Title Page
 Table of Contents
 List of Tables
 FCND discussion papers

Group Title: FCND discussion paper
Title: Pathways of rural development in Madagascar
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00089742/00001
 Material Information
Title: Pathways of rural development in Madagascar an empirical investigation of the critical triangle of environmental sustainability, economic growth, and poverty alleviation
Series Title: FCND discussion paper - International Food Policy Research Institute ; 82
Physical Description: vii, 48 p. : graphs, tables. ;
Language: English
Creator: Zeller, Manfred
International Food Policy Research Institute -- Food Consumption and Nutrition Division
Donor: Peter E. Hildebrand ( endowment ) ( endowment )
Publisher: International Food Policy Research Institute
Place of Publication: Washington, D.C.
Publication Date: 2000
Copyright Date: 2000
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Spatial Coverage: Madagascar
Bibliography: Includes bibliographical references.
General Note: "March 2000."
Statement of Responsibility: Manfred Zeller ... et al..
 Record Information
Bibliographic ID: UF00089742
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 50385691

Table of Contents
    Title Page
        Title Page
        Page iii
        Page iv
    Table of Contents
        Page v
    List of Tables
        Page vi
        Page vii
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
    FCND discussion papers
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
Full Text

FCND DP No. 82


Food Consumption and Nutrition Division

International Food Policy Research Institute
2033 K Street, N.W.
Washington, D.C. 20006 U.S.A.
(202) 862-5600
Fax: (202) 467-4439

March 2000

FCND Dscuwssion Papers contain preliminary material and research results, and are circulated prior to a full
peer review In order to stimulate discussion and critical comment. It is expected that most Discussion Papers
will eventually be published In some other form, and that their content may also be revised.



This paper is based on community-level data from 188 villages in rural Madagascar.

The survey that was conducted in 1997 made extensive use of long-term recall questions

ascertaining changes during the past 10 years in rice yields, wages, population, soil

fertility, and other pertinent variables of rural development. We find that-on average for

all villages-the yields of irrigated rice, the major food crop, and real agricultural wages

declined, while the communities expanded their upland area by nearly a quarter and

experienced deteriorating fertility of their upland soils. These patterns are consistent with

the wide-held belief that rural areas in Madagascar have witnessed increased poverty,

economic stagnation, and a continued degradation of the natural resources.

Yet, the five agroecological regions in our sample exhibit quite different patterns of

rural development, and at least one of them has experienced increases in yields and wages.

From a policy perspective, it is important to better understand the driving forces of such

diverse rural change. The overall decline in rural wages over the past 10 years is expected

to have contributed to increased poverty, food insecurity and malnutrition in rural areas, as

rural wage laborers traditionally belong to the poorest of the poor in Madagascar. In this

paper, we present an econometric analysis of the determinants of and interdependencies

between the three components of sustainable development: economic growth,

environmental sustainability, and poverty alleviation.


We develop a two-stage, least-squares fixed-effect model that attempts to explain

rice yields, rural migration and wages, endogenous placement of microfinance institutions,

and the observed change in upland and soil fertility, with exogenous and policy variables,

such as the communities' social capital, their exposure to weather risks and their access to

commodity and financial markets and to public services. The paper concludes with a

number of implications for policy and future research.


A cknowledgm ents .................................................... vii

1. Introduction ....................................................... 1

2. Conceptual Framework ............................................... 1

Induced (Technological and Institutional) Innovation Theory ................ 1
The Critical Triangle of Rural Development: The Case of Madagascar ......... 2

3. Changes in Rural Development: A Descriptive Analysis ...................... 4

Data Source and Sampling Frame ..................................... 4
Major Trends Characterizing the Critical Triangle of Rural Development ....... 5
The Expansion of the Agricultural Frontier in Rural Madagascar ............. 7
The Three Major Pathways of Rural Development in Madagascar ............ 9
Trends in Access to Financial and Commodity Markets and to Public Services .. 10
Access to Microfinance Institutions ............................. 10
Access to Agricultural Input and Output Markets ................... 11
Community Access to Public Services ........................... 12

4. A Causal Diagram of the Critical Triangle ............................... 12

5. Econometric Model and its Results ..................................... 18

Determinants of Community Access to Member-based Financial Institutions ... 20
Determinants of Lowland Rice Yields ................................ 21
Determinants of Migration ......................................... 23
Determinants of Rural Wages ....................................... 25
Determinants of Expansion of Upland ................................ 26
Determinants of Soil Fertility of Upland .............................. 28

6. C conclusions ...................................................... 31

T ables ............................................................. 35

R references .......................................................... 45


1 Indicators of economic growth, poverty and resource degradation: Rice yields,
wages for rural laborers, and fertility of upland soil, by region and year ....... 37

2 Indices of change in area of grassland, forest, arable upland and irrigated
lowland during past 10 years ....................................... 38

3 Participation in microfinance programs and travel time to financial
institutions,\ by region and year ...........................

4 Descriptive statistics of variables used in regressions (n = 376 from
188 villages) .........................................

5 Determinants of participation in member-based financial institutions

6 Determinants of yields of irrigated rice ......................

7 Determinants of migration ...............................

8 Determinants of rural wage rate ...........................

9 Determinants of expansion of arable upland ..................

10 Determinants of soil fertility on upland .....................

.......... 38

.......... 39

. . . . .. 40

. . . . .. 41

.......... 41

.......... 42

. . . . .. 42

.......... 43


1 The critical triangle of rural development: Pathways and causes ............. 14


The funding of this research by the United States Agency for International

Development (USAID) is gratefully acknowledged. The authors thank John Pender and

Sara Scherr for providing comments on the design of the questionnaire modules covering

forest and land resources. Thanks are also extended to Steve Vosti for reviewing this


Manfred Zeller
C6cile Lapenu
Bart Minten
International Food Policy Research Institute

Eliane Ralison
Ddsir6 Randrianaivo
Claude Randrianarisoa
Ministare de la Recherche Scientifique (MRS)
Antananarivo, Madagascar


Most of the world's poor and food insecure depend directly or indirectly on

agriculture for their livelihood. Much-needed increases in agricultural production can in

principal come about through two pathways of rural development: expansion of cultivated

area or increases in yields. We term the former the pathway of agricultural extensification

since it does not require increases in agricultural productivity by using modem inputs, such

as high-yielding seeds and mineral fertilizers. Shifting cultivation, by burning tropical

forests for gaining arable land, is a prime example for the first pathway. The second is

called a pathway of agricultural intensification (Boserup 1965; Pingali, Bigot, and

Binswanger 1987; Ruthenberg 1980). Which factors induce rural communities to follow a

certain pathway of use of land and forest resources, and how can policy assist to achieve

environmentally sustainable pathways of development while at the same time improve

incomes and alleviate poverty? In this paper, we seek to shed some light on this complex

question, using a descriptive and econometric analysis of community-level data from




Induced innovation theory suggests that environmental degradation can be self-

correcting, as population growth, increasing resource scarcity, and environmental

externalities induce new agricultural and resource management practices (Boserup 1965)

and new forms of collective regulation of common property resources (Kikuchi and

Hayami 1980; Ruttan and Hayami 1984). However, this early work on induced

institutional innovation is overly optimistic, and tends to overlook, as pointed out by

Baland and Platteau (1998), that evolution, including long-run institutional equilibria and

the corresponding pathways of development, are themselves dependent on the initial stock

of social capital, on the actions by the state, and on the distributive consequences of

institutional change.

Ruthenberg's (1980) survey of literature on farming systems in the tropics

documents many agricultural innovations that were associated with increasing population

growth and market integration in different agroecological zones. He explains the observed

technical changes in crop and soil management with the occurrence of increased scarcity

of land and declining soil fertility. More recent literature, reviewed by Scherr and Hazell

(1994), has described similar endogenous processes leading to pathways of agricultural



Vosti and Reardon (1997) conceptualize a critical triangle that links three

development objectives: economic growth, poverty alleviation, and environmental

sustainability. These authors emphasize the importance of simultaneous consideration of

all three development objectives, plus possible links between and trade-offs among them.

In recent years, Madagascar has become internationally known for its rich and

unique bio-diversity that is threatened by rapid deforestation (Jarosz 1993). Since the mid-

1980s, Madagascar has been a focus country of international conservation efforts when

international development organizations provided loan and assistance programs that

explicitly aimed to attain environmental objectives. Already in 1984, the Government of

Madagascar embarked on a National Strategy for Conservation and Development that

recognized the critical triangle by including environmental conservation, economic

development, and human needs into the policy framework (Larson 1994). However, as

pointed out by Gezon (1997), a number of important environmental projects implemented

during the early 1990s sought to address the links between poverty and environmental

degradation, but failed to take into account their economic sustainability.

The national survey Enquate Permanente aupr6s des M6nages (EPM), conducted in

1993 by the National Statistical Office (INSTAT), has a few questions that touch upon the

critical triangle. Apart from some in-depth surveys in or near protected areas, no

statistically representative data on a national or regional scale exist to the knowledge of

the authors that contain both agroecological and socioeconomic data and that can be

meaningfully exploited for causal analysis. Yet, the existence of such combined databases

is a precondition for in-depth analysis of the underlying determinants of rural development,

economic growth, poverty alleviation and environmental sustainability. One recent data

set collected by the authors attempts to combine these topics, and is the basis for the

analysis of this paper.



In May 1997, the International Food Policy Research Institute (IFPRI) and the

Ministry of Scientific Research (MRS) of Madagascar undertook a survey that covered

188 communities in five agroecological regions: Fianarantsoa Highlands, Fianarantsoa

coastal and escarpment region, Majunga Lowland area, Majunga Highland area, and the

Vakinankaratra, a diverse region in the central Highlands of Madagascar. While the

general purpose of the survey was to document past trends in rural development and to

explain how they have responded to the liberalization of agricultural commodity and rural

financial markets, several sections were added to cover changes in the natural resource

base.' The respondents to the survey were the mayor of the village and other community

leaders. Common to all modules was that the respondents were asked to provide

information for the situation prevailing in 1997 and 10 years ago.

