LABOUR PRODUCTIVITY AND EFFICIENCY IN SMALLHOLDER MAIZE

PRODUCTION: A CASE OF MPINGU EXTENSION PLANNING AREA

BY

HENRY KANKWAMBA

A RESEARCH PROJECT REPORT SUBMITTED TO THE FACULTY OF

DEVELOPMENT STUDIES, DEPARTMENT OF AGRICULTURAL AND APPLIED
ECONOMICS IN PARTIAL FULFILMENT OF THE REQUIREMENTS OF THE
BACHELOR OF SCIENCE DEGREE IN AGRICULTURAL ECONOMICS

Bunda College of Agriculture

P.O. Box 219

Lilongwe

Malawi

i
Abstract
Malawi allocates most of its labour to maize production in the rainy season. Literature
admits that there seems to be a labour surplus in spite of persistent low maize output.
Nevertheless, this might be misleading because there is insufficient information on the local
smallholder labour market.

The main objective of this research is to assess productivity and efficiency of smallholder
labour and other factors spent on maize production and their effects on output and net profit
in Mpingu EPA. The research focuses on the relationship between labour inputs and output
while holding other factors constant. A production function is used in the analysis to show
these relationships. The study also seeks to evaluate other factors that affect productivity on a
small holder farm. This paper shows that a positive relationship exists between total output,
labour land, fertilizer, and seed.

The less than unity elasticity of production in labour, 0.199, seed, 0.205, fertilizer 0.487, and
land 0.192, respectively present compelling evidence that diminishing returns prevail in
smallholder maize production. This further indicates that production occurs in the rational
region. Furthermore, increasing returns to scale, 1.1258, indicates that the smallholders can
still increase. Finally, the research gives recommendations on agricultural policy.

ii
Acronyms
AEDO : Agricultural Extension Development Officer

EPA : Extension Planning Area

DADO : District Agricultural Development Officer

MoAFS : Ministry of Agriculture and Food Security

NSO : National Statistical Office

SDNP : Sustainable Development Network Programme

NGO : Non-Governmental Organizations

SPSS : Statistical Package for Social Scientists

RTS : Returns to Scale

iii
1 Contents
Abstract ................................................................................................................................................... ii
Acronyms ............................................................................................................................................... iii
Table of Figures .................................................................................................................................. vi
1 INTRODUCTION ............................................................................................................................... 7
1.1.1 Background ..................................................................................................................... 7
1.1.2 Problem statement ......................................................................................................... 7
1.1.3 Justification ..................................................................................................................... 8
1.1.4 Objectives........................................................................................................................ 9
1.1.5 Hypotheses.................................................................................................................... 10
2 LITERATURE REVIEW ..................................................................................................................... 10
2.1 The maize productivity situation in Malawi.......................................................................... 10
2.2 Labour situation in Malawi ................................................................................................... 10
2.3 Impacts of labour market liberalization on employment, maize production and productivity
in Malawi........................................................................................................................................... 12
3 The Impact of HIV and AIDS on Agriculture .............................................................................. 13
4 METHODOLOGY ............................................................................................................................ 14
4.1 The Study Area ...................................................................................................................... 14
4.2 Data Collection Methods ...................................................................................................... 15
4.2.1 Survey............................................................................................................................ 15
4.2.2 Key informant interviews .............................................................................................. 15
4.3 Sampling Design .................................................................................................................... 15
4.4 Data Analysis ......................................................................................................................... 16
5 THEORETICAL FRAMEWORK; JOHANNES AND MENDOZA-ESCALANTE, (2007) ........................... 16
5.1 Economics of agriculture production.................................................................................... 16
5.1.1 Returns to scale............................................................................................................. 18
5.2 The Basic Model .................................................................................................................... 19
5.2.1 The basic production model ......................................................................................... 19
5.2.2 Derivation of the Cobb – Douglas Functions ................................................................ 21
6 RESULTS AND DISCUSSION............................................................................................................ 22
Descriptive statistics ............................................................................................................................. 22

iv
 6.1 Socioeconomic Variables ...................................................................................................... 22
6.1.1 Age of the household head ........................................................................................... 22
6.1.2 Marital status of household head ................................................................................. 23
6.1.3 Education level of the household head ........................................................................ 23
6.1.4 Household size .............................................................................................................. 24
6.2 Factors of production............................................................................................................ 24
6.2.1 Labour ........................................................................................................................... 24
6.2.2 Land ............................................................................................................................... 27
6.2.3 Fertilizer ........................................................................................................................ 27
6.2.4 Seed............................................................................................................................... 29
6.3 Inferential Statistics: Cobb-Douglas Production function estimates .................................... 29
6.3.1 Productivity Analysis ..................................................................................................... 31
6.3.2 Economic optimization ................................................................................................. 32
7 CONCLUSION ................................................................................................................................. 34
8 RECOMMENDATIONS.................................................................................................................... 35
9 BIBLIOGRAPHY .............................................................................................................................. 35
10 APPENDIX 1 ............................................................................................................................... 38
Questionnaire ................................................................................................................................... 38
11 APPENDIX 2 ................................................................................................................................. 2
Economic Optimization Calculations .............................................................................................. 2

v
 List of Tables
Table 1 Marital status of household head ............................................................................................ 23
Table 2 Marital Status of the Household Head ..................................................................................... 24
Table 3 Household size ......................................................................................................................... 24
Table 4 Descriptive Statistics for potential labour source .................................................................... 25
Table 5 Hired Labour ............................................................................................................................. 26
Table 6 Labour Use................................................................................................................................ 26
Table 7 Reasons for not hiring extra labour.......................................................................................... 26
Table 8 Problems on the farm .............................................................................................................. 27
Table 9 Input Costs of maize as Used in the Gross margin Analysis ..................................................... 28
Table 10 Fertilizer and Manure Costs and Quantities .......................................................................... 28
Table 11 Quantity (Kg) and Cost of Seed (MK)...................................................................................... 29
Table 12 Model Summary (b) .......................................................................................................... 29
Table 13 ANOVA (b) ......................................................................................................................... 30
Table 14 Coefficients (a) .................................................................................................................. 30
Table 15 Elasticities of production ........................................................................................................ 31
Table 16 Gross Margin (GM) for Maize................................................................................................. 34

vi
1 INTRODUCTION

1.1.1 Background
Malawi like most Sub-Saharan African countries depends on agriculture as its main source of
national income. The agricultural sector is divided into smallholder and estate subsectors. The
smallholder subsector accounts for about 80 percent of domestic food production, 10 percent
of total exports and contributes 25 percent of real GDP and 65 percent of agricultural GDP
(SDNP, 1998).

Maize is a staple food of great socio-economic importance in the Sub-Saharan Africa

(Ogundari and Ajibefun, 2008), of which Malawi is arguably the largest consumer. It is the
most important cereal crop in Malawi since it comprises the main staple food nsima. The total
land area planted to maize in Malawi is above 2.5 million hectares with an estimated output
of about 1.4 metric tons per hectare. Most of this crop is produced by smallholder farmers.

A growing population in Malawi poses a challenge to find more productive and efficient
ways to produce enough food. However, constraints to smallholder agricultural production
abound. For instance population growth has led to increased pressure on land and natural
resources, (Nakhumwa, 2008). Increased prevalence of HIV/AIDS has led to scarcity of
labour and increased unemployment which has led to most people engaging into more off
farm activities as a survival strategy (Jayne et al., 2008). Lack of sufficient credit facilities,
imperfect and incomplete factor and product prices have led to downward spiraling
production and unpredictable prices overtime (Tchale, 2008). High fertilizer prices have
caused most able farmers unable to access inputs, (Rhorbauch, 2008). However, the coming
of the Input Subsidy Program has caused a surge in maize production over the past three
years (Tchale, 2008).

