening up small rice swiddens or cassava fields on remote forestry land, as those claimed to need the land to meet their food needs Despite their obligation to re-port violations and the threat to be fined them-selves, they did not inform the Forest Protection Unit Village solidarity was stronger than the state’s reach

Tree planting was popular for two reasons Most villagers were keen to plant trees for the cash payments associated with tree planting A few even planted trees on the land holdings of other households, though there was no guarantee that they would enjoy future benefits A few vil-lagers planted trees on their previous upland fields

Fig 3 Uplands under cultivation in Son La and Chieng Dong.

along National Road No 6, because they were

allowed to intercrop agricultural crops during the

first 2 years Both groups planted trees for their

short-term benefits ᎏ cash and continued access

to the land

It does not come as a surprise then that the

trees did not grow well Neither group was

inter-ested in their long-term benefits For example,

the Forest Protection Unit estimated that only

10᎐15% of the trees planted in 1995 survived the

first year The tree plantations along Road No 6,

in particular, looked very poor The trees were

means to the people, means that they used for

their own goals The potential of the trees to yield

significant long-term benefits was considered low

The larger reasons for the lack of success were

rooted in the relations between villagers and the

state Villagers had claimed control over most of

the uplands during the previous years

Particu-larly during the 1980s, they had opened up new rice swiddens all over the uplands Legal author-ity, which rested with the state, and actual

property relations were far apart Fig 4 As a consequence, land allocation and protection con-tracts did not imply a shift of control toward villagers, but had the potential to weaken vil-lagers’ control Villagers therefore resisted the implementation of the policy

Facing this situation, the district authorities did not invest the resources to enforce state control District cadres looked away from the discrepancy between formal regulations and village practice After 1995, they virtually stopped enforcing forest protection regulations They paid out the full amount of protection payments, though many parcels of forestry land showed visible evidence of agricultural use Even the provincial People’s Committee retracted In 1997, it revoked its

origi-Fig 4 The conflict between village land use and state regulation in 1989.

nal order to zone protection belts along the

natio-nal road

Local cadres accommodated village practice

because they maintained close ties with the

vil-lages Village leaders and commune officials

cont-inued to live in the villages and work in

agricul-ture full-time Many district cadres came from

farm households within the district The ties of

the cadres with the central state authorities were

limited Ethnic differences, difficult topography,

and a lack of infrastructure weakened the

link-ages with Hanoi In addition, the cadres were

hired and paid by the local state authorities, and

they spent their entire professional careers in the

district The close ties with villagers and weak

linkages with Hanoi motivated cadres to

accom-modate village practices, even if these collided

with national policy National objectives seriously

competed with conflicting village interests only

when the central and provincial governments

un-derwent particular effort to enforce their goals,

for example the forest protection campaign of

1994

6 The effects of changing markets and

technologies

Forestry land allocation and protection

con-tracts did not have much effect on village land

and tree tenure If not forest policy, the question

is, what factors caused the increase in forest

cover? The answer lies in the broader dynamics

of land use and the ways in which changing

mar-ket and technological opportunities affected

agri-cultural production Market expansion and newly

available technologies motivated two shifts in

vil-lage land use: a shift in rice production from

swiddens to wet-rice fields and a change in cash

cropping from cassava to corn

In terms of rice production, villagers had relied

on extensive rice swiddening until the end of the

1980s, as it had been more productive than

rice cultivation Fig 5 Around 1990 several

fac-tors increased the relative attractiveness of

wet-rice production New seed varieties and

increas-ingly available chemical fertilizer facilitated an

average yield increase of approximately a quarter

in wet-rice production At the same time, the average productivity declined and weeding re-quirements grew in rice swiddening, as fallow cycles shortened The returns to labor in wet-rice cultivation began to exceed those in rice swidden-ing In 1992, 1 labor day yielded an average of almost 5 kg paddy in wet-rice production, but only

3 kg in rice swiddening Consequently, villagers increasingly shifted their labor to intensify pro-duction on wet-rice fields

Villagers did not only intensify production on the existing fields, but also constructed new ter-races on previous upland fields and fallow land to expand the area of rice sowed Terracing required

an initial investment of household labor Once the terraces were built, labor requirements on the terraces were similar to those on existing fields Starting in 1991, the construction of new terraces exceeded 2 haryear Terracing reached a peak in

1995, when villagers built approximately 10 ha of new terraces As a result of the intensification of existing wet-rice fields and the construction of new fields, rice output dramatically increased

between 1990 and 1997 Fig 5

In later years, market and technological factors continued to favor wet-rice cultivation More fa-vorable prices motivated an increase in fertilizer application rates The district’s input store contin-ued to introduce new rice varieties with higher yields In addition, the store advanced fertilizer to households and collected its cost in paddy after harvest Cuts in taxes and co-operative fees also raised the share of output retained by households from 70 to 85% As a consequence, fertilizer application rates, average yield, and output cont-inued to grow as villagers intensified wet-rice production And as wet-rice production produced sufficient foods to cover subsistence require-ments, rice swiddening declined drastically Changing markets and technologies also moti-vated a shift in cash cropping from cassava to

corn Fig 6 Markets for agricultural output rapidly expanded in the late 1980s, when the Vietnamese state abandoned mandatory procure-ment Traders began to arrive in larger numbers

in search of agricultural crops to satisfy the de-mand for animal feed in the lowlands The prices