In order to improve the efficiency of the random sampling procedure, two

stratifying variables were used: the size of the village (above or below median of all

villages in the agroecological region), and the terciles of the distance from the village to

The authors thank John Pender and Sara Scherr for providing comments on the design of the
questionnaire modules covering forest and land resources.

the next national road. These stratifiers are exogenous variables for short-run rural

development, and influence communities' access to commodity and financial markets,

information, preventative and curative health care, education, and other public services.


Rice is the major crop and also the staple food in Madagascar. For the nation as a

whole, about half of calories consumed are from rice (Ravalosoa 1998; SECALINE

1996). Rice is mainly grown on irrigated lowland fields, or to a small extent on rainfed

upland plots. In Table 1, the yield (in kilograms per hectare) of irrigated rice is shown for

the agroecological regions in the sample. For all regions, average rice yields declined from

1,765 kilograms per hectare to about 1,540 kilograms per hectare between 1987 and

1997. Yet, important differences across regions exist. In the Vakinankaratra region, the

yields remained the same. Major declines in rice yields can be noticed for the Majunga

regions (Minten, Randrianarisoa, and Zeller 1998). Insofar as rice yields can be viewed as

a performance indicator for the agricultural sector as a whole, we conclude that there was

either stagnation or decline in productivity.

In the community survey, we sought to ask about the changes in poverty over the

past 10 years. Poverty encompasses many dimensions, such as food insecurity,

malnutrition, and illiteracy. Most of them are difficult to measure through community-

level surveys. Moreover, we were interested in the change of poverty over the past 10

years, and, therefore, had to find variables that could be easily recalled by our respondents.

One such variable, we believe, is the rural wage rate. In Madagascar, rural wages are

often paid in rice, and measured as the number of kapoaka2 paid per day. Table 1

differentiates wage rates by agroecological regions. We note that wage rates have

declined over the past 10 years for all regions as a whole, but important differences among

the regions exist. The wage rate improved in the Vakinankaratra region, whereas the

other regions experienced modest (Fianarantsoa) or severe declines in wage rates

(Majunga Plateau). Insofar as wage rates determine the ability of the poorest of the rural

poor to feed their families in Madagascar, we conclude that food insecurity and rural

poverty increased.

The change in soil fertility is the third indicator shown in Table 1, reflecting the

environmental objective of the critical triangle. Community leaders were asked how the

fertility of the upland (tanety) in their community has changed. The variable "soil fertility"

ranges from 3, if soil fertility improved over the past 10 years; 2, if no change; 1, if soils

degraded somewhat; and 0, if soils became severely degraded. We find that soil fertility

significantly decreased, on average, for all regions, with major degradation observed in the

Fianarantsoa Highlands (HT), the Vakinankaratra and the coastal and escarpment area of

the Faritany of Fianarantsoa. The rice yield, wage rate, and soil fertility variables serve in

our analysis as indicators of the trends in agricultural growth, in poverty, and in

2 This common measure is the tin can used by Nestle Company for its condensed milk, which holds
285 grams of white rice.

environmental degradation, respectively. Overall, the data reveal quite disturbing trends.

Only one of the five regions improved in two of the three criteria over the 10-year period.


The yields of rice have stagnated or declined in Madagascar during the past 10

years. This is not evidence that an intensification strategy had been chosen by the rural

households and communities. Table 2 supports this notion. While rice land area grew by

about 5.3 percent as an average for all regions, the area of cultivated upland (tanety)

increased by about 24.0 percent. Considering that the population in the survey regions

grew by about 3.5 percent per year over this time, or by more than 35 percent over the

past 10 years, the rate of growth in arable area is actually below that of the growth in


The expansion of the agricultural frontier came at the expense of forestland,

bushland, and grassland. These land types are usually held as common property resources

but become, under Malagasy customary law, private property when cultivated successively

3 The natural population growth rate of 3.5 percent, which was measured in the community-level
survey through recall questions on the number of people residing 10 years ago and now in the village,
correcting for the number of people in-migrating and out-migrating, is considerably higher than the one
estimated at 2.9 percent for the country for the period of 1980 to 1989. We note that our growth rate is based
on quite simple, and therefore, imprecise questions on population size of village 10 years ago and now.
Growth rates, obtained from census data at two points in time, are admittedly much more exact. Second, the
rate reported here is not weighted by the initial size of the village (in comparison with other villages) so that
high population growth rates observed in small villages biases our estimate upwards. When we weigh the
population growth rate by the size of each village in the sample, the weighted population growth rate for all
survey regions as an aggregate is measured at 3.0 percent. In other words, when correcting for the faster
population growth in smaller villages, our estimates coincide very well with the national ones.

by the same family.4 On average for all regions, the largest losses were experienced for

area under primary forests (Table 2). Indicating the level of primary forests as 100 for

1987, the index fell to 67.4 in 1997. The loss in secondary forest amounts to about 28

percent, whereas the loss in grassland and bushland (kijana) was about 20 percent. If this

rate of deforestation continues, the survey regions could be without primary forests in two

to three decades.5

That the agricultural frontier can still be expanded in many Malagasy villages is

evident from the survey data On average for all regions, 59 percent of villages report that

there is additional land available for expanding upland cultivation. Fifty percent of villages

have such possibilities for expanding irrigated land (tanimbary), and 35 percent of villages

dispose over possibilities to expand both types of land. Despite this, the average holding of

upland per household declined from 1.74 hectares in 1987 to 1.28 hectares in 1997.

Clearly, growth in population has outpaced growth of agricultural land in many

communities, and pressures for agricultural intensification have tended to increase over



4 On land law in Madagascar, see Keck, Sharma, and Feder (1994), Rakotomanga (1976), and
Rakotonirainy (1984).

5 This result coincides with an estimate by the Economist Intelligence Unit in 1990, cited from Keck,
Sharma, and Feder (1994).

We conclude from this analysis that about half of the villages in the survey regions

will still be able to choose the extensification pathway for the near and distant future.

These more land-abundant communities may maintain their living standards with their low-

input agriculture for some time, while continuing to expand arable area and degrade soils.

Yet, while such extensive expansion of agricultural production may meet subsistence

needs in the short run, the continued degradation of forests, watersheds, and soils has

implications for future agricultural productivity, health, and nutrition (von Braun 1997).

Current subsistence and food security are then likely to be at the expense of future

generations. The households are most certainly aware of these trade-offs; yet, extreme

poverty can lead to high time preference rates. For example, Lapenu et al. (1998) show

that households in a significant number of sample villages needed more time to collect

firewood now compared to 10 years ago. In these villages with increasing firewood

shortages, the need for reforestation was perceived more frequently than in villages with

below-average time needed to collect firewood.

The remaining half of the villages does not have the possibility of further land

expansion anymore. In order to maintain or improve their living standards, households in

these land-constrained villages can in principal pursue two other generic development

pathways: (1) agricultural intensification, and related increased trade and diversification

into off-farm enterprises, or (2) migration. The main, and certainly most promising and

viable strategy in the long run, is to intensify agriculture, and to make it more productive

and environmentally sustainable.6 The third strategic response to land scarcity is to

migrate, i.e., to leave the village seasonally or forever. Its consequences for poverty

alleviation and the environment can be quite different from an agricultural intensification

strategy, as is shown later.


A number of factors that are considered to have exogenously evolved during the

period 1987-1997 are expected to have influenced the phenomena that villages differ in

their development paths. These factors include access to financial and to agricultural input

and output markets, and access to public services.

Access to Microfinance Institutions

As in most low-income countries in Sub-Saharan Africa, the formal banking sector

has barely penetrated rural Madagascar. Until 1986, the state-owned Banque pour le

D6veloppement Rural (BTM) was the main institution providing loans to farmers. Apart

from the requirement of tedious paperwork and collateral that poor households neither can

'Higher rural incomes could possibly be achieved by growth in the rural nonfarm sector. While this
paper focuses on agricultural income growth, the promotion of nonfarm enterprises (for example, for food
processing and marketing and for provision of rural services) can be an important strategy for overall rural
growth. However, the success of this strategy depends to a large extent on the relative magnitude of forward
and backward linkages of the (dominant) farm sector with the nonfarm sector. Islam (1997) finds that the
growth multiplier was estimated between 1.35 to 1.90, i.e., $1 of additional farm income increases nonfarm
rural income by $0.35 to $0.90. On farm-nonfarm growth linkages, see, for example, Hazell and R6ell (1983)
and Hazell and Haggblade (1993).

understand nor provide, the one-way travel time of the average smallholder to the nearest

bank branch is 4.75 hours. In other words, it takes an entire day to apply for a loan. The

travel times to banks, post offices, and postal savings bank also increased during the past

10 years (Table 3). In the early 1990s, a number of village bank programs and credit and

savings cooperative societies were introduced and subsequently expanded to better reach

rural smallholders, especially in the Vakinankaratra, the Fianarantsoa Highlands, at Lake

Alaotra and in Marovaoy (Fraslin 1997). The survey obtained information on the existence

and size of member-owned financial institutions in the village that allowed defining a

variable that measures the density of village-level participation in financial institutions.

This information was only obtained for 1997, but it is safe to assume that none of the

villages had access to similar organizations in 1987 because they were all formed after that

year. Eleven percent of households participate in member-based financial institutions in


Access to Agricultural Input and Output Markets

Survey data not presented here for lack of space show that access to markets for

seed rice and farm equipment improved, while access to mineral fertilizer, pesticides,

herbicides, vegetable seeds, and veterinary inputs worsened over the survey period.