1.1.2 Problem statement

Having a population of 13.4 million people (NSO, 2008), smallholders constitute 71.3
percent of the total population and about 9 percent of the Malawi’s poor. They also constitute
more than 80 percent of the total labour force. 71 percent of the smallholder labour is used to
produce maize, the country’s staple food. However, the smallholders’ production is almost
exclusively rain fed using hand hoes and other simple implements. Hence most of the labour
is employed during the rainy season when production is at its peak. With a single rainy
season, in Malawi, this results in pronounced seasonality in factor markets especially labour
(Siegal and Alwang, 2003).

7|Page
From the information above it can be noted that labour is a key asset for smallholder
households in rural Malawi. The quality and quantity of labour available to the household in
terms of numbers, educational level, skills, and health constitute the human capital that
becomes the basis for constructing household livelihood strategies. In the context of
Malawi’s smallholder production where farm mechanization is virtually nonexistent and all
farm work is done manually, having access to necessary labour for agricultural production
directly affects the levels of household farm income. In addition to working on a household’s
own farm, labour may also be deployed in off farm economic activities, thus providing
additional income to the household (Takane, 2008).

The literature on labour use in rural Malawi has tended to focus on ganyu 1. Ganyu is a range
of short-term casual labour contracts that are widely practiced in the country. Based on
ethnographic research in central Malawi, ganyu is interpreted not as individualized contracts
but outcomes of socially embedded relationships. On the other hand, the importance of both
ganyu as the major source of income for poor households and its potential conundrum, as the
need to engage in ganyu to obtain an immediate income may conflict with own-farm
production. From different perspectives, ganyu represented the intensifying inequalities and
conflicts between haves and have-nots in rural Malawi (Takane, 2008).

1.1.3 Justification
Noteworthy, there is a particular need to improve the efficiency of the smallholder sub-sector,
which is by far the largest, with nearly three million farm families cultivating on over 70% of
Malawi’s arable land held under customary tenure. There is a wide gap between outputs
observed in on-farm and experimental trials and the actual outputs obtained by farmers. For
example, while potential outputs for hybrid maize range from 5 to 8 tons per hectare, the
average actual outputs range from 1.5 to 2.5 tons and rarely exceed this. Evidence from past
studies suggests that levels of efficiency among the majority of Malawian smallholders are
low to moderate. This gap between potential and actual average farm crop outputs suggests
abundant scope for improvements in productivity (Tchale, 2008). To maintain high
productivity in the face of declining land holding sizes, inefficient labour use, and rising
fertilizer and input prices optimal resource allocation is required.

1
Ganyu: The English translation of this word is uncertain but it refers to a survival mechanism that makes poor
farmers engage in off farm activities in exchange for cash, food or other materials.
8|Page
More importantly, Edriss et al. (2004), notes that there is no easy way of establishing causation
since poor agricultural performance is the root cause of the core problems in the agricultural
labour market. It should also be pointed out that previous efficiency studies in Malawi have
looked mainly at technical efficiency (see Chirwa, 2003; Edriss et al., 2004 as cited in
Tchale, (2008)). Technical efficiency (the relative position of the farmer on the frontier)
derives from an agronomic view and it is possible that the farmer could achieve this kind of
efficiency, though at a much higher cost. An economic view, on the other hand, considers the
use of inputs in optimal quantities while keeping their cost in proportion to the price the
farmer receives for the outputs. It is therefore useful to examine the factors that affect overall
economic efficiency rather than focus only on technical efficiency (Tchale, 2008).

The thrust of this research therefore is to assess productivity and efficiency in the smallholder
maize production. A production function provides the technical efficiency criterion.
Economic optimization has been looked at with much thoroughness to achieve allocative
efficiency. Coming up with output levels that a smallholder can maximize profit and least
combinations, will comprise the bulk of this study. Labour among others has been the factor
that will mostly be used in the calculations. This is so because economic optimization of this
resource has not been given considerable focus. However, other factors of production have
also been looked into.

1.1.4 Objectives

1.1.4.1 Underlying objective

To assess productivity and efficiency of labour and related factors of production spent on
maize production and their effects on output and profits on different farm conditions.

1.1.4.2 Specific objectives

i. To measure output responses to changes in labour and other factors of production
(production function).

ii. To determine levels of optimum labour usage on and resultant output and net benefit
on different farm conditions.

iii. To identify and optimize other factors that affect smallholder maize productivity.

9|Page
1.1.5 Hypotheses
The following hypotheses are postulated for the research:

a. That increase in labour use improves output.

b. That increase in variable input costs reduces the productivity of labour employed and
total physical product.

c. That labour is a constraint to smallholder maize production.

2 LITERATURE REVIEW

2.1 The maize productivity situation in Malawi

Conroy (2005), reported that smallholder agriculture in Malawi is dominated by maize
cultivation at low levels of productivity. A large proportion of smallholders cultivate less than
one hectare of land. Many studies conducted on Malawi indicate declining levels of
productivity that pose serious food security concerns, since maize is a staple food for most
parts of the country (Johannes, 2005). Low incomes of the majority of smallholders prevent
them from purchasing any inputs on cash, and they do not have sufficient collateral (in the
form of livestock assets) to make them credit worthy (Conroy, 2005).

Given the vicious cycle of low farm incomes and low productivity (or low output) farming
practices, it is obvious that one of the lasting solutions to alleviate rural poverty includes
addressing issues of agricultural productivity and sustainability. Given that most people are
employed in the agricultural sector, there is need to understand how the dynamics in the rural
labour market affect rural poverty, agricultural productivity, as well as, economic, ecological
and social sustainability of the agricultural sector. There is no easy way of establishing
causation since poor agricultural performance can be the root cause of the core problems in
the agricultural labour market. Simultaneously, problems in the agricultural labour market
have repercussions in terms of agricultural performance (Edriss et al. 2004).

2.2 Labour situation in Malawi

The agricultural labour market in Malawi supports the livelihoods of nearly 90% of the rural
population. Most of the population work as full-time farmers on their own land, others are
involved in casual agricultural labour and on tenancy arrangements with landlords.

10 | P a g e
Agriculture absorbs 88% of the national labour force leaving only 12% for non-agricultural
sectors (GoM, 2000). Edriss et al. (2004) notes that there is overwhelming evidence that due
to the chronic food insecurity, most farmers are involved in casual labor (called ganyu in the
local language), especially during the peak agricultural season.

Firstly, the fact that people work off their own farms and thus cannot spend the time on their
own fields means that most agricultural husbandry practices such as early cultivation,
weeding and bunding are either performed late or not performed at all. Secondly, since most
of the wage income is used for subsistence, as a bridging or coping mechanism, it is highly
unlikely that these farmers can also apply adequate quantities of costly inputs in their farming
activities in order to compensate for the decline in labour intensity. This results in a spiral of
declining agricultural productivity and real wages within the agricultural labour market.

With declining real wages, it is unlikely that wage earners can invest in more intensive
agricultural production. In order to bridge the gap between the times they run out of food and
the next harvest, smallholders frequently resort to off-farm casual employment locally known
as ganyu (Phiri, 2006). Moreover, Kumwenda et al. (1997), notes that many farm households
are forced to sell their labour in return for food, which in turn, compromises their own
agricultural efforts. Phiri (2006) asserts that the common practice of supplying off-farm
labour such as ganyu gives an impression of surplus labour for on farm activities. However
evidence suggests acute labour shortages. This leaves the smallholder with an allocative
inefficiency which rises when the inputs are not consistent with the higher input requirements
on the farm (Ashok et al, 1995). Given the vicious cycle of low farm incomes and low
productivity (or low output) farming practices, it is obvious that one of the lasting solutions to
alleviate rural poverty includes addressing issues of agricultural productivity and
sustainability (Edriss et al, 2004).