Fig 5 The shift from rice swiddens to wet-rice fields in Chieng Dong.

they offered became increasingly favorable to

corn Corn prices rose against cassava,

particu-larly in 1995 and 1996 In terms of technology,

new corn varieties became available in the 1990s

They allowed corn yields to triple within a few

years, without the use of fertilizer and pesticides

By 1993, the returns for corn cropping exceeded

those of cassava cultivation: 1 labor day produced

an average return of VND 20 000 in corn

produc-tion, compared to VND 15 000 in cassava In

addition, corn output was more secure, as the

maturation period of corn was much shorter

In sum, changes in markets and available

tech-nologies changed the relative costs and benefits

of agricultural crops In response, rice production shifted from swiddens to wet-rice fields Corn replaced cassava as the major cash crop The shift

in land use reduced the demand for land, as wet-rice and corn production took place on per-manent fields They required a smaller total area

of land for the same annual area under cultiva-tion, as villagers had previously rotated rice swid-dens and cassava fields As a result, forests quickly regenerated in Chieng Dong, as in the whole of Son La, given the favorable natural growth

tions see Fig 2 The expansion was due to an increase in open-canopy forest Closed-canopy forest remained unchanged The forest grew back,

Fig 6 The shift from cassava to corn in Chieng Dong.

although the actual area under cultivation in a

given year expanded

7 Summary and conclusions: the limits of

devolution

The Vietnamese government radically changed

forest policy in the early 1990s The new policy

mandated the devolution of authority over forest

management to the local level The People’s

Committee of Son La province implemented

forest land allocation in 1994 Yet the concurrent

expansion of forest cover was mainly due to other

factors Changing markets and newly available

technology motivated villagers to intensify land

use Forest policy reform had little effect because

policy implementation did not affect actual

property rights Village land and tree tenure

per-sisted because villagers reper-sisted the

implementa-tion of the new policy, and because local state

cadres did not enforce it In addition, tree

plant-ing met with little success, and the payments

associated with tree planting and forest

protec-tion were small in relaprotec-tion to agricultural income

The findings demonstrate how forests are

em-bedded in the broader dynamics of land use The

changing forests of Son La reflected the

combina-tion of a variety of economic, technological and

political factors These factors shaped the set of

opportunities available to villagers, as well as the

relative costs and benefits of those opportunities

When key economic and technological factors

changed, they modified the relative costs and

benefits of opportunities Villagers

correspond-ingly changed their allocation of labor, which

affected land use and forests In contrast, the new

forest policy did not affect greatly labor

alloca-tion, as it left opportunity sets virtually

un-changed

The findings also indicate how the effects of

forest policy and other large-scale forces are

spe-cific to local contexts They highlight various

fac-tors that led to particular policy outcomes in

Chieng Dong and Son La, in particular the nature

of village᎐state relations, social relations among

villagers, favorable access to product markets, the

absence of off-farm employment, and the

avail-ability of credit for fertilizer Policy outcomes will

be different in other local contexts

In Son La, policy makers at the national and provincial level have several options to promote forest regeneration They could improve the state’s capacity at enforcement Measures that improve local cadres’ commitment to national objectives could strengthen the implementation

of forest policy Combining enforcement efforts with incentives, such as preferential access to credit, could improve their attractiveness to vil-lagers Second, market-based intervention could alleviate agricultural pressure on marginal land Suitable instruments might include taxes on land under cultivation, subsidies for wet-rice cultiva-tion, and credit support for the improvement of rice terraces The state could also use public investment into infrastructure and industry pro-jects to create off-farm employment and increase off-farm wages Third, directing agricultural re-search toward labor-intensive technology could reduce the demand for land Continuing intensi-fication will satisfy the demands and absorb the labor of a rapidly rising population New rice varieties that tolerate seasonal water shortages or cold spells could facilitate a further expansion of rice terraces into land considered unsuitable un-der current technological conditions

Most likely, the best policy includes elements

of all three strategies and provides flexibility to respond to different conditions The tree planting along the national road serves as an example of bad targeting Provincial policy-makers prioritized forest protection on land that was highly attrac-tive for cultivation Economic and technological forces provided strong motivation for villagers to use the land for cash cropping The same economic and technological forces, however, ame-liorated cultivation pressures in more remote ar-eas The potential of the new forest policy lies in those areas It is there that land allocation and protection contracts may tip the balance of rela-tive costs and benefits in favor of forestry The research findings attest to the importance

of in-depth village studies for forest policy re-search Local property rights may display signifi-cant discrepancies from national legislation A policy that promotes the devolution of forest