Particularly noteworthy is the large decline in access to fertilizer markets. On the other

hand, the access to agricultural output markets improved somewhat, in particular for rice.

Community Access to Public Services

We asked about the travel time from the village to over 15 public services that are

provided by central or local government The travel time in the rainy season-using the

commonly used modes of transport in the village--ranges between two and four hours.

Travel time to access a public phone booth is highest among all public services. It takes

an average of two hours to travel to either the nearest primary and secondary school, or

the nearest health post or hospital. Compared to 1987, the average travel times for all

regions taken as a whole in 1997 are either the same or slightly higher.


Three major pathways of development are distinguished in this analysis. The first is

the extensification pathway, that is, to continue with low-input, low-output agricultural

technology by expanding cultivated land area and subsequently reducing fallow years until

soil fertility has been depleted. The second pathway is to intensify agricultural production

and increase yields by exploiting access to financial and commodity markets, thus adopting

more capital-intensive, higher-yielding technology and reaping the benefits of interregional

specialization. Migration is the third major pathway that we observe from the data. The

migration pathway is similar to the extensification pathway: migrants are attracted by the

possibility of farming new land, as we will confirm in the econometric analysis, since the

pressure on natural resources is simply shifted from one region to another.

Hence, our conceptual framework stipulates four endogenous outcome variables:

rice yield, wage rate, changes in soil fertility, and in the area of upland. These are

influenced by three strategies leading to different pathways of rural development. The

villagers' choice of strategy is, of course, endogenous as it is influenced by the

endogenous wage rate, rice yields, and soil fertility, by the socioeconomic and

agroecological characteristics of the region, by the access to financial and commodity

markets and public services, and by the various institutional arrangements that regulate the

use of common and privately-owned forestland, grassland, and arable land resources.

Figure 1 presents a causal diagram of the critical triangle of rural development.

Arrows between the variables represent relationships. A positive sign implies that an

increase in the value of the variable from which the arrow originates is expected to induce

an increase in the value of the other variable. For example, we hypothesize that higher

lowland rice yields will decrease the incentives for further expansion of upland since the

returns to lowland rice production become relatively higher. This relationship is reflected

by an arrow with a negative sign.

Let us consider first the arrows that feed into the upland area from the other

endogenous outcomes or strategies. These arrows come from access to financial markets,

the wage rate, the soil fertility of upland, and the net immigration rate. First, improved

access to financial markets-reflected either through lower interest rates, lower

Figure 1-The critical triangle of rural development: Pathways and causes


Participation in member-based
micro-finance institutions

L, A +

Immigration rate -

p Change in upland area

transaction costs for the clients, or higher amounts available to borrow-will make the

capital used in rice production and in upland fields less costly. With the reduced cost of

capital, the profitability of these enterprises rises, leading to increases in yields in both

lowland and upland areas. Second, with increasing wage rates, farm households will tend

to choose more labor-extensive crop production technologies. Since cultivation of

irrigated rice is more labor-intensive than crop production, higher wage rates are

hypothesized to induce lower labor input and therefore lower yields of rice, but increased

expansion of upland area. Third, the expansion of upland area is expected to be positively

influenced by a higher net immigration into the village, since migrants are expected to

cultivate new lands not already occupied by villagers. Fourth, and last, expansion of

upland cultivation will have negative effects on soil fertility due to shorter fallow periods

for already cultivated land or result in increased cultivation of marginal or infertile land.

The negative arrow from soil fertility to upland area change reflects the vicious,

unsustainable cycle of extensification and soil degradation: lower soil fertility will induce

farmers to abandon fields after only a few harvests and seek new farmland to cultivate.

Since most investments in conserving soil fertility are labor-intensive (for example,

spreading manure, planting trees, or terracing land), rising labor costs will reduce the

relative profitability of soil-conserving investments. Moreover, we hypothesize that

increased participation in member-based financial institutions will improve soil fertility

because it reduces the cost of capital and increase its availability for investment in


In the causal model of Figure 1, the net immigration rate is perceived to be a

function of the endogenous wage rate. We hypothesize that villages with high wage rates

will attract migrants. As more migrants come into those high-wage villages, they will

increase the supply of labor and therefore push wage rates down, an effect that is

expressed through a negative feedback loop. In addition to immigration, wage rates are

hypothesized to be influenced by two other endogenous variables: access to financial

markets and lowland rice yields. As access to financial markets improves, the cost of

capital to the community and its households decreases, thus increasing the marginal return

to labor and pushing wage rates up. Similarly, if yields of irrigated rice (a labor-intensive

crop) increase, the marginal return to labor used in rice production increases, resulting in

an increase in wages.

Lowland rice yields are influenced by four variables. Two of them, improved access

to financial markets and wage rates, have already been discussed. We further expect that

increases in crop production on upland will reduce incentives for households to intensify

rice production, in particular if rice production is more capital and labor-intensive than

upland production and if ample opportunities to expand upland continue to exist. The

same type of argument can be made for the expected effect of soil fertility of upland on

lowland rice yields. If upland soils are fertile and productive, and one can get good

harvests on upland, little economic incentive exists for increasing yields of irrigated rice.

We choose to model the access to member-based financial institutions as an

endogenous outcome variable for the following reasons. First, as mentioned above,

financial institutions that reach smallholders in Madagascar have only been formed during

the past 10 years. This is the period of analysis. The placement of financial institutions is

driven by a variety of factors that determine directly or indirectly the loan repayment

performance or the transaction costs for the financial institution and the client (Sharma

and Zeller 1999). Second, as Pitt, Rosenzweig, and Gibbons (1995) demonstrate, failing

to correct for the possible nonrandom placement of government programs can result in

substantial biases in the estimates of the programs' effects. In the case of Madagascar, one

can observe that placement of financial institutions was concentrated mostly in areas with

high agricultural potential and above average infrastructure. More particularly, we

hypothesize that lowland rice yields are a good indicator of agricultural productivity, and

that financial institutions tend to place programs in villages with above-average lowland

rice yields in order to increase their performance. There are, as Pitt, Rosenzweig, and

Gibbons (1995) argue, other, unobserved variables that determine the placement of

programs. In this case, simple ordinary least squares of cross-section data may

overestimate or underestimate the effect of access to credit on yields. In order to control

for endogenous placement of programs, Pitt, Rosenzweig, and Gibbons (1995) use a

fixed-effect model that sweeps out the unobservable characteristics.

The causal diagram in Figure 1 could be further extended by including additional

endogenous variables if one would wish to extend the time horizon of the analysis beyond

the 10 years considered in this paper. In the very long run, everything is endogenous. We

choose to conceive only the six variables shown in Figure 1 to be endogenous during the

time span of our analysis. The main reason is that the 10-year period is sufficiently short

so that access to public services and infrastructure is not likely to change much because of

endogenous processes.


With information on endogenous strategies and outcomes at two points in time, the

potential biases resulting from unobservable characteristics can be avoided by using a

fixed-effect model (Pitt, Rosenzweig, and Gibbons 1995). Figure 1 shows a total of six

endogenous variables that depend on each other. To obtain consistent and unbiased

estimates, we apply a two-stage least squares model with fixed effects (Greene 1995).7

The structural equations of the model are

(1) Participation in financial institutions (Yi) = f(Y2, X, BANK),

(2) Rice yield (Y,) = f (YI, Y4, Y6, X,, TSTRIZ, PEREAU),

(3) Net immigration rate (Y3) = f (Y1, Y4, X,, TRANHUMD,


(4) Wage rate (Y4) = f (Y1, Y2, X,, SHARINCO),

7 The model is estimated with LIMDEP, using a two-way fixed-effects model that, in addition to
village-specific fixed effects, measures time-specific effects of the two years, 1987 and 1997 (Greene 1995).

(5) Change in upland (Y,) = (Y1, Y2, Y3, Y4, Y6, X,, TSTPAT,


(6) Change in soil fertility (Y6) = f (Y,, Y2, Y,, X,, CONFLICT),

where X, is a vector of exogenous variables, and Y( (for i = 1,2,3,4,5 and 6) are the

endogenous variables. For brevity, the variables contained inXi are only shown in the

regression results itself. They have been included in the regression as hypothesized causal

determinants of the independent variables, following the conceptual framework outlined in

the previous sections. The variables that are explicitly named in above equations serve as

instrumenting variables to identify the estimation.' The name, definition, mean, median,

and standard deviation of all variables used in the regression model are shown in Table 4.

Tables 5-10 list the results of the regressions for each of the six equations (1)-(6),

respectively. With very few exceptions, all signs of regression coefficients are as

hypothesized. In general, the models for participation in member-based financial

institutions (Table 5) and for net immigration (Table 7) have a low adjusted R2 of 0.15 and

0.16, respectively. The other models have a higher explanatory power with a R-squared

ranging from 0.34 (Table 9) for the change in upland to 0.68 for rice yield (Table 6).

8 To test whether these instruments are good ones, an F-test was performed to test the joint significance
of the instrumental variables in the first-stage regression. The F-statistic is the ratio of the explained variation
in the dependent variable and the unexplained variation, adjusted for the number of independent variables.
The probability of error resulting from the F-test for the estimated coefficients of the instrumental variables
is 0.74 for equation (1) (OUTREACH), 0.07 for equation (2) (rice yield), 0.00 for equation (3) (net
immigration RNETIMIG), 0.08 for equation (4) (wage rate), 0.02 for equation (5) (change in upland), and
0.49 for equation (6) (soil fertility).