As further confirmation, based on data from the Second Integrated Household Survey for
2004, the potential objectives and determinants of ganyu supply were explored. A
confirmation of the neoclassical theoretical hypothesis whereby households use ganyu as a
consumption smoothing option and supply ganyu in response to a higher marginal product of
this particular type of labor as opposed to own farm labor. While the results therefore reject
the alternative hypothesis of a backward bending labor supply curve, they reveal interesting
11 | P a g e
insurance and cultural patterns of the allocation of labor across own farm and foreign farm
activities in rural Malawi. Village level shocks increase the gross amount of ganyu supply
across families that tend to supply labor and families that tend to demand ganyu labor.
Women headed households appear to make a particularly strong use of ganyu as an insurance
through social networks and for this purpose simultaneously supply and demand ganyu. And
ganyu is particularly widely spread in ethno-cultural contexts with matrilineal systems, i.e.
predominantly in the South and the Center of the country. However, enhanced education for
individual peasants as well as for the community as a whole, encouragement of cash crop
growing practices and crop diversification appear to open up some alternatives to traditional
ganyu labor (Weber, et al, 2008).

2.3 Impacts of labour market liberalization on employment, maize

production and productivity in Malawi
Edriss et al., (2004) quantifies the effects of labour market liberalization on employment,
maize production and productivity in Malawi. Applying a frontier production function and
Divisia Index, it was found that total factor productivity declined at 1.2% per annum from
1985 to 2000. Prior to market liberalization (1985 – 1995), productivity increased at an
average annual growth rate of 2.0%. However, there was a sharp decline in maize
productivity after 1995, implying that market liberalization had a significant impact on maize
production efficiency. Coupled with recent droughts experiences, the sharp maize
productivity decline could be attributed to sharp decline in the input use, that is, labor (-
6.7%), fertilizer (-1.5%) and land (-3.5%).

Maize productivity constantly declined from 1995 onwards, and decreased by 2.8% per
annum in the post liberalization era. The largest contributing factor to the maize productivity
decline was the decrease in farm labour input share, which was largely affected by shift in
policy reform imposed. Market liberalization including labour market reform had directly or
indirectly shifted allocation of labour from own farm to other farms or non-farm activities,
including critical times of farming activities that resulted in lower technical efficiency and
maize productivity in Malawi. Consequently, market liberalization reforms via their influence
on the price level and inflation have led to a persistent decline in real wages by 65.3%
between the pre- and post-market liberalization periods. This directly implies the weakening
of the purchasing power of resource poor farmers who become continuously more food

12 | P a g e
insecure, and this aggravates the poverty level especially in rural areas of Malawi (Edriss et
al, 2004).

3 The Impact of HIV and AIDS on Agriculture

Anticipating the future effects of AIDS on the agricultural sector requires an understanding of
how labour is likely to shift between urban and rural areas and between agricultural and non-
agricultural activities if/when agricultural labour shortages may arise. Many studies reveal
that the informal sectors of eastern and southern Africa have swelled massively in recent
decades largely because of inadequate income opportunities in rural areas and the pushing out
of labor from rural areas where population densities are high and where farm sizes have
declined to levels inadequate to sustain all the members of succeeding generations. These
points, taken together, indicate that incipient agricultural labour shortages caused by AIDS
are likely to induce labour migration out of the urban informal sector into agriculture, a
phenomenon which has already been detected in census data in some areas of the region
(Jayne et al, 2008).

Land-labour-capital ratios vary greatly within most countries of eastern and southern Africa.
Most nationally representative household surveys find that 25 percent of smallholder
households with the smallest farms typically controlled less than 0.1 hectares of land per
person. At least for this stratum of smallholder households, land is likely to remain a primary
constraint on income growth, and it is not clear that the loss of a household member would
change this much. There are strong reasons for anticipating that AIDS will progressively de-
capitalize highly-afflicted rural communities – meaning a loss of savings, cattle assets, draft
equipment, and other assets. The depletion of savings and capital assets for many households
may indeed pose the greatest limits on rural productivity and livelihoods for the hardest-hit
communities (Jayne et al, 2008). For instance, the case of Malawi proves the severity of the
predicament at hand. It should also be noted that the infection rates presented vary between
urban and rural areas and that affect the demographic structure and growth of the population
and labour force, but labour is further affected by culturally determined responsibilities, such
as women’s common responsibilities to care for the sick (Mwale, 2005).

Although at a macro level its impact seems insignificant, the Acquired Immunodeficiency
Syndrome, AIDS, has devastating implications on agriculture in Malawi. Whiteside, 2008,
paints a bleak picture. Nationally, HIV prevalence was estimated at 14.1 percent of adults. In
13 | P a g e
the central region, a survey by Care International found that a significant number of
households suffered from chronic illnesses and were unable to provide the labour needed for
even low-productivity subsistence agriculture. Between 22 percent and 64 percent of
households in study sites suffered from chronic sicknesses leading to loss of labour. In
households with labour loss, 45 percent delayed agricultural operations, 23 percent left land
fallow, and 26 percent changed the crop mix. Resources were being used for health care and
funerals, and this led to even lower levels of household income and nutrition. Female headed
households were worst affected. Malawian women do much of the agricultural work and
combine this with child bearing and rearing and household responsibilities. They have the
‘double burden of care’, as they are most likely to suffer from HIV/AIDS and are also
responsible for caring for others.

Whiteside, 2008 reports that AIDS is eroding the resilience of rural livelihoods, by
undermining coping strategies. Food coping strategies critically depend on labour
availability, skill, and experience. The disease and impacts cluster at the household level and
(to a lesser extend) with communities. Because AIDS changes age and gender distribution,
there are fewer mature adults (especially women), and relatively more teenagers and people
in early 20s. The loss of older means skills and knowledge are not passed on – ‘institutional
memory’ is lost. Women are particularly important in time of famine as they often have
knowledge of wild fruits that can be gathered.

Agriculture even the most basic subsistence type, does not operate in a vacuum. AIDS means
key services such as marketing cooperatives and extension are less efficient due to staff
attrition and declining morale. In Malawi attrition due to death among the Ministry of
Agriculture and Industry staff rose from 0.45 to 1.1 percent annually between 1996 and 1998
(Whiteside, 2008).

4 METHODOLOGY

4.1 The Study Area

The research was carried out in Mpingu EPA in Lilongwe ADD. Mpingu EPA is situated
in the southern part of the capital city, Lilongwe. Mpingu EPA has 19 sections and is
divided into 152 blocks. It has a population of 18,402 farming families. The area is
supervised by 21 AEDOs. Receiving an annual rainfall of 850mm (MoAFS, 2007) and
14 | P a g e
 lying along the Lilongwe-Kasungu plain, Mpingu EPA has a comparative advantage in
maize production.

4.2 Data Collection Methods

The project used primary and secondary data. Surveys, key informant interviews and
focused group discussions are methods the research used to obtain primary data.
Secondary data was collected from NGOS, National Statistical Office, Bunda library and
internet.

4.2.1 Survey
The project carried out a verification survey. A structured questionnaire was designed to
collect data from households in a representative sample. The questions were arranged in
a logical sequence. The researcher carried out interviews and recorded the responses
from the questionnaire. After that, the questionnaire was pretested to remove any errors
that may have occurred.

4.2.2 Key informant interviews

A number of interviews were conducted from various people who know the problem at
district level and in the study area. Some of the key informants were officials from the
MoAFS headquarters, DADO, AEDOs, village headmen and seasoned researchers in the
field of agricultural production economics. These interviews were conducted by
producing a carefully prepared checklist of important issues related to the research.