The participation can be perceived as the joint outcome of determinants of supply

and demand for financial services. The model estimates these determinants in a reduced

form as we are not interested here in studying the formation of financial institutions (Table

5). The regression results indicate that formal market participation significantly increases

with higher yields (YLDRICI) in the village and with higher social capital (SOCCAP).

The former effect is likely to be a supply-driven one as program managers seek to place

institutions in villages with high economic returns so as to improve loan repayment rates.

As most credit programs target agricultural enterprises (Zeller 1998), the yields of rice as

the major crop can be a good predictor of such returns. The effect of social capital is

interpreted as a demand-side effect: social capital encourages and promotes villagers'

participation in semi-formal credit groups or cooperative societies. Social capital is

measured by the number of informal self-help groups, multiplied by the number of years of

their existence during the previous 10 years. Informal group action is expected to benefit

participation in other groups since experiences gained in the formation and management of

informal groups make it easier for community members to perform similar functions in

member-based microfinance institutions. The identifying variable in the regression is the

travel time from the village to the nearest bank branch (BANK). Its coefficient has the

expected negative sign but is highly insignificant All credit programs in Madagascar rely

to some extent on branches of the commercial or the agricultural banks for depositing

savings, and, in many instances, for channeling lending funds from the donor to the village

groups. Since the travel time from the village network to the bank influences the

transactions costs of the program, and since credit programs are presumably concerned

about cost recovery, villages close to a bank branch are found to have a higher chance to

obtain access to credit.

We conclude from this regression that villages with higher agricultural productivity

seem to be preferred by financial institutions since most of them lend for agricultural and

food-processing activities.


Table 6 shows the results of the regression concerning rice yields. Again, all signs

are as expected. Yields significantly increase with improved access to credit

(OUTREACH). The coefficient is very robust to alternative specifications of the model,

and is significant at the 1 percent level. The direct marginal effect of an increase of 1

percent more households in the village being a member of a microfinance institution raises

average lowland rice yields at the community level by 20 kilograms per hectare. This

yields an elasticity of lowland rice yields with respect to credit access of 0.069. This

translates into an elasticity of 0.07. This positive and significant effect would not occur if

capital were not a binding factor in the farm sector of Madagascar. As Zeller (1994)

shows, informal and formal lenders alike frequently ration loans, and about half of the

rural sample households have been subjected to credit rationing.

The access to fertilizer markets (A34ACC1) has the expected positive effect on

lowland rice yields, but is insignificant. When multiplying access to fertilizer (A34ACC1)

with access to financial markets (OUTREACH), the interaction term (OUTRFERT) has a

negative coefficient significant at the 10 percent level. This implies that the effect of

access to credit on yields diminishes at the margin as access to fertilizer markets improves.

Similar reasoning holds for the effect of fertilizer when the access to capital markets

improves. This result suggests that access to fertilizer dealers acts as a substitute for

access to financial institutions, and vice versa. Indeed, fertilizer dealers have been found

to advance loans for fertilizer, and obtain repayment when buying the harvest.9 Moreover,

some of the microfinance programs provide credit in kind, mostly in the form of seed and


This finding is in line with the negative, though highly insignificant, coefficient of

the change in upland area (CHTAN) and soil fertility of upland (SOILFERT). With higher

soil fertility of upland, mostly found in the hillsides and slopes, and a greater possibility of

expanding upland, the lower is the incentive to intensify rice production, for example,

through investments in improving existing irrigation and drainage systems. Similar results

are obtained for the possibility of expanding irrigated lowlands (TSTRIZI). If a

community still has additional area that can be terraced for irrigation, the yields of rice are,

on average, 95 kilograms per hectare lower than in land-constrained villages. Again, if

further land expansion is possible-be it upland or lowland, less incentives for agricultural

9 On interlinked credit-cum-marketing transactions in rural Madagascar, see Barrett (1997).

intensification exist These results empirically confirm Boserup's induced innovation

model: when land becomes scarce and its opportunity cost rises, rural households and

communities seek to adopt technology and alter their institutional arrangements for the

management of land in order to increase its productivity and return. As expected, higher

wages (WAGERATE) reduce the yield of labor-intensive rice, although the coefficient is

insignificant The elasticity of rice yield with respect to wage rate is measured as 0.12. In

summary, the most important determinants of the rice yield are the access to financial

markets and the availability of idle land that can be used for expanding cultivation.


The dependent variable net immigration (RNETIMIG) is the average annual rate of

net immigration into a village over the previous 10 years. Table 7 shows the regression

results. As expected, migrants respond to wage differentials, although this relationship is

highly insignificant. The elasticity of immigration with respect to wage rates is measured

at 1.74, that is, a 1 percent increase in wage rates in a village increases the net

immigration rate into this village by 1.74 percent. In contrast, the net immigration rate

(with a mean of 0.13 percent per year for all villages) significantly increases by an absolute

0.66 percent if a village has possibility to both expand upland and irrigated lowland

(TSTRITA). In other words, villages with ample land have a five times higher net

immigration rate compared to villages with constraints for future land expansion. The size

of this coefficient shows that rural-to-rural migration is largely driven by the possibility of

further expanding the agricultural frontier. The negative coefficient for the distance to

grazing areas of cattle (TRANHUMD) supports this finding.

Migrants prefer villages that have lower travel time to institutions for transferring

money or sending letters (MONEYTRA). As migrant families maintain their ties with

their extended family in the home region, sending and receiving money and letters are

important. Each additional hour of travel time to the post office or savings bank

significantly reduces the net immigration rate by 0.12 percent. This translates into an

elasticity of 7. Controlling for all other factors, migrants seem to prefer villages with

higher population density, either at the village (FOKDENS) or at the level of district

(Firaisana) (POPDENS). While the former is significant at the 20 percent level, the latter

is not. Social capital (SOCCAP) has a positive, but insignificant effect on the rate of

immigration from the village. With an increasing percentage of households owning less

than one hectare of upland, the positive effect of social capital on immigration diminishes

(LANDSCAP). In other words, as more virtually landless households live in the village,

pressure on idle land presumably increases and a countervailing power against further

immigration may begin to gain strength, reducing the positive effect of social capital on

immigration at the margin. Just as trade unions seek to protect the wages of their

members, the landless class in rural villages may be able to form coalitions and exert

political voice that perhaps undermines and works against the interest of the land-rich

households for importing cheap labor from other regions.


An improved access by the community to financial markets (OUTREACH)

increases the wage rate, although the estimated coefficient is insignificant. Villages with

higher population density (FOKDENS) and higher diversification of income sources

(SHARINCO) have significantly higher wage rates. These results appear to be caused by

the wage-enhancing effects of increased diversification of the rural economy that comes

along with higher population density and, linked with this, lower transaction costs in trade.

As can be seen from the regression coefficient (RNETIMIG) in Table 8, rural migration

depresses wages in villages that receive migrants and increases wages in villages that sent

migrants. This result confirms the previous argument that the landless class in migrant-

receiving villages may negatively be affected by migrants and therefore built up political

and social coalitions that dampen further immigration. As expected, since most wage

laborers perform labor services for agricultural enterprises, improved agricultural

productivity-measured by the yields of the major crop (YLDRICI)-increases the wage

rate through the increase of the marginal product from labor used in rice production. The

elasticity of the wage rate with respect to yields is measured at 0.33.

As expected, the consumer price of rice has a negative effect on the wage rate.

Since wages are paid in rice, an increase in the price of rice will result in paying less of

physical quantity of rice per labor day, if all other factors are held constant. Finally, with

an increased self-sufficiency index of rice (a traditional indicator of prosperity of a village),

the wage rate increases (see coefficient for variable AUTORIZ). This positive effect

becomes smaller at the margin with an increasing price of rice (PRICAUTR).


Holding everything else constant, the agricultural extensification pathway gains in

importance for villages with better access to financial institutions (OUTREACH), with

higher wage rates (WAGERATE), and with higher net immigration (RNETIMIG). Less

forestland and bushland is taken under cultivation if lowland rice yields are higher

(YLDRICI) or if soil fertility of upland is higher (SOILFERT). While the direct effect of

access to financial institutions (OUTREACH) on upland expansion is positive but

insignificant (an absolute 1 percent more of credit program participants increases upland

area by 0.247 percent), the indirect effects of improved credit access through other

endogenous variables are negative with the result that credit access actually reduces

upland expansion. The indirect effects can be calculated from the regression coefficients

for credit access (OUTREACH) on other endogenous variables, multiplied by the

regression coefficient of the endogenous variable for upland in Table 9. As this has

important policy implications, we discuss these results. From Table 6 we see that credit

access improves rice yields. In turn, one additional kilogram of rice reduces upland

expansion by 0.026 percent. Assuming constant marginal returns, the indirect effect is a

reduction of 0.52 percent of upland area with an absolute 1 percent increase in households

participating in member-based credit programs. The indirect effects of credit access on

upland expansion, through higher wage rates and higher soil fertility, are 0.017 and

-0.106, respectively. The combined effects, both direct and indirect, of a 1 percent

increase in the number of households participating in a credit program in the village are

then a reduction of 0.36 percent in upland area. Thus, improved access to credit markets

promotes agricultural intensification and preserves soils and forests.

The effect of population growth (RNATPOP) on upland area is very small. A 1

percent increase in the mean population growth rate increases upland by only 0.0092

percent This shows that other factors play a much more important role in explaining

expansion. Research in other countries also shows that the effect of population growth on

the expansion of cultivated area and resource degradation is mediated through and often

overshadowed by pressures arising from changing market, policy, or institutional

conditions (Cropper and Griffiths 1994; Foster, Rosenzweig, and Behrman 1997; Jodha


More land is put under the spate (angady), if the villages still possess the possibility

to extend its agricultural frontier (TSTPAT). The increase in upland is 16.7 percent higher

in such villages. With improved access to markets for non-rice crops (A34BAAC2) that

are usually cultivated on upland, the villagers seek to expand their upland area: with an

increase by one in the qualitative indicator for access to non-rice markets, the upland area

increases by an absolute 19 percent. The positive effect of access to non-rice crop

markets significantly diminishes at the margin with a higher consumer price of manioc

(MARKMAN). The results further suggest that the expansion of arable land can be

slowed down by rural electrification (DELEC). Electricity provides an alternative energy

source for the few that currently can afford it More important, access to electricity

enables a range of food processing, trading and off-farm enterprises that increase the

demand for labor and offer alternative employment for the poor.


The descriptive analysis showed that soil fertility has declined over the past 10

years. What caused this degradation? As shown in Table 10, holding all other

determinants constant, an increase in upland (CHTAN) significantly reduces soil fertility

(SOILFERT). Two effects explain this result First, as more land is put into cultivation,

more and more infertile soils are taken. Second, as the demand for land increases, the

time for fallow is reduced, giving upland less and less time to regenerate.

As expected, improved lowland rice yields, that is, agricultural intensification, have

a positive effect on soil fertility of upland as higher productivity of lowlands reduces the

incentives to mine upland soils. An increase in the index for soil fertility of upland by 0.2

percent is associated with a 1 percent higher yield of lowland rice. The results show that

soil fertility can also be enhanced by improving the community's access to financial

markets (OUTREACH).10 Given that rural households in Madagascar face credit

to As Pender and Kerr (1996) point out, improved access to credit (or higher initial liquid assets of the
household or village) is not expected to have an effect on conservation investment and soil fertility, if capital
markets are perfect, that is, every household can borrow and save as much as it wants at the prevailing market
rate. However, empirical evidence leads to the rejection of the assumption of a perfect credit market in rural

constraints, improved access to credit reduces the opportunity costs of capital, leading to

lower discount factors or time preference rates when valuing future income streams

against current ones. A major hypothesized determinant of the decline in soil fertility is

the high time preference rate that poor households use for discounting the value of future

use of soils.

A third potential option to preserve soil fertility is to improve access to input

markets (ACCINPUT). Yet, the coefficient on this variable is highly insignificant As was

the case for yields, the interaction term (OUTRINPU) between input and credit access is

negative and significant, indicating that the two types of access can act as substitutes for

each other. Population growth (RNATPOP) and its squared term (PSQRNAT) lead to

improved soil fertility, perhaps indicating that as land gets scarcer through population

growth, institutional arrangements and response mechanisms can be found to improve soil

fertility. However, these results are highly insignificant. More conflicts over grassland

and forest resources (CONFLICT) in the village during the previous 10 years are used as

an indicator of insecurity of land tenure. As tenure insecurity increases, investment in the

maintenance of soils is reduced as the investor cannot be sure to be able to reap the

benefits of his or her investment The sign of the regression coefficient confirms this

hypothesis, although the result is not statistically significant. Lower access to school

education, reflected by the travel time from the village to primary and secondary schools

combined (EDUC), appears to have a negative influence on soil fertility.

Social capital of the village significantly contributes to preserve and maintain soils

(SOCCAP). The results confirm the notion that common property management requires

collective action, and that collective action, in turn, necessitates social capital. It is further

interesting to note that the rice-self sufficiency index in the village (AUTORIZ) has a

significant and positive effect on soil fertility, confirming our hypothesis that a high

dependency on highly seasonal markets for buying rice increases consumption and

production risks that negatively impacts on the overall investment capacity of households.

In other words, if one has to rely on uncertain provisions of rice by the market, one wants

to reduce this dependency by growing more food even if this is at the expense of soil

fertility and future food production. The results suggest that improved access to financial

and agricultural input markets can help to preserve soil fertility, and-through the negative

effect of soil fertility on change in upland-to slow down the expansion of agriculture into



The agendas of environmental sustainability, economic growth, and poverty

alleviation are linked: pursuing one without regard of the other two is a pathway to failure

in the long run (Vosti and Reardon 1997). Addressing this critical triangle means putting

rural households and communities first. The critical triangle-and its specific problems

and ramifications in Madagascar-point to one overarching issue that needs to be better

understood: what makes households and individuals choose the environmentally more

sustainable intensification pathway over the extensification pathway, and how can

macroeconomic and sectoral policies and institutional and technological innovations in the

financial, agricultural, health, and education sector contribute to make the intensification

pathway more attractive in the short and long runs.

The analysis in this paper points to a number of implications for policy and further

research. Access to member-based financial institutions, such as credit groups, village

banks, or savings and credit cooperative societies, seems to play an important role for

enabling an agricultural intensification pathway in Madagascar. We find that access to

financial institutions has significant positive effects on lowland rice yields and on soil

fertility of upland. However, participation in member-based financial institutions also

leads to the expansion of upland, as capital becomes cheaper and more available to

farmers. The combined effect of access to financial institutions on upland is the sum of the

direct effect on upland change and the indirect effects of credit access through higher

yields, higher soil fertility, and higher wages. The sum of all effects shows that increasing

the percentage of households participating in a microfinance institution by one percent

reduces upland area by 0.36 percent. We conclude therefore that promoting microfinance

institutions for rural households can have beneficial effects on agricultural productivity,

poverty, and natural resources.

Our research further suggests a positive effect of improved access to markets for

rice and agricultural inputs on rice yields, on soil fertility of upland, and on the reduction

of newly cultivated hillsides. Yet, improved access to output markets for non-rice crops

that are grown on upland seems to lead to an expansion of cultivated upland in the short

run. As such market access increases, farmgate prices increase while farmers' transaction

costs for selling and buying crops decrease, ultimately pushing up the value of land.

Investments in soil conservation will consequently become more economical. Neglecting

the build-up of rural infrastructure and markets would only condemn rural villages and

households to continue with the low-input, low-output, but land-mining agricultural


We find that a major incentive for migration is to seek out villages with further

possibility for expanding the agricultural frontier. In so far as migrants search for better

living conditions, poverty in the villages and regions sending migrants (namely the Faritany

of Fianarantsoa) is a driving force for natural resource degradation elsewhere on the

island. Thus, alleviation of poverty through improving access to public services, such as

schools and health services, enhancement of domestic trade, and generation of off-farm

income opportunities has beneficial effects on conservation of soils and natural resources.

Our analysis weakly suggests that as land becomes scarcer, conflicts about common

property may reduce investments in soil conservation because of tenure insecurity. On the

other hand, social capital is found to significantly enhance soil fertility, presumably by

enabling villagers to agree on more sustainable property rights regimes.

While the potential of generating employment and income opportunities in

Madagascar's rural nonfarm sector should not be underrated, we note that most rural

households will have to continue to depend either directly or indirectly on agriculture and

related animal production. Agricultural intensification in the major food crops, i.e., rice,

cassava, potatoes, and maize, is therefore called for: Madagascar is a country where the

Green Revolution still needs to take place. The results strongly suggest a greater role of

public policy in improving agricultural productivity on irrigated lowlands and hillside

uplands through increases in investments for agricultural research and extension.

Because of lack of data, the analysis could not properly include a number of factors

that deserve further research. These include (1) the interdependence between food

security, health, and nutritional status and related pathways of rural development, (2) the

linkages among livestock, cropping, and off-farm income generation within the household

and village economy, (3) the role of (changing) property rights and related determinants

and institutional processes of change, and (4) the determinants of formation and

maintenance of social capital and of local-level institutions that enhance sustainable use of

forest and arable land.



Table 1-Indicators of economic growth, poverty and resource degradation: Rice yields, wages for rural laborers, and
fertility of upland soil, by region and year

Majunga: Plaines Majunga: Plateaux Fianar: HT Fianar: Cote/Falaise Vakinankaratra All regions

1987 1997 1987 1997 1987 1997 1987 1997 1987 1997 1987 1997

Yield irrigated rice 2,664 2,246 1,927 1,470 1,705 1,611 1,475 1,213 1,854 1,852 1,765 1,540

Number of kapaoka of rich paid per
day worked 9.5 8.5 14.7 8.0 4.2 3.7 6.1 5.0 5.3 6.7 7.6 5.7

Index for change in fertility of upland
soil 1.3 1.2 0.3 0.9 0.9 0.8

Source: IFPRI/FOFIFA Community Survey 1997.
Note: The index for soil fertility has the value 2 if no change over the past 10 years. The other values recorded are 3 = improved, 1 = reduced,
0 = severely degraded.

Table 2-Indices of change in area of grassland, forest, arable upland and irrigated
lowland during past 10 years

Majunga Majunga Fianar Fianar All
Plaines Plateaux HT Cote/Falaise Vakinankaratra regions

Index for increase in kijana area
1987 = 100 79.2 86.7 67.5 86.5 84.3 79.9

Index for increase in primary
forest area 1987 = 100 83.5 88.5 43.3 73.0 69.8 67.4

Index for increase in secondary
forest area 1987 = 100 80.4 79.8 63.2 76.3 67.5 72.0

Index for increase in tanety
area 1987 = 100 131.6 129.6 134.2 107.3 119.8 124.0

Index for increase in tanimbary
area 1987=100 112.6 104.6 105.7 104.0 105.7 105.3

Source: IFPRI/FOFIFA Community Survey 1997.
Note: The figures are indices for which the value 100 was given for the area held in 1987 by the
community. For example, an index of 120 implies a 20 percent average increase in the area of the
sample villages belonging to a particular region.

Table 3-Participation in microfinance programs and travel time to financial
institutions, by region and year

All regions
1987 1997

Rainy season travel time in hours to nearest bank 4.3 4.8

Rainy season travel time in hours to nearest post office or postal savings bank 3.6 4.0

Percent of households participating in credit 0.0 11.4

Source: IFPRI/FOFIFA Community Survey 1997.