4.3 Sampling Design

Probability sampling methods were used to come up with a representative sample. A
combination of cluster and stratified random sampling was used. A multistage sampling
technique was carried out starting with simple random sampling in order to pick 5
sample blocks. This was done in this way because the study area has already been
divided into 152 geographic blocks (clusters) which were too big for the budget. The
sample cluster was divided into strata of male headed households and female headed
households. Then using proportional probability sampling method, the number of
households in each stratum was decided. To ensure that there was minimal random error,

15 | P a g e
 a sampling fraction was used. In this way the random error was minimized and the data
was used directly.

To find a sampling fraction:

 = /

Where:

SF= sampling fraction

N= population size

 = Sample size

Since Mpingu EPA’s population was 18402, a 95 percent confidence level was chosen.
The sampling error was estimated at 0.5 and p was 0.5. Therefore, to come up with a
sample size the following formula was used:

=
 


= = 
. ..


Nevertheless, the sample size was increased to 100 households because the research
needed sufficient data to come up with correct estimates. It was also assumed that design
effect and non response errors would be accounted for within the sample.

4.4 Data Analysis

After the data was collected, quantitative and econometric analysis was carried out to
come up with descriptive and inferential statistics. Hence, SPSS was useful to come up
with descriptive statistics.

5 THEORETICAL FRAMEWORK; JOHANNES AND MENDOZA-

ESCALANTE, (2007)

5.1 Economics of agriculture production

16 | P a g e
The basic thrust of economics of agricultural production at the micro level is to assist farmers
to attain their objectives through efficient farm allocation of resources over a given period of
time. Profit maximization could be achieved by maximizing output from a given resource or
minimizing the resources required for a given output. Agricultural productivity is
synonymous with resource ¨C productivity which is the ratio of total output to the
resource/inputs being considered. The basic concepts in productivity measurement are
Average Product (AP), Marginal Product (MP), Marginal Rate of Substitution (MRS),
Elasticity of Production (EP) and Returns to Scale (RTS). The knowledge of these concepts
can be used to study the three stages of the production surface (Fasoranti Olayiwola
Olujenyo, 2009).

Schultz’s (1964) ‘poor-but-efficient’ hypothesis – i.e. small farmers in traditional agricultural

settings are reasonably efficient in allocating their resources by responding positively to price
incentives – can be fairly considered as one of the enduring themes in rural
development economics over the past three decades. With respect to the long-term
effectiveness of the individual development strategy applied on small-scale farming the
level of efficiency of those farming activities has important implications: If farmers are
reasonably efficient, then an additional increase in efficiency requires the usage of more
productive inputs and/or the application of a more productive technology to shift the
production frontier upwards. If on the other hand current inputs and/or technology could be
used more productively, an improvement in the institutional setting - e.g. input markets,
infrastructure endowment, available extension systems, management and training services -
should be targeted to increase the efficiency on farm level. Hence, the two broad approaches
- technology development and transfer versus more efficient use of available technology
and resources on the individual farm level - can be considered as a continuum in the process
of development (Ali and Bayerlee, Poverty Incentives, and Development 1991 p. 2; Schultz,
Transforming Traditional Agriculture, 1975, as cited in Johannes and Mendoza-Escalante,
(2007)).
Assuming efficiency of small-scale farming could be based on the notion that farmers in a
more traditional agricultural setting depend largely on their own resources and
consequently managed to adjust their coordination and management efforts in the long-run to
the most efficient use of these resources. Assuming on the other side inefficiency in a more
dynamic and developed agricultural setting could be based on the reasoning that the

17 | P a g e
individual producer find it more difficult to adjust the allocative decisions to a
continuously changing production environment: “Farmers in this situation are likely to
be in a continual state of disequilibrium, and there will be high returns to improving
their information and skills to help them to adjust more rapidly and reduce technical and
allocative errors.” (Ali and Byerlee, Poverty Incentives, and Development 1991, p. 2 as cited
in Johannes and Mendoza-Escalante, (2007)).

Johannes and Mendoza-Escalante, (2007) reports that very poor performance of a small
farmer relative to others operating on the production frontier can be simply due to the
small scale of his/her agricultural operations and vice versa a good performance
relative to others can be simply due to the large scale of his/her operations compared
to the peer group average. Considering also the scale effects on efficiency could
deliver a more precise picture of the relative economic efficiency of small scale farms
in developing areas. If this could be empirically verified then a viable policy option in
both a more traditional as well as a more dynamic setting would be to enhance overall
economic performance on the firm level by delivering incentives for an increase in
the scale of operations and forming bigger production units by fostering farm
cooperations and/or mergers.

5.1.1 Returns to scale

As is well known the concept of returns to scale (rts) reflects the degree to which a
proportional increase in all inputs increases output. We refer to constant, increasing,
or decreasing rts as a proportional increase in all inputs results in the same, in a more
than proportional, or less than proportional increase in output. This basic economic concept
refers to a long-run factor-factor relationship where output may be increased by simply
changing all factors by the same proportion i.e. by altering the scale of the operation (see
e.g. Chambers, 1988). Hence, the observation that a farm has increased its productivity from
one year to the next does not imply that the improvement has been resulted from pure
technical and/or pure allocative efficiency improvements alone, but may have been (also)
due to technical change or the exploitation of scale economies or from some
combination of these three factors.
Consequently, beside technical inefficiency failure to maximize profit – i.e. maximize output
and minimize cost - in a given period has a systematic allocative inefficiency

18 | P a g e
component, which can involve an inappropriate input mix, an inappropriate output mix (i.e.
the scope of production in the case of multiple outputs) and an inappropriate scale. For a farm
to be profit efficient it requires technical efficiency and both input and output allocative
efficiency to be achieved at the proper scale.

Proposition: The overall economic efficiency of a small scale agricultural enterprise can only
be adequately assessed by also investigating its relative scale efficiency.

5.2 The Basic Model

5.2.1 The basic production model

Since labour is one of the factors of production, the theory of production was used. There
was a modification to the production theory in order to determine the optimum labour
usage. This is considered a modification because values and net benefits are used in the
analysis, rather than market prices and profits normally used in production theory
(Ng’ong’ola, 1987). Ng’ong’ola (1987), argues that because the production process
entails taken some risks and use of non marketed resources such as family labour, hence
non - money costs, e.g. risk premium and family labour opportunity cost, must be taken
into account.

Optimum labour was computed to give maximum profit. Although farmers may
maximize other objectives, they are assumed to maximize net profit since in reality
farmers do not deviate much from the norm of net profit maximization (Ng’ong’ola,
1987).

The general equation to find the optimum labour usage that was to maximize profit is
given below

 =  −   −   − ⋯ −   (1)

And optimal levels of the inputs are given below by simultaneous equations of the partial
derivatives:

= − =0
! $
"# "#
(2)

= − =0
! $
"& "&
(3)

19 | P a g e
 . . . .

. . . .

. . . .

= − =0
! $
"' "'
(4)

Where:

π = net profit (MK/ ha)

y = quantity of product

x1,x2 , ...,xn = levels of inputs

x1,= Labour; x2=Land, X3=Seed, X4=Fertilizer used in production of y

p = value of product (MK/kg)

w1 ,w2, ...,wn = wages of man hours worked(MK/hour) and costs of inputs in production of
y

Below are the axioms for the theory above:

1. The production period is monoperiodic.

2. All inputs and outputs are homogeneous i.e. only maize is grown.
3. The production function is given by a single twice continuously differentiable
function.
4. Constrained optimization.
5. The goal of the smallholder is to maximize net benefits i.e. most smallholders
grow maize for food.