Table 4-Descriptive statistics of variables used in regressions (n = 376 from 188

Name of variable Definition Mean Median deviation













Percent of households participating in credit program
Annual natural population growth rate in percent
Net annual immigration rate in village in percent
Number of kapoaka of rice per labor day
Index for increase in tanety area 1987 = 100
Yield irrigated rice 1,
Soil fertility change 3=improved, 0=very degraded
Mean of indices of access to seven agriculture input
markets, 100 = 1987
Change in access to fertilizer dealers during past
5 years, 1=no change, 0=worse, 2=better
5 year change in access to other crop buyers,
1-=no change, 0 -worse,
Interacted term: price of rice index of access to
rice markets
Interacted term: price of manioc index of access
to nonrice agriculture output markets
Rainy season travel time in hours to nearest bank
Interacted term: BANK times number of covariate
risks in last 10 years
Rainy season travel time in hours to nearest post
office or caisse 6pargne
Electricity in village l=yes, 0=no
Rainy season travel time to education institutions
Rainy season travel time to hospital or health post
Percent of rice land irrigated
Share of primary income source in village income
Consumer price of 1 kapaoka of white rice
Interacted term: price of rice (PRIZ) rice self-
sufficiency index (autoriz)
Number of people per square kilometers in village
Number of people per square kilometers infiraisana
Number of households in village owning less than
1 hectare of upland (tanety)
Interacted term: social capital (SOCCAP) with
Social capital: number of present informal
groups* years of existence
Average upland holding in hectare per household


5.68 0.00
3.56 3.45
0.01 0.00
6.67 5.00
11.98 100.00
52.40 1,500.00
1.40 2.00

0.99 1.00

0.92 1.00

1.04 1.00





89.43 59.52 78.44

96.90 50.91
4.54 3.00


33.00 0.00 75.13







29.59 20.00 29.93

211.52 0.00 14,235.01

68.24 0.00 360.96
1.51 1.15 1.20


Mean Median deviation

Name of variable Definition




Interacted term: price of rice (PRIZ) with
LANDLT1 1,945.07
Index of rice self-sufficiency of village, 100 = 1987 1.07
Possibility of extension of tanety l=y, 0=n, 1987=1 0.79
Possibility of extension of rice land l=y, 0=n, 1987=1 0.75
Interacted term: tstpat*tstriz 0.67
Number of conflicts about grassland/kijana 0.15
Distance in kilometer from village to cattle
grazing area 28.35
Consumer price of rice divided by price of manioc 0.80
Interacted term: WAGERATE* TSTPAT 5.00

Source: IFPRI/FOFIFA 1997 Community Survey.

Note: All descriptive in this and the following tables are weighted averages, using the sampling weights
computed for the stratified random sampling frame.

Table 5-Determinants of participation in member-based financial institutions

Variables Coefficient Standard error t-value

YLDRICIR 0.12817E-01 0.71157E-02 1.801**
BANK -0.21288 0.72728 -0.293
DELEC 15.190 22.492 0.675
SOCCAP 0.18258E-01 0.11007E-01 1.659**
FOKDENS -0.10264E-01 0.13716E-01 -0.748
LANDLT1 0.20022 0.17806 1.124
LANDSCAP -0.16480E-03 0.28227E-03 -0.584
Constant -19.687 13.338 -1.476*

Notes: Dependent variable is OUTREACH with mean of 5.68 and standard deviation of 29.28. Number
of observations: 376. Degrees of freedom = 179.
R-squared 0.59 and adjusted R-squared = 0.15.
*, **, and *** indicate significance at the 15 percent, 10 percent, and 5 percent level.

Table 4 (continued)





Table 6-Determinants of yields of irrigated rice

Variables Coefficient Standard error t-value

CHTAN -0.47583 3.5631 -0.134
SOILFERT -36.605 114.32 -0.320
WAGERATE -30.095 32.101 -0.938
OUTREACH 20.107 7.6738 2.620***
OUTRFERT -12.407 6.6113 -1.877**
PRIZ -3.6858 2.7553 -1.338
MARKRIZ 0.39802 0.94583 0.421
MARKMAN 0.45386E-01 0.46810 0.097
TSTRIZI -95.094 91.067 -1.044
A34ACC1 136.62 105.60 1.294
PEREAU 2.8289 2.6442 1.070
PRICEREL 83.247 21.707 3.835***
PRICAUTR -1.1088 1.7382 -0.638
AUTORIZ -269.62 256.21 -1.052
Constant 2,177.7 588.08 3.703***

Notes: Dependent variable is YLDRICIR with mean of 1,652.40 and standard deviation of 760.22.
Number of observations: 376. Degrees of freedom = 172.
R-squared 0.86 and adjusted R-squared = 0.69.
*, **, and *** indicate significance at the 15 percent, 10 percent, and 5 percent level.

Table 7-Determinants of migration

Variables Coefficient Standard error t-value

WAGERATE 0.17356E-01 0.70694E-01 0.246
WATPAT -0.87747E-01 0.52580E-01 -1.669**
TSTRITA 0.66022 0.39019 1.692**
HOSPITAL 0.12580 0.85311E-01 1.475*
EDUC 0.38938E-01 0.83788E-01 0.465
MONEYTRA -0.12592 0.46989E-01 -2.680***
POPDENS 0.13163E-03 0.20789E-02 0.063
TRANHUMD -0.58849E-03 0.29993E-02 -0.196
SOCCAP 0.36156E-03 0.45638E-03 0.792
LANDSCAP -0.12310E-04 0.11700E-04 -1.052
FOKDENS 0.84745E-03 0.60349E-03 1.404
Constant -0.25177 0.59452 -0.423

Notes: Dependent variable is RNETIMIG with mean of -0.07 and standard deviation of 1.21.
Number of observations: 376. Degrees of freedom = 175.
R-squared 0.61 and adjusted R-squared = 0.16.
*, **, and *** indicate significance at the 15 percent, 10 percent, and 5 percent level.

Table 8-Determinants of rural wage rate

Variables Coefficient Standard error t-value

RNETIMIG -0.45322 0.38361 -1.181
YLDRICIR 0.13507E-02 0.12537E-02 1.077
OUTREACH 0.87814E-02 0.26850E-01 0.327
FOKDENS 0.27656E-02 0.15360E-02 1.801**
PRIZ -0.21659E-02 0.17181E-01 -0.126
AUTORIZ 2.4720 1.6974 1.456*
PRICAUTR -0.21634E-01 0.10671E-01 -2.027***
SHARINCO -0.59400E-01 0.30764E-01 -1.931**
Constant 6.3875 3.5893 1.780**

Notes: Dependent variable is WAGERATE with mean of 6.67 and standard deviation of 4.36.
Number of observations: 376. Degrees of freedom = 178.
R-squared 0.7961 and adjusted R-squared = 0.56.
*, **, and *** indicate significance at the 15 percent, 10 percent, and 5 percent level.

Table 9-Determinants of expansion of arable upland

Variables Coefficient Standard error t-value

YLDRICIR -0.26405E-01 0.14709E-01 -1.795**
SOILFERT -14.013 6.3488 -2.207***
WAGERATE 2.0264 1.4973 1.353
RNETIMIG 1.4897 3.2619 0.457
OUTREACH 0.24726 0.24391 1.014
TSTPAT 16.705 5.4775 3.050***
DELEC -28.826 23.325 -1.236
AUTORIZ -21.228 14.796 -1.435
PRICAUTR 0.17182 0.87940E-01 1.954**
PRIZ -0.25321 0.14753 -1.716**
MARKRIZ 0.37519E-01 0.48049E-01 0.781
MARKMAN -0.12102 0.44332E-01 -2.730***
A34BACC2 19.863 9.5550 2.079***
RNATPOP 0.55790 0.70146 0.795
PNATPAT -0.41441 1.1443 -0.362
PRICEREL 2.0634 1.3793 1.496*
Constant 159.26 36.857 4.321***

Notes: Dependent variable is CHTAN with mean of 111.98 and standard deviation of 25.55.
Number of observations: 376. Degrees of freedom = 170.
R-squared 0.70 and adjusted R-squared = 0.34.
*, **, and *** indicate significance at the 15 percent, 10 percent, and 5 percent level.

Table 10-Determinants of soil fertility on upland

Variables Coefficient Standard error t-value

CHTAN -0.82909E-02 0.39462E-02 -2.101
YLDRICIR 0.17701E-03 0.24547E-03 0.721
OUTREACH 0.76829E-02 0.13813E-01 0.556
OUTRINPU -0.19943E-01 0.12347E-01 -1.615
EIDWI241E-01 0.37853E-01 -1.090
ACCINPUT 0.59672E-02 0.21238 0.028
CONFLICT -0.56723E-01 0.54192E-01 -1.047
AUTORIZ 0.35286 0.18792 1.878
RNATPOP 0.23778E-02 0.18927E-01 0.126
PSQRNAT 0.75183E-03 0.15944E-02 0.472
LANDLT1 -0.59350E-02 0.42577E-02 -1.394
FOODPOOR 0.16124E-04 0.28418E-04 0.567
LANDSCAP -0.68352E-05 0.62478E-05 -1.094
SOCCAP 0.53611E-03 0.29026E-03 1.847
Constant 1.9094 0.73036 2.614

Notes: Dependent variable is SOILFERT with mean of 1.40 and standard deviation of 0.80.
Number of observations: 376. Degrees of freedom = 172.
R-squared 0.77 and adjusted R-squared = 0.50.
*, **, and *** indicate significance at the 15 percent, 10 percent, and 5 percent level.


Baland, J.-M., and J.-P. Platteau. 1998. Dividing the commons-An assessment of the

new institutional economics of property rights. Photocopy, University of Namur,


Barrett, C. 1997. Food marketing liberalization and trader entry: Evidence from

Madagascar. World Development 25 (5): 763-777.

Boserup, E. 1965. The conditions of agricultural growth: The economics of agrarian

change under population pressure. Chicago: Aldine Publishing Co.

von Braun, J. 1997. The links between agricultural growth, environmental degradation,

and nutrition and health: Implications for policy and research. In Sustainability,

growth, and poverty alleviation: A policy and agroecological perspective, ed. S. A.

Vosti and T. Reardon. Baltimore, Md., U.S.A., and London: Johns Hopkins

University Press.

Cropper, M., and C. Griffiths. 1994. The interaction of population growth and

environmental quality. American Economic Review 84 (2): 250-254.

Foster, A., M. Rosenzweig, and J. Behrman. 1997. Population growth, income growth

and deforestation: Management of village common land in India. Brown

University, University of Pennsylvania.

Fraslin, J.-H. 1997. Quel credit pour les agriculteurs? Economic de Madagascar 2:


Gezon, L. 1997. Institutional structure and the effectiveness of integrated conservation

and development projects: Case study from Madagascar. Human Organization 56

(4): 462-470.

Greene, W. 1993. Econometric analysis. Second Edition. Englewood, N.J., U.S.A.:

Prentice Hall, Inc.

Greene, W. 1995. LIMDEP Version 7.0, Chapter 17.4, pp. 309-314.

Hazell, P., and S. Haggblade. 1993. Farm-nonfarm growth linkages and the welfare of

the poor. In Including the poor, ed. M. Lipton and J. van der Gaag. Washington,

D.C.: World Bank.

Hazell, P., and A. Roell. 1983. Rural growth linkages: Household expenditure patterns

in Malaysia and Nigeria. Research Report 41. Washington, D.C.: Intemational

Food Policy Research Institute.

Hossain, M. 1988. Credit for alleviation of rural poverty: The Grameen Bank.

Research Report 65. Washington, D.C.: International Food Policy Research


Islam, N. 1997. The nonfarm sector and rural development-Review of issues and

evidence. Food, Agriculture and the Environment Discussion Paper 22,

Washington, D.C.: International Food Policy Research Institute.

Jarosz, L. 1993. Defining and explaining tropical deforestation: Shifting cultivation and

population growth in colonial Madagascar (1896-1940). Economic Geography 69

(4): 366-379.

Jodha, N. 1985. Population growth and the decline of common property resource in

India. Population and Development Review 11 (2): 247-264.

Keck, A., N. P. Sharma, and G. Feder. 1994. Population growth, shifting cultivation,

and unsustainable agricultural development: A case study from Madagascar.

World Bank Discussion Paper, Africa Technical Department Series, No. 234.

Washington, D.C.: World Bank.

Kikuchi, M., and Y. Hayami. 1980. Inducements to institutional innovations in an

agrarian community. Economic Development and Cultural Change 29 (1): 21-36.

Lapenu, C., M. Zeller, B. Minten, E. Ralison, D. Randrianaivo, and C. Randrianarisoa.

1998. Extensification Agricole et son impact sur les pratiques agricoles, la

fertilite des sols et les resources naturelles. Cahier de la Recherche sur les

Politiques Alimentaires No 11, FOFIFA-IFPRI, Antananarivo.

Larson, B. A. 1994. Changing the economics of environmental degradation in

Madagascar: Lessons from the National Environmental Action Plan Process.

World Development 22 (5): 671-689.

Minten, B., C. Randrianarisoa, and M. Zeller. 1998. Niveau, Evolution etFacteurs

Diterminants des Rendements du Riz a Madagascar. Cahiers de la Recherche

IFPRI/FOFIFA sur les politiques alimentaires (IFPRI/FOFIFA working paper

series) No. 8. Antananarivo, Madagascar: Centre National de la Recherche

Appliqu6e au Ddveloppement Rural (FOFIFA).

Pender, J., and J. Kerr. 1996. Determinants of farmers' indigenous soil and water

conservation investments in India's semi-arid tropics. Environment and

Production Technology Division Discussion Paper 17. International Food Policy

Research Institute, Washington, D.C.

Pingali, P., Y. Bigot, and H. P. Binswanger. 1987. Agricultural mechanization and the

evolution of farming systems in Sub-Saharan Africa. Baltimore, Md., U.S.A.:

Johns Hopkins University Press.

Pitt, M., M. Rosenzweig, and D. Gibbons. 1995. The determinants and consequences of

the placement of government programs in Indonesia. In Public spending and the

poor, ed. D. van der Walle and K Nead. Washington, D.C.: World Bank.

Rakotomanga, G. 1976. Le Fokonolisme et le droit foncier. Unpublished dissertation.

University of Madagascar.

Rakotonirainy, G. 1984. La dynamique des structures fonciers sur les Hautes Terres

Malgaches. Unpublished mimeograph.

Ravalosoa, R. J. 1998. Les disparites &conomiques a l'heure de la decentralisation: une

image regional de la consommation des manages en 1993/94, Projet Madio,

INSTAT, Document 9637E.

Ruthenberg, H. 1980. Farming systems in the tropics. 3r edition. Oxford: Clarendon


Ruttan, V., and Y. Hayami. 1984. Toward a theory of induced institutional innovation.

Journal ofDevelopment Studies 20 (4): 203-223.

Scherr, S., and P. Hazell. 1994. Sustainable agricultural development strategies in

fragile lands. Environment and Production Technology Division Discussion Paper

1. International Food Policy Research Institute, Washington, D.C.

SECALINE. 1996. Evaluation de la situation alimentaire et nutritionelle a Madagascar.

Sharma, M., and M. Zeller. 1999. Placement and outreach of group-based credit

organizations: The cases ofASA, BRAC, and PROSHIKA in Bangladesh. Food

Consumption and Nutrition Division Discussion Paper 59. International Food

Policy Research Institute, Washington, D.C.

Vosti, S. A., and T. Reardon, eds. 1997. Sustainability, growth, and poverty alleviation:

A policy and agroecological perspective. Baltimore, Md., U.S.A., and London:

Johns Hopkins University Press.

Zeller, M. 1994. Determinants of credit rationing: A study of informal lenders and formal

credit groups in Madagascar. World Development 22 (12): 1895-1909.

Zeller, M. 1998. Determinants of repayment performance in credit groups: The role of

program design, intragroup risk pooling, and social cohesion. Economic

Development and Cultural Change 46 (3): 599-620.


01 Agricultural Technology and Food Policy to Combat Iron Deficiency in Developing Countries,
Howarth E. Bouis, August 1994

02 Determinants of Credit Rationing: A Study of Informal Lenders and Formal Credit Groups in
Madagascar, Manfred Zeller, October 1994

03 The Extended Family and Intrahousehold Allocation: Inheritance and Investments in Children in the
Rural Philippines, Agnes R. Quisumbing, March 1995

04 Market Development and Food Demand in Rural China, Jikun Huang and Scott Rozelle, June 1995

05 Gender Differences in Agricultural Productivity: A Survey of Empirical Evidence, Agnes R.
Quisumbing, July 1995

06 Gender Differentials in Farm Productivity: Implications for Household Efficiency and Agricultural
Policy, Harold Alderman, John Hoddinott, Lawrence Haddad, and Christopher Udry, August 1995

07 A Food Demand System Based on Demand for Characteristics: If There Is "Curvature" in the Slutsky
Matrix, What Do the Curves Look Like and Why?, Howarth E. Bouis, December 1995

08 Measuring Food Insecurity: The Frequency and Severity of "Coping Strategies," Daniel G. Maxwell,
December 1995

09 Gender and Poverty: New Evidence from 10 Developing Countries, Agnes R. Quisumbing, Lawrence
Haddad, and Christine Pefia, December 1995

10 Women's Economic Advancement Through Agricultural Change: A Review of Donor Experience,
Christine Pefia, Patrick Webb, and Lawrence Haddad, February 1996

11 Rural Financial Policies for Food Security of the Poor: Methodologies for a Multicountry Research
Project, Manfred Zeller, Akhter Ahmed, Suresh Babu, Sumiter Broca, Aliou Diagne, and Manohar
Sharma, April 1996

12 Child Development: Vulnerability and Resilience, Patrice L. Engle, Sarah Castle, and Purnima Menon,
April 1996

13 Determinants ofRepayment Performance in Credit Groups: The Role of Program Design, Intra-Group
Risk Pooling, and Social Cohesion in Madagascar, Manfred Zeller, May 1996

14 Demand for High-Value Secondary Crops in Developing Countries: The Case ofPotatoes in
Bangladesh and Pakistan, Howarth E. Bouis and Gregory Scott, May 1996

15 Repayment Performance in Group-Based Credit Programs in Bangladesh: An Empirical Analysis,
Manohar Sharma and Manfred Zeller, July 1996

16 How Can Safety Nets Do More with Less? General Issues with Some Evidence from Southern Africa,
Lawrence Haddad and Manfred Zeller, July 1996

17 Remittances, Income Distribution, and Rural Asset Accumulation, Richard H. Adams, Jr., August 1996


18 Care and Nutrition: Concepts and Measurement, Patrice L. Engle, Pumima Menon, and Lawrence
Haddad, August 1996

19 Food Security and Nutrition Implications ofIntrahousehold Bias: A Review of Literature, Lawrence
Haddad, Christine Pefia, Chizuru Nishida, Agnes Quisumbing, and Alison Slack, September 1996