From these axioms the following assumptions were formulated:

1. Smooth causal relationship exists between the inputs and products i.e. first
$
"
derivative prevail. (5)

2. Diminishing returns prevail with respect to each factor i.e.

< 0.
&$

"&
(6)

20 | P a g e
 3. Decreasing returns to scale prevail. An equal proportionate increase in factors results
in a less than proportionate increase in product i.e.

∑* , * +, < 1
$ "
"+ $
(i=1,2,3,...,n) (7)

(Ng’ong’ola, 1987).

5.2.2 Derivation of the Cobb – Douglas Functions

Salvatore (2003) notes that the Cobb – Douglas production function is the simplest and most
widely used production function in empirical work today. The formula for the Cobb –
Douglas production function is
. = /01 23
Where Q= output in physical units, L = quantity of labour, k = quantity of capital, and A, α
and β are positive parameters estimated in each case from the data. The parameter α refers to
the percentage increase in Q for a 1 percent increase in L, while holding K constant. Thus α is
the output elasticity of labour. Similarly the parameter β refers to the percentage increase in Q
for a 1 percent increase in K, while holding L constant. Thus β is the output elasticity of
capital. The 4 + 6 = 1 shows degree of homogeneity (returns to scale). Thus we have
constant returns to scale. Even though we experience constant return to scale, the
Cobb –Douglas production function was chosen because it is easy to calculate factor
elasticities. However, the Cobb-Douglas function was used in its log linear transformation to
run a production function regression using Ordinary Least Squares Method (OLS). In its
stochastic form, the Cobb-Douglas production function is represented as

7 = 89:3; <= ; = , , ?, … , A
Where :
e is the base of the natural logarithm
is the stochastic disturbance term
is the vector of variable resources

The logarithmic transformation

21 | P a g e
The model is linear in parameters and ℰ and was estimated using ordinary
least squares of the linear regression model Gujarati, 1998, Edriss et al., 2005.
The variables that were run in the model include both physical and qualitative variable.
These include, land, potential available labour, seed, fertilizer, age of household head,
sex of the household head, family size and marital status. Thus the model becomes

cd 7 = cd 3 + e cd 0fghij
+ e cd 0fk + e? cd lk + em cd njopqpj + 3 lr + 3 /s
+ 3t ufjpofqvofoiv + =

6 RESULTS AND DISCUSSION

Descriptive statistics

6.1 Socioeconomic Variables

6.1.1 Age of the household head

Age of the household head is very critical in understanding most production decisions made
on the farm. Mostly, age of the household head can be taken as a proxy of experience while
we suppose that the older the person the more experience you have and therefore the better
the decisions you can make. It is quite easy to hypothesize that older persons are bound to
make better production decisions than younger people. On one hand, this proposition seems
reasonable for child headed households since children do not have enough tacit knowledge
and experience to make sound and well informed decisions. This results in very bad
production decisions which end up in low productivity per unit factor of production.
Eventually, child headed households are more likely to have low outputs and be chronically
food insecure. The proposition is not likely to be true on the converse. The research proves
the contrary. Having an average age of the household head at 40 years old with a standard
deviation of 15 years, the age range is widely spread among the population. This is quite
paradoxical to the hypothesis. However, the researcher observed that older people were more
inclined to older and obsolete technologies which contributed little per unit factor
productivity – for instance, use of old planting spacing in maize like 90cm×90cm×1cm
reduces the productivity of land. In addition, older people were more inclined to local maize
varieties than hybrids – another component that reduces output effectively. The oldest
22 | P a g e
household head was 87 years old while the youngest was 21 years old. The ages between 21
and 45 indicated 68 percent of the total labour force. No incidences of household heads below
20 years old were identified. These results evidently falsify the hypothesis that the older the
farmers the better the decisions they make.

6.1.2 Marital status of household head

Smallholder farming is almost exclusively dependent on family labour. In this case the family
poses as an institution that provides farm labour. The previous statement and empirical
evidence asserts that the larger the family the more labour will be available for most farming
activities. More importantly, a married couple is in a better position to make good decisions
through consultations with the spouses and siblings. The research findings further
authenticate these assertions – a thing which underscores the importance of understanding the
marital status of the household head. Most household heads reported that they were married.
This is evidenced by 88 percentage prevalence. 7 percent were single and 4 percent were
widowed.

Table 1 Marital status of household head

Frequency Percent
Valid single 7 7.0
married 88 88.0
divorced 1 1.0
widowed 4 4.0
Total 100 100.0
Total 123

6.1.3 Education level of the household head

Education level of household head
In a firm skilled labour is valued differently from unskilled labour. In this case, skilled labour
is more valuable than unskilled labour. The microeconomics of agricultural production
employs the same principles from theory of the firm. As people move further up the
education ladder, they acquire more skills which help them in their production decisions. This
points towards a proposition that those who spend more years in school are likely to get
higher output because of the skills they acquire. Literacy rates in the study area present a
serious problem for it was found that 81 percent of the respondents reported that they
23 | P a g e
attended only primary school. Of these, a larger proportion, over 70 percent attended junior
primary school. This implies that their skills are largely underdeveloped and they are likely to
make poor farming decisions. 12 percent attended secondary school. This category may be
that which may possibly make better decisions concerning agricultural production. 7 percent
attended tertiary education – the people that can make sound production decisions.

Table 2 Marital Status of the Household Head

School level Frequency Percent

Primary 81 81.0
Secondary 12 12.0
Tertiary 7 7.0
Total 100 100.0

6.1.4 Household size

Household size is one of the major determinants of labour productivity and efficiency on the
farm. The research reported that the average household size was 5 people which is slightly
lower than the national average (NSO, 2004). This is so particularly because the population is
still in its productive stage and they are still producing children. A standard deviation of 2
people per household indicates that utmost a household can be expected to have at least three
members and utmost 7 people.
Table 3 Household size

N Mean Std. Deviation

household size 99 4.90 1.992

Valid N (listwise) 99

6.2 Factors of production

The factors that were used in the production of maize were land, labour, fertilizer and seed.
Below is a thorough description of how these factors are utilized on the farm.

6.2.1 Labour

6.2.1.1 Descriptive Statistics for labour use

The importance of labour as a factor of production on a farm cannot be overemphasized. In a
country where mechanisation is a farfetched dream which creates a big productivity gap, the
only way to take the edge off this gap is intensifying labour use. Further, in a family economy
24 | P a g e
a smallholder farm is seen as a micro firm (Sachs, 2005) and it can be seen that productivity
of this firm can be enhanced by intensifying labour use since mostly it cannot afford capital
goods. The research findings show that 85 percent of the respondents reported that they used
family labour while 15 percent hired casual labour. This shows that family labour is used
intensely on the farm.

Potential labour source and labour utilization

Table 4 Descriptive Statistics for potential labour source

N Mean Std. Dev

household size 99 4.9 1.99
Family Labour 100 3.96 3.57
No. of labourers
hired 97 2.1 9.1

There were no traces of communal labour. On average the household size was 5 and potential
available labour was 4 people per household with a standard deviation of four people per
household. Of the houses that hired casual labour an average of 2 persons was found.

Figure 1 Labour Use Intensity Source: Researcher

It was observed that land preparation took most of the time by taking 36 percent of all the
time spent farming. This was followed by weeding, bunding, processing, fertilizer
application, harvesting, planting and transportation with 22, 15, 7, 6, 6, 5 and 3 percent
respectively. Land preparation took on average 22 days while weeding took 13 days.
However, the statistics show that 69.1 percent of the respondents said that labour was
sufficient to complete all the farm operations while 30.1 said it was not sufficient.68 percent
of the respondents stated that they do not hire extra labour but solely depend on family
labour. Only 31percent hired extra labour.

25 | P a g e
 Table 5 Hired Labour

Frequency Percent
No 24.4 30.9
Yes 54.5 69.1
Total 78.9 100
Total 100

Table 6 Labour Use

Activity Mean number of man days

land preparation 22.03
planting 3.24
weeding 13.02
bunding 9.51
fertilizer application 3.34
harvesting 3.77
transport 1.79
processing 5.67

Reasons for not hiring labour.

Among the reasons for not hiring labour on the farm, 91 percent reported that they do not
have sufficient finances, while 1.4 percent said they do not trust the labourers to do a good
job on the farm.
Table 7 Reasons for not hiring extra labour

Frequency Percent
Insufficient Finances 67 91.8
Untrustworthiness of
labourers 1 1.4
Total 73 100.0

The main problem that farmers were facing was poor health. 50 percent of the respondents
attributed this problem to HIV/AIDS, malaria, malnutrition prevalence. 25percent said that
26 | P a g e
the marginal benefit from the extra input was too little. While 12 percent reported
repo that the
main problem was food insecurity.

Table 8 Problems on the farm

Frequency Percent
lack of skill 11 11.82
poor health 47 50.53
low incremental
gain for extra input 24 25.80
food insecurity 11 11.82
Total 93 100

6.2.2 Land
Land as a factor of production is one of the constraints to agricultural production in Malawi.
It is not very surprising that most of the farmers in Malawi are constrained by a per capita
holding size of 0.5ha
5ha per family. Surprisingly,
S at Mpingu the average land
nd holding size was
1.039 ha per household. Almost all farmers cultivate their farms under customary land tenure.
It is this reason that made 98.7 percent of the respondents’ report that they farm in their own
land.

Land Ownership

100
80
Percentage

60
40
20
0
Own land Rented Borrowed
Land Tenure

Figure 2 Land Tenure Source: Researcher

The remaining 1.3 percent is shared between rented and borrowed land.
land There were no
incidences of leasehold or freehold.

6.2.3 Fertilizer

27 | P a g e
Fertilizer is the major constraint to agricultural production. The fact that maize is a heavy
feeder, taking a lot of nitrogen for its vegetative growth, it requires huge amounts of fertilizer.
This is why farmers that did not apply any fertilizer got barely enough to sustain them all
year round. Of late, prices of fertilizer have soared. As a result a lot of farmers have no access
to this precious commodity which in turn results in chronic food insecurity. However, the
coming of the targeted fertilizer and input subsidy program has made fertilizer accessible to
some farmers. In this case farmers buy a 50kg bag of fertilizer at K800.00 ($5.71) from
K9500.00 ($67.86). The subsidy program caused a lot of discrepancies in terms of estimating
the price of fertilizer in the EPA. This came up because the study did not consider separating
beneficiaries and non beneficiaries of the subsidy program. This anomaly is further explained
by the huge standard deviation in the cost of fertilizer. The average total cost of fertilizer
applied on the farm was K13564.80 ($96.89). However, this can be clarified if the standard
costs estimated by the Crop Science Department at Bunda College can be used as a proxy for
costs of production per hectare.

Table 9 Input Costs of maize as Used in the Gross margin Analysis

Inputs Costs in MK

Seed 8375

Basal fertilizer + subsidy 1600

Top dressing fertilizer + subsidy 3200

Basal dressing no subsidy 1100

Top dressing fertilizer no subsidy 22000

Storage Sacks 12000

Storage insecticides 1500

Total with subsidy 26675

Total without subsidy 44975

Source: Kabambe V. H., 2009

Table 10 Fertilizer and Manure Costs and Quantities

N Mean Std. Dev

fertilizer quantity 96 2.8344 4.76
cost of fertilizer 96 13564.80 21339.71
manure quantity 98 636.2760 2524.55

28 | P a g e
 Valid N 95

From the farmers that were sampled, it was indicated that on average, farmers applied 3 bags
of fertilizer per hectare. The usual fertilizers used are 23:21:0+4S and UREA. As the use of
fertilizer is continued, environmentalists pose concerns that fertilizers cause negative
externalities such as soil exhaustion and water pollution. As such the use of organic manure is
being advocated. Many people have not adopted application of manure to their farms. It
should, however, be noted that diffusion of this innovation is still a long way to go as only an
average 636.27kg of manure was applied and only 30 percent of the respondents applied
manure to their farms.

6.2.4 Seed
Farmers in Mpingu EPA plant local and hybrid maize varieties. Some of the seed varieties are
are DK8071, SC6031 and some older varieties. On average people used 25kg of seed per
hectare and the average total cost of seed was K3303.59 ($23.60). It should be noted that
some farmers did not buy seed because they either used local varieties or recycled the seed
from the previous harvest so the cost of seed can be taken as the opportunity cost of the local
seed which is valued at K50 per Kg. It is also recommended that the average cost of seed
should be taken from table 7.
Table 11 Quantity (Kg) and Cost of Seed (MK)

N Mean

seed quantity 99 25.07

cost of seed 77 3303.59

6.3 Inferential Statistics: Cobb-Douglas Production function estimates

Table 12 Model Summary (b)

Model R R Square Adjusted R Std. Error of Durbin-Watson

29 | P a g e
 Square the Estimate

1 .776(a) .603 .573 .54347 1.813

a Predictors: (Constant), sex, ln land, age, ln labour, ln fertilizer, ln seed
b Dependent Variable: ln output

The table above shows the model summary. Having an R2=0.603 means that 60.3 percent of
variations in output were explained by the explanatory variables included in the model. The
standard error of the estimate was 0.54. The Durbin-Watson Coefficient was 1.813 which
indicates that there was no incidence of multicollinearity in the variables of the collected
data.

Table 13 ANOVA (b)

Sum of
Model Squares df Mean Square F Sig.
1 Regression 36.715 6 6.119 20.718 .000(a)
Residual 24.220 82 .295
Total 60.934 88
a Predictors: (Constant), sex, ln land, age, ln Labour, ln fertilizer, ln seed
b Dependent Variable: ln output

The ANOVA table presents the overall significance of the model. In this case, an F-value of
20.718 is significant at 1 percent which implies that the functional form and model
specifications were correct.
Table 14 Coefficients (a)

Unstandardized Standardized t Sig.

Coefficients Coefficients
Model B Std. Beta
Error
1 (Constant) 3.798 .608 6.245 .000*
Potential Labour .199 .119 .136 1.663 .100***

Seed .205 .078 .248 2.621 .010**

Fertilizer .487 .092 .486 5.278 .000*
Land .192 .127 .153 1.517 .133
Age 1.688E-03 .004 .031 .383 .703
Sex 5.124E-03 .111 .003 .046 .963
Access to Extension 3.599E-02 .197 .013 .182 .856
a. Dependent Variable: ln Output

30 | P a g e
b. Predictors: (Constant), sex, ln land, age, ln Labour, ln fertilizer, ln seed
c.*significant at 1%, **significant at 5%, ***significant at 10%, none = not significant

The production function analysis is presented in the table above. The Table shows that a
positive relationship exists between total output and Land, Fertilizer, Labour, Seed, Age, and
Sex. This implies that as more of these variables are employed, there will be an increase in
total output of maize. Labour was significant at 10 percent, seed at 1 percent, fertilizer at
1percent while age, sex and access to extension services were not significant.

6.3.1 Productivity Analysis

Table 15 Elasticities of production

Variables Elasticities
Potential Labour 0.1990
Seed 0.2050
Fertilizer 0.4870
Land 0.1920
Age 0.0017
Sex 0.0051
Access to Extension 0.0360
Return To Scale 1.1258

The input elasticities are presented in Table 8. Potential labour showed an inelastic elasticity
i.e. if labour is increased by 1 percent, output will increase at a diminishing rate of 0.1990
percent. If seed is increased by 1 percent output is expected to increase by 0.2050 percent.
This can be achieved using modern farming methods for example the Sasakawa2 method
otherwise the results may be inconclusive. If fertilizer application is increased by one percent
output is expected to increase by 0.4870 percent. Recommended fertilizer application rates
are key to achieve this. If a farmer increases land by one percent, output should be expected
to increase by 0.1920.

Age, Sex and access to extension services were not significant but they positively influence
output though negligibly. They are not factors of production but are socio-economic factors
that affect production. Age, which may be a proxy of experience, was not able to capture the

2
In this method, you plant one seed per planting station. Spacing is about 25cm X 75cm X 1cm apart
31 | P a g e
influence on output as older people have been using very obsolete technologies and their
productivity tends to even out when they age.

To all intents and purposes, the results showed that farm size (land), labour, seed and
fertilizer have positive but less than unity elasticities indicating a decreasing positive returns
to each of the factors. They are therefore efficiently utilized and hence their use is in stage II
(i.e., the rational zone) of the production function. Thus, diminishing returns prevail.
However, the return to scale (RTS) estimated as 1.1258 implies increasing returns to scale,
i.e. that an equal proportionate increase in all factors results in a greater than proportionate
increase in product i.e.
y z
wx {x { > 1
yz 
(i=1,2,3,...,n).
This tells that farmers have potential to increase their production if they increase all factors
proportionately. In addition, it can be observed that if farmers increase their farm productivity
in one year and the following year it does not simply lead to a conclusion that the
improvement is as a result of the technical and allocative efficiency improvements alone. It
may also be due to technical change or the exploitation of scale economies or from some
combination of these three factors (Johannes and Mendoza-Escalante, 2007). From the
proposition in the theoretical framework, the results clearly authenticate that the overall
economic efficiency of a small scale agricultural enterprise can only be adequately assessed
by also investigating its relative scale efficiency.

These results have considerable implications on small scale agriculture in Malawi. As it can
be observed, to spur the small scale farming development approach there is need to use more
productive inputs for instance improved seeds, conservation agriculture and improved access
to information and technology. Usage of more productive technology like Sasakawa,
conservation agriculture, and irrigation can shift the production frontier upwards. It is also
evident that if the current technology could be used productively, there should be a need to
increase the efficiency at farm level. There should be improved input markets, infrastructure
endowment and management, training and extension systems.

6.3.2 Economic optimization

32 | P a g e
The profit maximization criterion was used in this analysis. Microsoft Math software was
used to come up with first order, second order and all the other mathematical computations to
come up with profit maximization levels and finally, holding output constant, least cost
combinations were arrived at. The profit maximization problem was denoted by:

 =  −   −   − ⋯ −  

 = 50 − 129 − 131 − 190~

 = 50 .€€ . ~.‚ƒ„  − 129 − 131 − 190~

First Order Necessary Conditions

∂y 199 . ~ .‚ƒ„

= −129 + =0
∂x 20 .ƒ

∂y 41 .€€ ~ .‚ƒ„

= −131 + =0
∂x 4 .„€

∂y 487 .€€  .

= −190 + =0
∂x~ 20~ .~

199 . ~ .‚ƒ„

−129 + =0
Š 20 .ƒ ’
‰ 41 .€€ ~ .‚ƒ„ ‘
nsolve ‰{ −131 + = 0 Œ,Œ{ ŽŒ,Œ{~ ŽŒ,Œ{ Ž‘
4 .„€
‰ ‘
487 .€€  .
−190 + =0
ˆ 20~ .~ 
Second Order Sufficient Conditions

∂ y 159399 . ~ .‚ƒ„

= − <0
∂x 20000.ƒ

∂ y 8159~ .‚ƒ„
= <0
∂x 4000x .„€  .ƒ

∂ y 96913 .
= <0
∂x~ 20000~ .~  .ƒ

Where ” , ” , ”~ … ” are as defined in section 5.2.1. After solving for the x’s the optimal values
can be found.

33 | P a g e
Due to the complexity of the equations a gross margin analysis can be used as a close estimate of
potential labour and optimum input use.

Table 16 Gross Margin (GM) for Maize

Input/income value, MK Maize

Total labour man days/ ha 139

Total value of labour 27800

Variable costs, + subsidy 26,675

Variable costs no subsidy 43,875

Expected yield kg/ha 4716

Value of product MK 41.96

GM/ha with subsidy 197,883

GM/ha without subsidy 143,408

GM/man days with subsidy 1032

GM/man days without subsidy 908

Source: Kabambe V. H, 2009

7 CONCLUSION
Malawi’s smallholder farmers are among the poorest but efficient farmers in the region. The
results of this research provides evidence for this assertion as critical factors of production
such as land, labour, seeds, fertilizer all show considerable statistical significance.
Furthermore, these farmers have potential to increase production by increasing all factors of
production as evidenced by increasing returns to scale estimate. It should also be noted that
labour is, by large, among the constraints to efficient agricultural productivity since most
smallholder farmers indicated that they did not have enough labour to complete their farming
activities. Costs of production, malnutrition, HIV and other diseases are some of the
constraints to efficiency in agricultural production. The socioeconomic variables such as age,
marital status, and education of the household head have shown proven importance in
decision making on the farm. Household size ensured availability of labour within the
farming activities while marital status and education level indicated the quality of decisions

34 | P a g e
made on the farm. In conclusion, labour is one of the constraints affecting the productivity
and efficiency of smallholder farmers and smallholders are efficient but poor.

8 RECOMMENDATIONS
The research has considerable implications on agriculture, extension, marketing and input
policy. The recommendations therefore are grouped from general agricultural to input policy.
Below are the recommendations:

a. Labour is one of the factors of production which is also a constraint to most

smallholders. To reduce this problem, extension workers should emphasize on proper
allocation of small holder labour on the farm. In addition,

b. Since most smallholder farmers are efficient but are not operating under economies of
scale, a viable policy option can be to facilitate formation of cooperatives in
smallholder farmers.

c. There should be improved input markets, infrastructure endowment and management,

training and extension systems in order to ensure improved flow of information from
farmers to extension workers, researcher, and consumers vice versa and maximize
profits.
d. Farmers should be using improved farming technologies such as conservation
agriculture, improved seeds, and herbicide technology to ensure increased production.

9 BIBLIOGRAPHY

Ashok P. et al, (1995) Measurement of Economic Efficiency in Pakistan Agriculture.

American Journal of Agricultural Economics. 77(1 - 2)

Conroy A. C. (1993) the economics of small holder maize production in Malawi with
reference to the market for hybrid seed and fertilizer, PhD thesis, University of Manchester,
Manchester, UK.

Conroy, A.C, (2005) The history of development and crisis in Malawi. In: Conroy et al.,
(2005) Poverty, AIDS and Hunger Breaking the Poverty Trap in Malawi. Palgrave
Macmillan. New York.

35 | P a g e
Edriss A. K, et al. (2005) The impact of labour market liberalization on maize
productivity and rural poverty in Malawi [online]. Policy Analysis for Sustainable
Agricultural Development (PASAD) Project, University of Bonn, Germany. Available from:
www.pasad.org/

Johannes S. et al. (2005) The Global Food and Product Chain – Dynamics, Innovations,
Conflicts, Strategies Smallholder Maize Productivity Under Alternative Soil Fertility
Management Options In Malawi. Deutscher Tropentag, Hohenhein [online] Available from:
http//www.tropentag.de/2005/abstracts [03/10/2008]

Jayne T.S. et al, (2005) HIV/AIDS and the Agricultural Sector in Eastern and Southern
Africa: Anticipating the Consequences [online]. International Food Policy Research Institute.
Availablefrom:http://www.ifpri.org/events/conferences/2005/durban/abstracts/jayneabstract.p
df

Kumwenda J. T. D. et al (1997) Soil Fertility Management for Maize Cropping Systems of

Smallholders in Southern Africa: A Review [online]. NRG, paper 96-02. Mexico D.F:
CIMMYT. Available from: http://www.cimmyt
.org/research/nrg/SFMR/htm/SFMRcontents.html

Lovewell, M. (2008). Economics. [online].Mc-Graw Hill. Available from: McGraw-Hill

Website: http://highered.mcgraw-
hill.com/sites/0070891478/student_view0/chapter7/key_terms___glossary.html [11 /12/
2008]

Ministry of Agriculture and Food security, (2007) 2006/2007 Annual Agricultural Statistical
Bulletin. Government of Malawi. Planning Department, Lilongwe.

Ng’ong’ola H.D, (1987). An economic analysis of smallholder fertilizer use in the Lilongwe
Rural Development Project (LRDP), Malawi: implications for agricultural policies. PhD
thesis, University of Wales, Aberystwyth, UK.

36 | P a g e
Phiri, M. T, (2006) Dynamics of internal labour migration in Malawi, unpublished PhD
thesis, Bunda College of Agriculture, University of Malawi, Lilongwe Malawi.

Rubey, R (2003) Malawi’s food crisis: Causes and solutions. [online]. Available from:
http://www.sarpn.org.za/focus.php [03/10/2008]
Sauer J. and Mendoza-Escalante A. (2007) Farming in the Eastern Amazon – Poor but
allocatively efficient [online] Available From:
http://nhjy.hzau.edu.cn/kech/nyjjx/zl/xueshulw/en/Poor%20but%20allocatively%20efficient
%E2%80%94evidence%20from%20the%20Eastern%20Amazon.pdf [18/1/2010]

Siegal, J. And Alwang. (2000). Measuring the impacts of agricultural research on poverty
reduction: An Application to Malawi. Agricultural Economics , 29(1).

Topouzis, D. (2000). Addressing the Impact of HIV/AIDS on Ministries of Agriculture: Focus

on Eastern and Southern Africa. FAO Corporate Repository Document [online]. Available
from:FAO Website: http://www.fao.org/docrep/005/y4636e/y4636e01.htm#TopOfPage [12/
11/ 2008]

Whiteside A. (2008). HIV/AIDS an Introdution. Oxford Press. London

Whiteside, M. 2000. Ganyu Labour in Malawi and its Implications for Livelihood Security
Interventions: An Analysis of Recent Literature and Implications for Poverty Alleviation.
Agricultural Research and Extension Network Paper, No. 99. Overseas Development
Institute, London.

37 | P a g e
10 APPENDIX 1

Questionnaire

University of Malawi

Bunda College of Agriculture

P.O. Box 219

Lilongwe

LABOUR PRODUCTIVITY AND EFFICIENCY IN SMALLHOLDER MAIZE PRODUCTION:

A CASE OF MPINGU EXTENSION PLANNING AREA.

Questionnaire

SECTION A: IDENTIFICATION OF INFORMATION

Name of interviewer Henry Kankwamba Village

Name of respondent Date of Interview
Name of EPA HH ID
Section Checked by
Name of ADD

SECTION B: DEMOGRAPHIC CHARACTERISTICS

Name Age Sex Marital Education Occupation Availability Household

status Level size
(years in
school)

Code number Variable code 1 2 3 4 5

i. sex male female
38 | P a g e
 ii. Marital status single married Divorced Widowed
iii. Education level Primary Secondary Tertiary Illiterate
iv. Occupation Farming Business Formal Casual Other
employment Labourer
v. Availability Available Unavailable

SECTION C: IDENTIFICATION OF FACTORS THAT AFFECT SMALLHOLDER MAIZE PRODUCTION

C.1.0 POTENTIAL AVAILABLE LABOUR (circle where appropriate)

C.1.1 What is the source of labour on your farm?

A. family labour B. Casual labour D. Contract workers E. Communal workers

How many people stayed in your house and worked in the maize garden?

Family Gender Age Education Major Time spent

members level activity farming

Gender codes: 1=male 2=female

Activity code: 1= farming 2. Full time education 3=off-farm employment 4=own

business

C.1.2 How many labourers did you employ last growing season?

C.1.3 How much labour did you use for : (for this question check on landscape page)

C.1.4 How much money do you pay the labourers.

C.1.5 Is the labour that you use sufficient to complete all the tasks on your farm? A. Yes B. No

C.1.6 If No, do you hire extra labour? A. Yes B. No

C.1.7 If No to C.1.6, Why don’t you hire extra labour?

A. Lack of sufficient money B. Labour is scarce C. Quality of labour is poor D. Lack of trust

E. other (specify)

C.1.7 What was the total cost incurred for hiring out labour on your farm?

39 | P a g e
C.1.8 What labour problems do you experience on your farm?

A. lack of skill B. Poor health and diet C. Low extra gain for extra input D. Food insecurity

E. Other (specify)

C.1.9 How many months did the maize season last?

SECTION D: LAND TYPE, SIZE , ACQUISITION AND OUTPUT OBTAINED

D.1. 1 What was the total land area for your gardens? (Specify hectarage)

CODE 1 2 3 4 5 6
Hectares <0.5 0.5 - <1.00 1.00- <1.50 1.5-<2.00 2.00<2.50 >2.50
No. gardens

Land type Land Size Land Output obtained

(actual size) Acquisition
Land under cultivation
Own land
Rented land
Borrowed
Land under fallow

D.1.2 If rented how much was the rent? MK

SECTION E: SEED

E.1.1 What was the type of seed that you used last season?

Type of seed Amount (Kg) Cost (MK)

SECTION F: FERTILIZER APPLICATION

F.1.1 Did you apply any chemical fertilizer to your maize garden? A. Yes B. No

F.1.2 If Yes, how much fertilizer did you apply (number of bags) to all maize varieties last season?

FERTILIZER LOCAL MAIZE HYBRID COMPOSITE COST (MK)

TYPE

40 | P a g e
UREA
CAN
23:21:0+4s
20:20:0
Other (specify)

F.1.3 Did you apply any organic fertilizers (manure) to your garden? A. Yes B. No

F.1.4 if yes, how much manure did you apply to your garden last season?

MANURE LOCAL HYBRID COMPOSITE AMOUNT COST

TYPE MAIZE (NGOLO) (MK)
Poultry
Goat
Cattle
Mixture
Other
(specify)

SECTION E: ACCESS TO CREDIT

1. Do you have access to credit facilities? A. Yes B. No

2. If yes, what is the microfinance institution that gives you credit?

3. If no, what are the reasons why?

A. Lack of creditworthy B. Lack of collateral C. Lack of institutions D. Fear of default E.

Other___________

SECTION F: ACCESS TO EXTENSION SERVICES

1. Do you have access to extension services? A. Yes B. No

2. If yes, how often does the extension worker teach you about how to utilize farm labour on
the farm? A. Very often B. Often C. seldom D. Never

3. If no, why

a.They favour other farmers b. they are very few c. I don’t like them d. other________

41 | P a g e
42 | P a g e
C.1.3 How much labour did you use for

Labour source Task Amount of Number of Cost (MWK)

labour days
(workers)
Family labour Land preparation
Planting
Weeding
Bunding
Fertilizer
application
Harvesting
Transporting
produce
Processing
Hired labour Land preparation
Planting
Weeding
Bunding
Fertilizer
application
Harvesting
Transporting
produce
Processing
Communal Land preparation
labour Planting
Weeding
Bunding
Fertilizer
application
Harvesting
Transporting
produce
Processing
Totals
11 APPENDIX 2
Economic Optimization Calculations

2|Page