20 Macroeconomic Crises and Poverty Monitoring: A Case Study for India, Gaurav Datt and Martin
Ravallion, November 1996

21 Livestock Income, Male/Female Animals, and Inequality in Rural Pakistan, Richard H. Adams, Jr.,
November 1996

22 Alternative Approaches to Locating the Food Insecure: Qualitative and Quantitative Evidence from
South India, Kimberly Chung, Lawrence Haddad, Jayashree Ramakrishna, and Frank Riely, January

23 Better Rich, or Better There? Grandparent Wealth, Coresidence, and Intrahousehold Allocation,
Agnes R. Quisumbing, January 1997

24 Child Care Practices Associated with Positive and Negative Nutritional Outcomes for Children in
Bangladesh: A Descriptive Analysis, Shubh K. Kumar Range, Ruchira Naved, and Saroj Bhattarai,
February 1997

25 Water, Health, and Income: A Review, John Hoddinott, February 1997

26 Why Have Some Indian States Performed Better Than Others at Reducing Rural Poverty?, Gaurav
Datt and Martin Ravallion, March 1997

27 "Bargaining" and Gender Relations: Within and Beyond the Household, Bina Agarwal, March 1997

28 Developing a Research and Action Agenda for Examining Urbanization and Caregiving: Examples
from Southern and Eastern Africa, Patrice L. Engle, Pumima Menon, James L. Garrett, and Alison
Slack, April 1997

29 Gender, Property Rights, and Natural Resources, Ruth Meinzen-Dick, Lynn R. Brown, Hilary Sims
Feldstein, and Agnes R. Quisumbing, May 1997

30 Plant Breeding: A Long-Term Strategy for the Control of Zinc Deficiency in Vulnerable Populations,
Marie T. Ruel and Howarth E. Bouis, July 1997

31 Is There an Intrahousehold 'Flypaper Effect'? Evidence from a School Feeding Program, Hanan
Jacoby, August 1997

32 The Determinants of Demand for Micronutrients: An Analysis ofRural Households in Bangladesh,
Howarth E. Bouis and Mary Jane G. Novenario-Reese, August 1997

33 Human Milk-An Invisible FoodResource, Anne Hatley and Arne Oshaug, August 1997


34 The Impact of Changes in Common Property Resource Management on IntrahouseholdAllocation,
Philip Maggs and John Hoddinott, September 1997

35 Market Access by Smallholder Farmers in Malawi: Implications for Technology Adoption,
Agricultural Productivity, and Crop Income, Manfred Zeller, Aliou Diagne, and Charles Mataya,
September 1997

36 The GAPVU Cash Transfer Program in Mozambique: An assessment, Gaurav Datt, Ellen
Payongayong, James L. Garrett, and Marie Ruel, October 1997

37 Why Do Migrants Remit? An Analysis for the Dominican Sierra, B6n6dicte de la Bribre, Alain de
Janvry, Sylvie Lambert, and Elisabeth Sadoulet, October 1997

38 Systematic Client Consultation in Development: The Case of Food Policy Research in Ghana, India,
Kenya, andMali, Suresh Chandra Babu, Lynn R. Brown, and Bonnie McClafferty, November 1997

39 Whose Education Matters in the Determination ofHousehold Income: Evidence from a Developing
Country, Dean Jolliffe, November 1997

40 Can Qualitative and Quantitative Methods Serve Complementary Purposes for Policy Research?
Evidence from Accra, Dan Maxwell, January 1998

41 The Political Economy of Urban Food Security in Sub-Saharan Africa, Dan Maxwell, February 1998

42 Farm Productivity andRural Poverty in India, Gaurav Datt and Martin Ravallion, March 1998

43 How Reliable Are Group Informant Ratings? A Test of Food Security Rating in Honduras, Gilles
Bergeron, Saul Sutkover Morris, and Juan Manuel Medina Banegas, April 1998

44 Can FAO's Measure of Chronic Undernourishment Be Strengthened?, Lisa C. Smith, with a Response
by Logan Naiken, May 1998

45 Does Urban Agriculture Help Prevent Malnutrition? Evidence from Kampala, Daniel Maxwell, Carol
Levin, and Joanne Csete, June 1998

46 Impact ofAccess to Credit on Income and Food Security in Malawi, Aliou Diagne, July 1998

47 Poverty in India and Indian States: An Update, Gaurav Datt, July 1998

48 Human Capital, Productivity, and Labor Allocation in Rural Pakistan, Marcel Fafchamps and Agnes R.
Quisumbing, July 1998

49 A Profile of Poverty in Egypt: 1997, Gaurav Datt, Dean Jolliffe, and Manohar Sharma, August 1998.

50 Computational Tools for Poverty Measurement and Analysis, Gaurav Datt, October 1998

51 Urban Challenges to Food and Nutrition Security: A Review of Food Security, Health, and Caregiving
in the Cities, Marie T. Ruel, James L. Garrett, Saul S. Morris, Daniel Maxwell, Arne Oshaug, Patrice
Engle, Pumima Menon, Alison Slack, and Lawrence Haddad, October 1998


52 Testing Nash Bargaining Household Models With Time-Series Data, John Hoddinott and Christopher
Adam, November 1998

53 Agricultural Wages and Food Prices in Egypt: A Governorate-Level Analysis for 1976-1993, Gaurav
Datt and Jennifer Olmsted, November 1998

54 Endogeneity of Schooling in the Wage Function: Evidence from the Rural Philippines, John Maluccio,
November 1998

55 Efficiency in Intrahousehold Resource Allocation, Marcel Fafchamps, December 1998

56 How Does the Human Rights Perspective Help to Shape the Food and Nutrition Policy Research
Agenda?, Lawrence Haddad and Ame Oshaug, February 1999

57 The Structure of Wages During the Economic Transition in Romania, Emmanuel Skoufias, February

58 Women's Land Rights in the Transition to Individualized Ownership: Implications for the Management
of Tree Resources in Western Ghana, Agnes Quisumbing, Ellen Payongayong, J. B. Aidoo, and Keijiro
Otsuka, February 1999

59 Placement and Outreach of Group-Based Credit Organizations: The Cases ofASA, BRAC, and
PROSHIKA in Bangladesh, Manohar Sharma and Manfred Zeller, March 1999

60 Explaining Child Malnutrition in Developing Countries: A Cross-Country Analysis, Lisa C. Smith and
Lawrence Haddad, April 1999

61 Does Geographic Targeting of Nutrition Interventions Make Sense in Cities? Evidence from Abidjan
andAccra, Saul S. Morris, Carol Levin, Margaret Armar-Klemesu, Daniel Maxwell, and Marie T. Ruel,
April 1999

62 Good Care Practices Can Mitigate the Negative Effects of Poverty and Low Maternal Schooling on
Children's Nutritional Status: Evidence from Accra, Marie T. Ruel, Carol E. Levin, Margaret Armar-
Klemesu, Daniel Maxwell, and Saul S. Morris, April 1999

63 Are Urban Poverty and Undernutrition Growing? Some Newly Assembled Evidence, Lawrence
Haddad, Marie T. Ruel, and James L. Garrett, April 1999

64 Some Urban Facts of Life: Implications for Research and Policy, Marie T. Ruel, Lawrence Haddad,
and James L. Garrett, April 1999

65 Are Determinants ofRural and Urban Food Security and Nutritional Status Different? Some Insights
from Mozambique, James L. Garrett and Marie T. Ruel, April 1999

66 Working Women in an Urban Setting: Traders, Vendors, and Food Security in Accra, Carol E. Levin,
Daniel G. Maxwell, Margaret Armar-Klemesu, Marie T. Ruel, Saul S. Morris, and Clement Ahiadeke,
April 1999


67 Determinants ofHousehold Access to and Participation in Formal and Informal Credit Markets in
Malawi, Aliou Diagne, April 1999

68 Early Childhood Nutrition and Academic Achievement: A Longitudinal Analysis, Paul Glewwe, Hanan
Jacoby, and Elizabeth King, May 1999

69 Supply Response of West African Agricultural Households: Implications of lntrahousehold Preference
Heterogeneity, Lisa C. Smith and Jean-Paul Chavas, July 1999

70 Child Health Care Demand in a Developing Country: Unconditional Estimates from the Philippines,
Kelly Hallman, August 1999

71 Social Capital and Income Generation in South Africa, 1993-98, John Maluccio, Lawrence Haddad,
and Julian May, September 1999

72 Validity ofRapid Estimates of Household Wealth and Income for Health Surveys in Rural Africa, Saul
S. Morris, Calogero Carletto, John Hoddinott, and Luc J. M. Christiaensen, October 1999

73 Social Roles, Human Capital, and the Intrahousehold Division of Labor: Evidence from Pakistan,
Marcel Fafchamps and Agnes R. Quisumbing, October 1999

74 Can Cash Transfer Programs Work in Resource-Poor Countries? The Experience in Mozambique, Jan
W. Low, James L. Garrett, and Vitbria Ginja, October 1999

75 Determinants of Poverty in Egypt, 1997, Gaurav Datt and Dean Jolliffe, October 1999

76 Raising Primary School Enrolment in Developing Countries: The Relative Importance of Supply and
Demand, Sudhanshu Handa, November 1999

77 The Political Economy of Food Subsidy Reform in Egypt, Tammi Gutner, November 1999.

78 Determinants of Poverty in Mozambique: 1996-97, Gaurav Datt, Kenneth Simler, Sanjukta Mukherjee,
and Gabriel Dava, January 2000

79 Adult Health in the Time ofDrought, John Hoddinott and Bill Kinsey, January 2000

80 Nontraditional Crops and Land Accumulation Among Guatemalan Smallholders: Is the Impact
Sustainable? Calogero Carletto, February 2000

81 The Constraints to Good Child Care Practices in Accra: Implications for Programs, Margaret Armar-
Klemesu, Marie T. Ruel, Daniel G. Maxwell, Carol E. Levin, and Saul S. Morris, February 2000

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs