This article was downloaded by: [New York University]

On: 18 October 2014, At: 07:22

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer Street,
London W1T 3JH, UK

The Journal of Peasant

Studies
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/fjps20

Output per acre and size

of holding: The logic of
peasant agriculture under
semi‐feudalism
a
Graham Dyer
a
Department of Economics, School of Oriental
and African Studies , University of London ,
Thornhaugh Street, Russell Square, London,
WC1H 0XG
Published online: 05 Feb 2008.

To cite this article: Graham Dyer (1996) Output per acre and size of holding:
The logic of peasant agriculture under semi‐feudalism, The Journal of Peasant
Studies, 24:1-2, 103-131, DOI: 10.1080/03066159608438632

To link to this article: http://dx.doi.org/10.1080/03066159608438632

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all
the information (the “Content”) contained in the publications on our
platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy,
completeness, or suitability for any purpose of the Content. Any
opinions and views expressed in this publication are the opinions and
views of the authors, and are not the views of or endorsed by Taylor
 & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information.
Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities
whatsoever or howsoever caused arising directly or indirectly in
connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access
and use can be found at http://www.tandfonline.com/page/terms-and-
conditions
Downloaded by [New York University] at 07:22 18 October 2014
 Output Per Acre and Size of Holding:
The Logic of Peasant Agriculture Under
Semi-Feudalism

GRAHAM DYER
Downloaded by [New York University] at 07:22 18 October 2014

The inverse relationship between farm size and productivity is widely

accepted as a 'stylised fact' of agriculture in developing countries, a
generalised phenomenon observed in widely differing agro-climatic
conditions and agrarian structures. The significance of the inverse
relationship as a crucial developmental issue cannot be overemphasised.
The inverse relationship constitutes a major component of the economic
rationale for redistributive land reform, and has obvious importance for
policy issues concerning land reorganisation.
The debate over the inverse relationship has been extensive, constituted
predominantly by two distinct agendas: first, a discussion of the factors
which give rise to and explain the inverse relationship in the static context;
second, and later, following the main thrust of the green revolution, a
discussion of the breakdown of the inverse relationship in the dynamic
context of changing technology. The former strand has focused on two main
areas: first, qualitative factor differentials across farm size (soil quality,
farm management, etc.) and the mechanisms that generate such
differentials; second (quantitative) factor intensity explanations, for the
most part within an explicitly choice theoretic framework.
The social context within which a differentiated peasantry operates, and
which can generate an inverse relationship, is rarely discussed. Only
Bharadwaj [1974a; 1974b; 1985] and Patnaik [1972; 1979; 1987] come
close to this approach. However, while Bharadwaj lays the foundations for
a political economy framework, she leaves the analysis at the level of
technical and market relations and the constraints on participation imposed
by resource positions. Patnaik, who operates very firmly within a class
analysis framework, uses it to explain away the inverse relationship as a
statistical illusion. This strand in the literature alludes only very briefly to
the possibility of the inverse relationship breaking down under changing
Graham Dyer, Department of Economics, School of Oriental and African Studies, University of
London, Thornhaugh Street, Russell Square, London WC1H 0XG.
 104 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

production conditions. This latter strand in the literature often relies on the
weaker (qualitative) explanatory foundations to discuss the disappearance
of the inverse relationship in the dynamic context of the impact of the new
technology and its differential adoption rate across farm size.
Byres [1972; 1974; 1977a; 1977b; 1981; 1986a; 1986b; 1990; 1991] and
Byres and Dyer [forthcoming] made the first serious attempt to integrate
both the static and dynamic analyses into a rigorous and unified class
theoretic framework in the context of agrarian transitions. This approach
explains both the generation of the inverse relationship and its breakdown.
Class relations of exploitation in a relatively backward agriculture exert
forces of economic compulsion on the poor peasantry to intensify
Downloaded by [New York University] at 07:22 18 October 2014

cultivation, thus generating an inverse relationship. However, with the

increasing penetration of capitalist relations of production and exchange,
and with capital deepening on (proto-) capitalist farms, the inverse
relationship breaks down as large farms capture the economies of scale
associated with the new technology.
The empirical evidence for the existence of an inverse relationship
between farm size and farm productivity is both historically and
geographically widespread, ranging from pre-revolutionary Russia and
China to contemporary poor countries in Asia, Africa and Latin America.
Indeed, the inverse relationship is widely regarded as a 'stylised fact' of
traditional agriculture [Bardhan, 1973: 1370], a generalised phenomenon
observed in many developing countries characterised by widely different
agro-climatic conditions, agrarian structures and cropping patterns [Cornia,
1985: 514-15; Ghose, 1979: 27].
Griffin [1974: 228] notes:
A striking feature of the agricultural systems of virtually all poor
countries is that yields per acre rise as average farm size declines.
That is, the smaller the farm, the greater the average productivity of
land. Conversely, the larger the farm, the greater the average
productivity of labour. Since land is usually the factor in most acute
shortage, the farms with the highest yields per acre are normally the
most efficient. Even in countries where the average farm is very small,
such as in India, it has been demonstrated that those farms which are
smaller than average are economically the most efficient.
The modern debate comprises a large theoretical and empirical literature
which came into being in the Indian context. Central to the modern Indian
debate on farm size and productivity is the 17-volume Farm Management
Studies (FMS) data collected by the Directorate of Economics and Statistics
and the Research Programmes Committee over three years: 1954-55,
1955-56 and 1956-57.
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 105

The Indian debate proper begins with Sen's seminal 1962 article in the
Economic Weekly [Sen, 1962]. There he notes the following observation:
'By and large, productivity per acre decreases with the size of holding.' Sen
further notes that this trend holds in most areas for value-added too. Khusro
[1964] and Bharadwaj [1974a] carry out more detailed studies using OLS
regression techniques on the grouped FMS data. The constant generation
and analysis of data since then continues to confirm the finding of an
inverse relationship.

SOME APPARENT POLICY IMPLICATIONS

Downloaded by [New York University] at 07:22 18 October 2014

The significance of the inverse relationship as a crucial developmental issue

cannot be overemphasised. Such a relationship has obvious importance for
policy issues such as land ceilings and redistribution, as well as co-operative
and other forms of land reorganisation, involving discussion of factors such
as market imperfections and the institutional framework of traditional
agriculture [Bardhan, 1973: 1371]. The apparent implications of such a
relationship constitute a major component of the economic rationale for
redistributive land reform and a small farm bias in agricultural development
strategy.
Cornia [1985: 532] writes:
Because of the demonstrated superiority of small vis-a-vis large
fanning, land redistribution would have, if thoroughly implemented,
immediate beneficial effects in terms of output growth, enhanced
income distribution and, as a result, of alleviation of rural poverty. It
would also bring about a resource use more in line with the factor
endowment of developing countries by increasing labour absorption
... while forestalling premature labour-displacing mechanisation.
Berry and Cline, regarded by many as the definitive statement on the
inverse relationship, argue that the finding of systematically higher land
productivity and comparable levels of total factor productivity on small
farms as opposed to large farms require redistribution of land to small
farmers who apply labour more intensively, and the improvement of small
farm access to credit and new technology. Both strategies will improve
equity and increase output levels. With such land redistribution, the authors
claim that: 'The optimal post-reform farm size, in the absence of technical
returns to scale, will be merely the total agricultural area divided by the total
number of families in the agricultural labour force' [1979: 128].
Estimates of output gains after land redistribution follow a two-step
procedure: first, the average farm size is computed by dividing the available
land by the number of families in the rural labour force; secondly, a
 106 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

statistical estimate of output per acre for that farm size is applied to the
former result. Berry and Cline present [1979: 132-3, Table 5-1] estimates
of the potential gains from such equalising land redistribution, ranging from
ten per cent for Pakistan to 79.5 per cent for north-eastern Brazil.
The only caveats to the above procedure for calculating output gains
given any prominence in Berry and Cline are those arising from price
changes following shifts in output mix, and changes in labour input
intensities [1979: 18-9]. However, the set of assumptions required for this
astonishingly simplistic calculation to hold are both numerous and highly
unlikely to occur in reality.
The first and most obvious problem with this procedure is that the
Downloaded by [New York University] at 07:22 18 October 2014

estimates are not constrained by product mix or land quality. They assume
that the total land available, defined as all land currently being used for
agricultural purposes including pasture and woodland, can be converted to
arable cultivation. Indeed, even marginal and waste land or land unsuitable
for arable purposes can be so converted. Clearly, this assumption is
inadmissible.
Furthermore, none of their calculations take into account the possibility
of there being a minimum efficient scale or of there being a floor
determined by subsistence income. In terms of the inverse relationship
evidence, the optimal size of farm would be around one acre [Sen, 1962].
But when other criteria are considered, relating to viability, we find that the
level of the floor rises considerably. One criterion might be that of minimum
income. That is, the farm should be able to provide an adequate level of
income for a peasant family — a subsistence income. This may turn out to be
significantly larger than the optimal size based on the inverse relationship
phenomenon (output maximisation). Other criteria might involve
employment absorption: that is, the farm should be able to gainfully employ
the working members of the average rural family; or technology absorption:
that is, the minimum farm size should be such as to make efficient use of
draught animals or machines. Consideration of these criteria may
significantly raise the minimum feasible size of holding. Here there is a
clear conflict between various criteria, in particular that based on the yield
postulate of the inverse relationship on one hand, and those based on
subsistence requirements or resource use on the other. Thus, the concept of
viability sets a limit to redistributive land reform.
To draw on the inverse relationship to justify redistributive land reform,
the general equilibrium effects also need to be considered: landless
labourers have normally been excluded from access to redistributed land,
and if land reform is implemented at the expense of large farms, then the
landless may be worse off due to a fall in employment opportunities.
Even if the inverse relationship is empirically valid because small farms
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 107

use more inputs per acre, land redistribution without ensuring the
availability of extra inputs may not produce the expected results [Bardhan,
1973: 1371]. Estimates of output gains after redistribution of land assume
that the required inputs exist and that no losses occur due to the process of
redistribution. Such estimates do not take into account the extra investment
costs of providing irrigation to unirrigated land, or of providing extra inputs
(seeds, fertilisers, pesticides, etc.). The process of land redistribution itself
may involve extra costs in terms of cadastral survey, boundary marking and
the provision of access to plots, as well as the potential for the disruptive
effects of land reform to reduce output.
The estimates further assume that the current input-output
Downloaded by [New York University] at 07:22 18 October 2014

characteristics on existing farms of the 'optimal' size would also

characterise farms of that size after land reform. This requires considering
not only the availability and distribution of production inputs, but also the
whole area of motivation governing labour effort. If incomes were to rise on
small farms, this may relax the subsistence income or debt obligations
constraint and allow small farmers either to relax labour effort with a lower
application of labour per acre, and hence lower output per acre, or to reduce
the proportion of output marketed. In such circumstances, while on-farm
family consumption might increase, the supply of marketed surplus may
well fall with serious consequences for economic activity outside
agriculture.

SOME CONCEPTUAL AND STATISTICAL PROBLEMS

Neither the existence of the inverse relationship, nor its implications, have
gone unchallenged. Indeed, Sen was one of the first to admit that the inverse
relationship was by no means a well established fact and had not been
proven 'beyond the legitimate doubts of exacting statisticians due to the fact
that average data can be misleading' [1964a: 323].
A notable conceptual problem with respect to the inverse relationship
was introduced into the debate by Utsa Patnaik [1972: 1613-24; also 1979
and 1987]. She argues forcefully that, if instead of taking acreage as the
measure of farm size (size grouping) we take either annual value of gross
output per farm or the value of tangible capital stock (scale grouping), we
then may get, in certain circumstances, 'diametrically opposite results'. On
the second measure of size (the scale grouping) output per acre may rise
sharply with increasing scale of production. It may be the case that 'an
intensive application of capital on the smaller holding results in its being
larger [than a holding of higher acreage] on every economic index except
acreage' [1972: 1614]. The reason for this is that holdings of 'varying levels
of intensity, i.e. with varying production techniques and even varying
 108 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

organisation' [1972: 1617] are lumped together within the same size
category, which may blur certain distinctions: between intensively and
extensively cultivated holdings; and between different categories within the
peasantry, so obscuring the extent of differentiation [1972: 1615]. It is
possible to identify, say, a rich peasant with 80 acres and a rich peasant with
just eight acres [1972: 1614].
If all of this is so, then the outcome is as follows, with the inverse
relationship emerging on the size grouping, and certain crucial aspects
concealed, which are clear on the scale grouping:
When grouped by size, a small number of high-productivity holdings
of small size, lumped with similar sized but low productivity
Downloaded by [New York University] at 07:22 18 October 2014

extensively run holdings (which constitute the majority), raises the

average yield for the small size group as a whole; while the large
extensive farms, lumped with similar sized but much more intensive
high-productivity capitalist farms, lower the average yield for the
large-size group as a whole ... Not only with respect to yield, but also
as regards all inputs, capital intensity and labour productivity, we find
the same striking differences in the results of grouping by size and by
scale respectively [1972: 1620].
Patnaik is making a very important point. It is, indeed, the case that 'when
we are studying a process of agricultural change ... it becomes especially
important to analyse the available data according to scale of production'
[1972: 1621]. Two holdings of the same size may well differ in their class
status. A holding may be smaller in size but larger as an economic unit
[1972: 1614]. But note how she formulates the argument:
What we are saying is that in the past, the set of rich peasants and the
set of farms above 20 acres in UP probably had a very large number
of common elements: while at present owing to the changes taking
place in use of techniques and intensification of production the
intersection of the two sets is getting smaller. Therefore the
identification of the properties of one set with those of the other is
now less justifiable and may be downright misleading [1972: 1624].
Observe the difference stressed between the past (a static situation) and
the present (a dynamic one): in the Indian case, between the pre-'new
technology' situation and the period when that 'new technology' had begun
to spread and have an impact. The validity of her argument surely relates to
circumstances of change. One might argue, on her own logic, that in the
relatively static situation of the 1950s, when the inverse relationship was
first established in the FMS studies, size was a useful measure,
encompassing 'sets with a very large number of common elements'; while
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 109

later, following the introduction of the green revolution technology, her

argument begins to apply. She stresses: 'If techniques were absolutely
uniform for all holdings then of course no problem of distinguishing
between size and scale would arise. However techniques are far from
uniform ... The possibility of adopting new, usually much more capital
intensive techniques exists [by the 1970s] for the farms with an investible
surplus' [1972: 1615].
This is the crucial point. We would argue that techniques tended to
uniformity in the 1950s (certainly, to a far greater extent then than in the
1970s); and so size was then a useful stratifying variable: an acceptable
index of economic status (class position). Where techniques of production
Downloaded by [New York University] at 07:22 18 October 2014

have been only partially transformed, farm size may still be the relevant
stratifying variable, but great care must be taken to avoid grouping farms
with different systems of production. The exclusion or separate treatment of
the latter of course requires theoretical and empirical investigation.
One further issue of crucial significance that needs to be flagged at this
point, as it relates closely to the question of agrarian change, is the
breakdown and disappearance of the inverse relationship in the context of
the new technology. This widely documented phenomenon [Chadha, 1978;
Roy, 1979; 1981; Khan, 1979; Dyer, 1991] is discussed later. Suffice it to
state here that abundant evidence from India and Pakistan shows that
following the introduction of green revolution technology, the inverse
relationship seems to have disappeared, especially in those areas where the
new technology has penetrated most deeply.
This, of course, is a very different phenomenon from that described by
Patnaik, whose approach involves a statistical breakdown of the inverse
relationship by switching from an area definition of farm size to a scale
measure. The former approach, however, suggests that it is larger farms, due
to their greater surpluses and access to and control over capital, which
capture the productivity gains from the new technology, thereby structurally
reversing the direction of the size-productivity relationship, despite the
continued intensive application of family labour on small farms.
Another crucial distinction to be made is that between aggregated and
disaggregated levels of analysis. At highly aggregated levels of analysis,
whether at the cross-country [Cornia, 1985], national or regional level, a
relatively high degree of land heterogeneity is to be expected. Those areas
with better than average soil quality, in particular water availability, and
hence higher than average natural land productivity, are historically likely
to have attracted greater population settlement. Higher population density
will, given limited land resources, Jead to small average farm size over long
time periods. Contrariwise for areas of relatively poor agricultural land
where population settlement is likely to be less dense and average farm size
 110 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

larger. Hence, at this 'macro' level, an inverse correlation between land

productivity (as measured by output per acre) and farm size is to be
expected.
Note that this type of inverse relationship at the 'macro' level is
fundamentally different from the inverse relationship which is the focus of
our study. At the aggregate level, the direction of causality runs from land
productivity (itself caused by better soil quality) to small average farm size.
However, at the micro level (village level, for example) the causality is
postulated to run in the opposite direction: specific factors associated with
smaller than average farm size generate higher than average farm
productivity. Moving from the macro to micro scale, the direction of
Downloaded by [New York University] at 07:22 18 October 2014

causation is reversed: while exogenous factors are assumed to cause the

inverse relation at the macro level, endogenous factors are assumed to cause
the same inverse relation at the micro level - the inverse relationship at the
'macro' level is caused by diversity in natural conditions.
There are two further dimensions to the problem of aggregation. There
is the related problem of aggregation over villages leading to a spurious
inverse relationship. Even if no inverse relationship existed within villages,
but there were different soil fertilities between villages, and high fertility
villages had smaller average farm size, then an inverse relation would be
shown by data aggregated over the group of villages. Alternatively, if
fertility (average land productivity) was the same over villages, and the
inverse relationship holds in each village, but villages have different size-
ranges, then such a relationship could be eliminated [Chattopadhyay and
Rudra, 1976: A-110]. Rudra [1968b] implies that the inverse relationship
does not exist within any particular village, but arises as a spurious
relationship when data for different villages are aggregated.
There is also the problem of grouped averages. Most of the analysis of
the FMS data was carried out using size class averages for the principal
variables, farm size and output per acre. Grouped data may generate an
aggregation bias when the in-group variance is in fact greater than the
between-group variance. Thus, farm level data which show no overall
relationship between farm size and output per acre may, when grouped into
size classes, show a spurious inverse relation.
Rudra [1968a: 1041] suspects the process of aggregation is responsible
for what he calls the 'spurious correlation' of the FMS data. In a later study,
Rudra presents data on rank correlation coefficients between farm size and
yield per hectare for individual crops [1968b: A-35, Table 1]: producing
only four statistically significant negative coefficients. Note, however, that
there is a clear distinction to be drawn between physical yields of individual
crops and total crop production in value terms (this reflects the importance
of cropping intensity and crop pattern). Unlike the inverse relationship in
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 111

value terms, little or no evidence has been produced for an inverse

relationship in terms of the physical yields of individual crops.

THE INVERSE RELATIONSHIP VINDICATED AND SOME CONJOINT

RELATIONSHIPS: CLUSTERED FACTORS
While Rudra is correct to assert the need to disaggregate the data, the
divergence between his results and those of the FMS would persist even
with disaggregation. Rudra's concept of yield per acre is biased toward
eliminating the inverse relationship. Yields have been calculated as gross
value of output per gross cultivated area [Rudra, 1968a: 1041]. Gross
Downloaded by [New York University] at 07:22 18 October 2014

cultivated area or gross cropped area includes farm size plus those parts of
the operated area multi-cropped.
However, factors such as double (or multiple) cropping and the
percentage of land cultivated reflect the economics of farming. They are not
exogenous factors. A farmer who cultivates land more intensively via
double cropping and raising the percentage of land cultivated may be argued
to be using land more efficiently [Rao, 1968: 1413]. Rudra's procedure
corrects for the efficiency of land use by using gross cropped area. Moreover,
it is important to establish why land is cultivated more intensively. Rudra's
procedure, however, obscures cropping intensity differentials between farm
size groups which may be of critical significance. If an inverse relationship
exists between cropping intensity (the ratio of gross cropped area to net
cropped area or farm size) and farm size, then Rudra's results are not
surprising. Nevertheless, much of Rudra's criticism of the earlier FMS
studies has a great deal of validity. While Chattopadhyay and Rudra
challenge the universal validity of the inverse relationship, they do not reject
that validity in all circumstances: in some places, at certain times, and for
certain size ranges the inverse relationship may hold [1976: A-104].
Other writers have sought to answer some of the criticisms by using
disaggregated data for individual farm households at the village level. Saini
[1971] and Bhattacharya and Saini [1972: A-63] analyse disaggregated
FMS data for 25 data sets in nine States, concluding: 'Thus, by and large,
the inverse relationship between farm size and productivity is a confirmed
phenomenon in Indian agriculture and its statistical validity is adequately
established by an analysis of the disaggregated data' [1971: A-81-2].
The general conclusion after the spilling of much ink is that the inverse
relationship between farm size and productivity has been confirmed as a
valid empirical phenomenon in India, but not in the way conceived of in the
earlier studies. Rudra and Sen in a joint paper [1980: 393] conclude: 'While
... the inverse relation is more frequently confirmed than rejected, it would
be a mistake to take it to be an empirical generalisation for Indian
 112 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

agriculture as a whole.' The inverse relationship findings in the Indian FMS,

then, are not entirely conclusive. Moreover, some exceptions and
complications are introduced by considering other factors besides land [Sen,
1964b: 441]. But the findings do suggest a persistent phenomenon that
needs to be explained.
A wide range of relationships between farm size and other important
factors are revealed by the various studies mentioned above that may throw
some light on the theoretical understanding of this phenomenon. The major
findings of the FMS studies and the wider debate which followed can be
summarised as follows:
(1) An inverse relationship between output per net cropped acre for the total
Downloaded by [New York University] at 07:22 18 October 2014

value of crop production and farm size exists for most regions of India, and
for a wide range of other countries.
(2) That inverse relationship appears to be weakened when gross cropped
area is used as the land measure in the productivity calculation.
(3) No inverse relationship is evident between the physical yield per acre of
individual crops and farm size. Indeed, in most cases, physical yields of
individual crops appear to be constant or even increasing across farm size.
(4) There is a strong inverse relationship between cropping intensity and
farm size, where the cropping intensity index represents the ratio of gross
cropped area to net cropped area.
(5) A further inverse relationship is evident between units of labour input
per acre and farm size. As farm size increases, less human labour input is
applied per acre.
(6) A related phenomenon is the declining ratio of family labour to total
labour as farm size increases. The ratio of hired labour input to total labour
increases with farm size.
(7) Along with increasing labour input intensity on the smaller farms, is a
higher intensity of application of capital inputs (including animal labour
power, seeds, fertilisers, and farm buildings).
(8) This latter association does not apply to purchased intermediate inputs
which tend to increase proportionately or more than proportionately with
farm size.
(9) An inverse relationship has also been noticed between the percentage of
cultivated area irrigated and farm size.
It is already fairly obvious that these findings are closely interdependent.
All the studies mentioned previously find these relationships clustered,
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 113

suggesting a priori explanations for the inverse relationship based on factor

input intensity, which will be discussed shortly.

THEORETICAL APPROACHES TO THE INVERSE RELATIONSHIP

We have seen then that empirical evidence points to a number of clustered
relationships involving factors of production. That evidence has spawned a
vast theoretical literature attempting to explain why an inverse relationship
arises. It is convenient and logical to examine the principal theoretical
approaches to the inverse relationship under three headings: (1) those that
attempt to explain the inverse relationship in terms of qualitative factor
Downloaded by [New York University] at 07:22 18 October 2014

differences between farm-size categories with respect to management,

labour effort, technique, and soil quality; (2) a second and more substantial
approach that attempts to explain the inverse relationship in terms of
differential factor use intensities between farm-size categories; (3) the third
is a class-based political economy explanation of the inverse relationship.

Management and Labour Quality Hypotheses

Given the essentially untestable nature of the concepts of management and
labour quality, the researcher is forced to utilize various proxies, and hence
this approach tends to be residual in nature. Some form of 'diseconomies of
large scale' peculiar to agricultural production are normally invoked as
giving rise to an inverse relationship.
This argument appears to have two distinct components: the first is a
proposition concerning the increase in complexity of organisation that
comes with size, and which appears to have inefficiencies inherent in it,
compared with the small farm situation; the second is a proposition relating
to supervision problems and incentives which seem to become problematic
as size increases.
Clearly, complexity of organisation does increase with size, but why this
should necessarily generate insuperable inefficiencies is not so clear. On the
contrary, one might argue that the scope provided for division of labour and
specialization leads to the possibility of increased efficiency.
Assuming that this argument has some validity, its force may hinge, to
some extent, upon whether or not technical economies of scale exist. If there
are indeed demonstrable potential economies of scale then these may
compensate for possible managerial diseconomies. This implies a possible
trade-off between technical economies of scale and managerial
diseconomies of scale.
Thus, the argument simply in organisational terms is not wholly
convincing, but the second part of the argument concerning incentives does
seem, at first sight, to have considerable force. This argument suggests that
 114 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

any economies of scale arising from indivisibilities will be offset by the

agency costs of managing wage labour and enforcing effort on the part of
the hired workforce. Supervision costs and incentive contracts, it is posited,
will have profound implications for the optimal size of farm [Mellor and
Johnston, 1984: 558; Mellor, 1966: 368; Nolan, 1988: 41-2].
Related to this argument is the proposition that small farms have a
productivity advantage with respect to the quality of labour, in turn related
to incentive structures. This is a classic populist proposition in the literature,
more often implied than stated explicitly. Unfortunately, no evidence is
presented in support of the proposition, or against which one might test the
hypothesis.
Downloaded by [New York University] at 07:22 18 October 2014

Lipton [1993: 1524] argues:

people will produce more as small family groups, working for
themselves and receiving the whole product of their labour, than as
employees in larger units ... This points to the frequent superiority of
the smaller, self-managed family unit, as expressed in the 'inverse
relationship': namely, the generally higher productivity of land, and to
a lesser extent all factors taken together, farmed in smaller holdings.
Because family members are residual claimants to profits, they therefore
have higher incentives to provide effort than hired labour. They share the
burden of risk and have no search or recruitment costs in the labour market:
thus, the claimed superiority of family over large wage labour-based
operations [Binswanger and Deininger, 1993: 1452]. Due to the advantages
of peer monitoring, agents such as family farmers benefit more from
supervising each other in small groups than from external supervision by
costly foremen. Small farms reduce unit labour-related transaction costs
(search, screening, supervision, shirking) by providing nearby, informed,
rapid and flexible family overview of labour, and by building on intrafamily
altruism and on the extended fungibility of family members between the
household and the family farm [Lipton and Lipton, 1993: 648]. It is further
argued that individual peasant farmers have better knowledge of local
natural conditions than managers of large farms, and that potential losses
from imperfect information are minimised by the ability of the small farmer
to adjust to micro variations in the natural environment [Binswanger and
Deininger, 1993: 1452]. This is accentuated by weather variability which
requires rapid and flexible response by the farmer.
Many of the foregoing arguments, if they have any validity at all, would
appear to have more relevance for vast landed estates or extensive ranch-
type farming involving thousands of acres. They will have much less
relevance for the types of large farms we are considering, in the range from
10 to 50 or even 100 acres.
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 115

Supervision costs, under such conditions, would seem to be greatly

exaggerated. The use of attached labour in a supervisory role, or even better,
the use of family labour, in setting work tasks and ensuring execution is
sufficient to ensure that the requisite labour effort has been expended on the
part of the workforce. The threat of losing access to wage labour
opportunities will provide the incentive for the worker to supply the quantity
and quality of labour effort demanded. The often highly personalised
relationship between employer and labourer, sometimes involving access to
credit and land, help to provide 'nearby, informed, rapid and flexible'
supervision of labour [Bhaduri, 1973].
On the small farm side of the equation, idealised notions of 'family
Downloaded by [New York University] at 07:22 18 October 2014

altruism' and 'fungibility of family members' ignores the unequal

distribution of income within the farm household and intra-household
exploitation. Further, the idea that small farmers have better information on
the local environment is a dubious proposition, and is probably outweighed
by the superior knowledge of large farmers concerning the use of modern
technology.
The really critical problem, however, for these approaches is that if small
farms are indeed characterised by superior management or labour quality,
these should be reflected in a productivity advantage with regard to
individual crops. However, as Bharadwaj [1974a] showed in her study of
the FMS data, no systematic or significant relation between farm-size and
the physical yields of individual crops exists (in fact, irrigated wheat-gram
and cotton in Punjab showed a positive relation). This would seem to refute
the hypothesis that the inverse relation arises from any superior
entrepreneurship of the small farmer.
This approach is also undermined by the fact that the inverse relationship
phenomenon is not limited to a simple comparison between small family
labour farms and large hired labour farms. All the evidence shows that the
inverse relationship runs across small family farms themselves. The adherents
of the above approaches appear not to have recognised this problem.

Land Fertility Hypothesis

A second qualitative approach centres on the proposition that small farms
are located on land of superior productive potential, leading to the fertility-
based explanation of the inverse relation. A problem with this approach is
that it is difficult to distinguish between inherent qualitative land differences
and those arising from the application of fertility-augmenting inputs.
Furthermore, superior soil quality may not be picked up by individual crop
regressions because to some extent crop choice is dictated by soil type, and
quality differences will be reflected predominantly in crop patterns
[Bharadwaj, 1974b: A14-15].
 116 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

Sen [1966] tends towards this fertility based approach. Over time, a
correlation between land fertility and size of holdings will be established via
population expansion on more fertile land. Faster population growth on
more fertile land (due to higher growth of income opportunities) leads to
greater subdivision of the land. This is easy to see in interregional variation
where population expands faster due to natural increase and immigration,
but also within regions, claims Sen, where the ability of a farm household
to withstand famine or crop disease is greater with more fertile land. This
approach ignores the crucial distinction noted earlier between the inverse
relation at the macro level and at the micro level.
Sen's argument regarding fertility differentials relies upon a rather
Downloaded by [New York University] at 07:22 18 October 2014

dubious Malthusian link between income and family size and between level
of income and fragmentation, and ignores alternative employment
opportunities off-farm. An alternative hypothesis proposed by Bhagwati
and Chakravarty [1969] suggests that large farms build up holdings by land
purchase and foreclosure on loans leading to a high degree of
fragmentation, and consequently low productivity.
Of course, neither version of the fertility hypothesis (Sen or Bhagwati
and Chakravarty) holds the promise of economic betterment for small
farmers nor suggests that they have any inherently progressive
characteristics. While the fertility hypothesis based on partible inheritance
or small farm distress sales may appear plausible, it is weakened by the fact
that small farmers may sell land to other small farmers. Likewise, land
reform laws may simply lead to large farmers divesting their worst quality
land. Bharadwaj [1974a] also points out that according to the FMS database,
whereas large holdings do tend to be composed of a greater number of
fragments, the intensity of fragmentation (number of fragments per acre) is
higher on the small farms.
Any discussion of the land fertility-based explanation of the inverse
relation would be incomplete without commenting on the perhaps curious
empirical finding that small farms appear to have a greater percentage of
acreage under irrigation. Many studies have focused on this fact as an
explanation for the inverse relationship [Rao, 1963: 2043; Rao, 1966: 7;
A.P. Rao, 1967: 1990].
Bharadwaj [1974b: A-19] offers two possible explanations for this: (1)
better irrigation leads to greater soil fertility which over time produces
greater land fragmentation; and (2) abundant family labour is deployed to
create and maintain irrigation facilities. However, the first explanation
along the lines of the macro inverse relation, would not explain why within
a given district, small farms have a higher percentage of irrigated area.
Secondly, if such irrigation facilities require capital investment, then the
purported advantage of the small farmer will be counterbalanced. Thirdly,
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 117

the irrigation ratio tells us nothing about the quality of irrigation facilities
on various farm sizes: it does not indicate the effectiveness, source, quality,
controllability or quantum of water supply. Indeed, a large proportion of the
area officially classified as irrigated is often no better than unirrigated land,
depending on rainfall as the source of water, which means that quantitative
comparisons of irrigation ratios can be misleading.

Labour Intensity and Labour Market Dualism: The Sen Model

The extensive list of complementary relationships revealed by the FMS data
tends to suggest that the inverse relationship is associated with variations in
cropping intensity, with the quantum of labour and other production inputs
Downloaded by [New York University] at 07:22 18 October 2014

per acre as the major explanatory factor: what we may call factor intensity
explanations. One influential example is the argument that labour input
intensity explains higher cropping intensity, with much attention, therefore,
directed towards factors which explain labour input intensity.
Sen, in his seminal 1962 article, states that the FMS observations
regarding the inverse relationship are to be expected given what he calls the
'mode of production' of Indian agriculture and its variation over farm size.
The Sen model divides the rural economy into two parts: a modern,
capitalist large farm sector based on hired labour, with the goal of profit
maximisation; and a traditional peasant small farm sector based on family
labour aiming to maximise gross output. While this approach, then, does
hint at the question of determinacy alluded to above, Sen leaves the reasons
for the primary motivation of output maximisation on the peasant farms
unexplained, and concentrates on the causal factors behind higher labour
input intensity.
On large wage labour-based farms labour is hired in up to the point
where the marginal product of labour is equal to the market wage, thus
maximising profits. On small family labour-based farms, 'provided labour
has no outside opportunity of employment and provided there is no
significant disutility of work in the relevant range of effort', labour will be
applied beyond the profit-maximising point until its marginal product is
zero [Sen, 1962: 245].
In a more sophisticated version of the model [Sen, 1966: 440], small
farms maximise utility in a trade-off between increased income from extra
output and leisure. Thus on the family labour-based farm, the marginal
product of labour is not equalised to the market wage, but is determined by
the subjective evaluation of the marginal disutility of effort. This family
labour allocation rule is similar to that advanced by Chayanov [1966] in his
theory of the peasant economy. Unlike Sen, however, Chayanov assumes
the existence of family labour farms to the exclusion of wage labour-based
farms. Nevertheless, they arrive at similar claims as to the relative
 118 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

superiority of the family labour farm. The subjective evaluation of labour is

lower on the peasant farm than on the capitalist farm and this provides the
former with greater resilience (via the ability of the peasant family to
compress income). Thus, the inverse relation, according to Sen, is the
natural result of an economy characterised by the existence of widespread
surplus labour and family-based non-wage cultivation. The crucial factor is
not size as such, but the system of fanning [1962: 246].
A number of writers have criticised the Sen approach as being
analytically deficient given that even small farms do not rely on family
labour exclusively, and a large number of small farms are engaged in off-
farm income generating activities. It is pointed out by these authors that the
Downloaded by [New York University] at 07:22 18 October 2014

Sen model requires the non-existence of off-farm employment opportunities

and the absence of labour-hiring on the small farms [Khusro, 1964; Bhagwati
andChakravarty, 1969;Rudra, l913\Taslim, 1989;Mazumdar, 1963:1259].
Of course, the fact that small farms hire in labour does not imply that
they should follow marginalist rules. Time constraints may necessitate the
hire of labour for urgent tasks to avoid harvest failure, which is quite
consistent with under-employment throughout the rest of the year. Further,
certain operations needing specialised skills may require hired labour which
cannot therefore be seen as a substitutive category, for example the hire and
operation of certain equipment. Furthermore, there may be sociological
factors behind labour market imperfections: barriers to employment of
women and children on account of status, or reluctance to work outside the
farm [Bardhan, 1973: 1380]. Family labour and off-farm labour are not
always coterminous. Social convention may determine the allocation of
tasks, dependent on class, caste and gender status.
Neither hypothesis (that family and hired labour are exclusively separate
categories so that family labour becomes a datum to the cultivating
household, or that they are perfect substitutes so that the wage rate measures
opportunity cost) is then justified. The wage rate is only one determinant of
labour use and possibly not the most important factor [Bharadwaj, 1974b:
A-18].
A second major criticism directed at the Sen model is an extension of the
above argument. Clearly, farms are neither exclusively family labour-based
nor wage labour-based, but this can be taken further: examination of the
FMS data shows that the proportion of exclusively family labour farms is
negligible among large farms (and not even very high among small farms)
so that the labour allocation pattern posited by Sen on such farms cannot be
used as a firm basis for an explanation of the inverse relation over the entire
range of farm size. Hired labour is necessary beyond ten acres. Once 15
acres is reached, the fixed quantity of family labour wears thin (per acre use
of family labour is negligible both on 15 acre and 50 acre farms, with both
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 119

depending on hired labour). So the use of family labour cannot be used as

an explanation of higher productivity on 15 acre as against 50 acre farms.
More explicitly neoclassical models exist which posit different sets of
factor prices facing small and large farms as an explanation for the inverse
relationship [Srinivasan, 1973; Bardhan, 1973; Griffin, 1974; Bhalla, 1979;
Berry and Cline, 1979]. Here, the explanation involves optimal use of factor
inputs in the context of relative scarcity: factor prices differ between large
and small farms, such that the effective prices of land and capital are low for
the former and the effective price of labour is low for the latter. The
presence of relatively abundant family labour on small farms and the
relatively low implicit price of land for large farmers dictate choices of
Downloaded by [New York University] at 07:22 18 October 2014

technique with different factor intensities. Thus small farms have high
labour/land ratios, whereas large farms use labour and land less intensively.
Small farms with a lower opportunity cost of labour can exploit more
marginal land, cultivate a larger proportion of their land, and achieve higher
yields.
The thesis of differential factor prices between large and small farms is
highly problematic. If indeed factor price differentials were the main
explanatory mechanism at work, then we would expect to find higher
capital intensities on large farms manifested in technological innovation,
both biochemical and mechanical. However, as we will see later, in this
context the inverse relationship breaks down. Indeed, this hypothesis would
appear to be more appropriate as an explanation for the non-existence of an
inverse relationship rather than its cause. Further, the supposed ability of
small farmers to exploit more marginal land (because of a lower opportunity
cost of labour) is hardly conducive to higher crop yields.

BEYOND THE MARGINALIST APPROACH: THE DEVELOPMENT OF

A CLASS-THEORETIC FRAMEWORK
One common element between the Sen model and these more explicitly
pricist variants is that they share the same marginalist conceptions and
categories. The critical problems arise from the production function
methodology employed. The only difference between Sen's model and the
factor market imperfection theories mentioned above is that the latter
consider all relative factor prices whereas Sen considers only one price: the
cost of labour, which varies according to the system of production.
An important point of criticism of both the Sen model and the
neoclassical variants concerns the simplifying assumption of a single crop
and a single production cycle, which misses entirely the crucial significance
of cropping intensity and cropping pattern. Thus, we see that both models
are geared to explaining a non-existent relationship: while the inverse
 120 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

relationship can be observed in value terms, no such relationship exists with

respect to the physical yields of individual crops. Of course, the models can
be respecified to look at total value of crop output and labour input intensity,
with a strong relationship between cropping intensity and labour
application.
Production function studies, however, suffer from the problem that they
are estimated with flow measures, which obscures the role of indivisibilities
in agricultural production [Sen, 1981: 204]. The whole exercise of
comparing marginal product and wage rates is meaningless for agricultural
production, which is very different from industrial production. Agricultural
production is sequential in nature, subdivided both temporally and
Downloaded by [New York University] at 07:22 18 October 2014

operationally. The production cycle in agriculture is prolonged and varies

greatly between crops. No straightforward marginal productivity rules can
be applied here as the productivity of family labour inputs prior to harvest
time will depend on the timely application of the necessary amount of
labour during the harvesting period. Labour demand at the harvesting stage
is largely determined by elements of previous stages. Labour at earlier
stages thus has to be seen in the nature of fixed costs. Labour demand at
harvest is a derived demand dependent on conditions earlier in the
production cycle. Under such conditions, the meaning of the marginal
product of labour in agriculture becomes extremely hazy almost to the point
of becoming operationally useless [Roy, 1979: 20-23].
A second common element is that both approaches assume the same
production function for all farm sizes and systems of production. Both
peasant and capitalist farms are assumed to be operating on the same
production function. Sen focuses on two different systems of production
(small peasant farms and large capitalist farms), which have fundamental
differences in their organisation of production and qualitative differences in
their division of labour. It is not obvious therefore why these two systems
are located on the same production function (the same marginal product of
labour curve).
Indeed, Sen himself was later to admit: 'it is illegitimate to eulogize
peasant farming on the basis of an analysis in which every type of farm has
access to the same production function and to the same factors of
production' [1966: 444]. A peasant farmer may be constricted to a less
efficient set of production conditions including lack of access to economies
of scale, lack of technical knowledge or access to particular factors, or risk
aversion to using new inputs.
The use of a single production function to compare two very different
systems of farming, points to the identification problem at the heart of the
Sen model and its disciples. This concerns the assumption that large farms
are capitalist farms, and small farms are peasant family labour-based farms.
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 121

This conceptual sleight of hand obscures the real relations at work which
produce the inverse relation. This crucial point is further discussed below
when we turn to a class-based approach to analysing the inverse relation.
We must go deeper than the size of holding categories to the underlying
social relations of production, in the dynamic context of agrarian transition.
Three possible explanations in terms of factor intensity have been
posited: (1) labour input, (2) cropping intensity, and (3) choice of crop mix.
A key point in the debate over the causal factors behind the inverse
relationship, but one that is more often ignored or obscured, is which of
these explanations is determinant? These findings are clearly inter-related,
but what is the direction of causality? No simple econometric analysis can
Downloaded by [New York University] at 07:22 18 October 2014

tell us the answer.

There are two possible interpretations however. The first, labour-based
approach, we have already surveyed critically. The second is that it is higher
cropping intensity and a cropping pattern associated with either higher labour
absorption or remuneration (and hence the higher income derived from
production) which implies greater factor use intensity, in particular labour
intensity. Bharadwaj [1974b: A-16], indeed, suggests that value productivity
differentials between farms boil down to the differential cropping intensity
components of crop patterns. Here, the primary motivation is output
maximisation, with cropping intensity appearing as the proximate cause of the
inverse relation. Higher cropping intensities imply a higher intensity in the
application of other factors of production, particularly labour.
Bharadwaj [1974b: A—16] notes from the FMS evidence that while there
is a strong inverse relationship between labour input and size, this is not
manifested in terms of individual crops. An explanation in terms of efficient
factor substitution is undermined by the fact that there is no such relationship
between labour productivity and size to be found in the FMS data.
Chattopadhyay and Rudra [1976: A-114], although taking a rather
eclectic view of the inverse relationship, make the following powerful and
incisive statement which takes us beyond the confines of the marginalist
approach:
Among the forces that drive a small farmer to more intensive effort the
most important one, of course, is his need for survival. There is a certain
basic minimum of consumption that a poor peasant family has to have
without which it will be simply wiped out. It is only understandable that
such a poor peasant family, depending on a small piece of land,
submerged in a vast population of surplus labour in the countryside, and
thus not having any alternative sources of employment and income,
would try to produce the maximum output on his piece of land. He
would not only ignore any marginal productivity calculations insofar as
 122 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

family labour is concerned, he would employ hired labour whenever

necessary to supplement family labour, and in doing that would pay no
heed to marginal productivities. He would also try to apply non-labour
and non-monetised inputs with maximum intensity, once again by using
labour without any calculations. He would try to improve the quality of
land by small-scale irrigation and other such means as can be procured
with the help of labour. He will tend to leave fallow as little land as
possible, and try to cultivate as many crops as possible and choose such
crops which after meeting his minimum consumption needs would meet
his minimum cash needs ... the tendency to intensify his effort would
be all the more so [under sharecrop'ping] because of the fact that he has
Downloaded by [New York University] at 07:22 18 October 2014

now to meet his minimum needs with only a share of the results of his
effort.
Despite the criticisms of the methodology and assumptions employed in
Sen's labour utilisation model and the neoclassical 'pricist' variants, the
former does have the great merit of attempting to locate an explanation of
the inverse relationship in terms of the different conditions of production
facing farm households. These Sen explores within the logic of what he
calls the 'mode of production of Indian agriculture' and its correlation with
farm size [1962: 245]. However, his notion of mode of production is
conceptually nebulous and constrained by its choice theoretic framework
based on relative resource endowments. There is clearly a need to go
beyond farm size as the relevant stratifying variable to examine the
underlying relations and forces of production.
The class-based approach proceeds from the proposition that the peasant
farm is embedded in the socio-economic context of an emerging capitalist
agriculture in which however, non-capitalist forms of surplus appropriation
are still prevalent. Such a transitional state has been described by Bhaduri
[1973] and Bharadwaj [1974b] as one of semi-feudalism, a situation in
which the relations of production have more in common with feudalism
than capitalism.
Agriculture remains backward inasmuch as 'the process of
commercialisation has not culminated necessarily or rapidly in the
pervasive dominance of capitalist relations' [Bharadwaj, 1985: 9]. The
process of commercialisation has intensified peasant differentiation but has
not generated a qualitatively changed peasantry: it has not resulted in a
fully-formed capitalist agriculture in which rich peasants are transformed
into capitalist farmers and poor peasants into wage labourers. Neither
commercialisation nor the development of wage labour, however, are
sufficient conditions for the development of capitalist agriculture.
This directly addresses Sen's and other writers' erroneous identification,
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 123

on the basis of the labour-hiring criterion alone, of the large farms in their
samples with capitalist farms. That identification needs an altogether more
complex specification. The transition from rich peasant to capitalist farmer
cannot simply be assumed (as it is generally in many studies of the inverse
relationship) - it requires demonstration. While the use of wage labour and
market participation may be important indications of processes of agrarian
transition, the evidence contained in Bhaduri, Bharadwaj, Patnaik and others,
of the lack of capital intensification on large farms, and of the continued
existence of unfree relations between labour hirers and workers would suggest
that Sen's (and others') elision of rich peasant/capitalist farmer is invalid.
It is within the context of a backward agriculture that we must seek the
Downloaded by [New York University] at 07:22 18 October 2014

factors that give rise to and sustain the inverse relationship. Within a semi-
feudal agriculture, the dominant classes are able to intensify and maximise
surplus extraction from subordinate classes via interlocking modes of
exploitation. By interlinking transactions in more than one market, the
dominant class can maximise the rate of exploitation over time.
Here then is a very powerful mechanism, rooted in the social relations
of production of an essentially pre-capitalist mode of production, that drives
poor peasants to maximise output because their survival depends upon it. It
would appear therefore that the factors driving poor peasants to intensify
labour effort are more important than the factors permitting them to do so.
The inverse relationship cannot be understood in terms of scale advantages
among isolated farms, or simply in terms of the poverty and unemployment
facing poor peasants. The inverse relationship arises because of factors
which are related to farm size, but not because of some independent size
effect per se. It is thus misguided to treat the inverse relationship as a sign
of relative efficiency rather than of distress. Chattopadhyay and Rudra
[1976: A-115] conclude: 'if the inverse relationship be made the basis of a
policy for preserving small farmers as they are, the result would be the
destitution and expropriation of poor peasants ... '. At the other end of the
class spectrum, rich peasants, who have not yet been transformed into
capitalist farmers, use the same technology, but with much lower labour
intensity, and thus achieve lower yields.

THE BREAKDOWN OF THE INVERSE RELATIONSHIP IN THE

DYNAMIC CONTEXT
We turn now to a further set of considerations that need to be taken into
account: the essentially static nature of the Sen approach. The extension of a
static result from a given set of unchanging circumstances to a dynamic
context is invalid. The most significant change is likely to be with respect to
technology.
 124 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

In those areas where we find an inverse relationship, it is the case that

all farm sizes have access to a more or less similar technology. Large and
small farms use the same set of production inputs, and small farms achieve
higher output per acre via far higher cropping intensities which in turn
imply higher application of labour effort per acre. With the introduction of
a new technology which favours large farms due to its associated economies
of scale, then the so-called advantage which the small farms have with
respect to labour input intensity, may be matched or more than matched by
the new advantages which large farms have with respect to technology. In
this situation, we might expect the inverse relationship to break down and
eventually disappear, and with it part of the case for redistributive land
Downloaded by [New York University] at 07:22 18 October 2014

reform.
In terms of the class-based approach, which is essentially a dynamic
theory, with the development of the forces of production in the form of
green revolution technology rich peasants are better placed to reap the
benefits. When the new technology is transferred from laboratory test beds
and research stations to agrarian production, large farmers benefit
disproportionately. Institutions and services (extension, credit, input/output
prices, information, marketing) and political power exhibit a strong bias in
favour of large fanners who are the early adopters and therefore early
gainers in terms of high output prices and low (subsidised) input prices
[Byres, 1972].
Rich peasants have the resources, which poor peasants lack, to gain
access to the package of new inputs (HYV seeds, chemical fertilisers, plant
protection materials), and are able to monopolise the available credit
necessary to purchase the green revolution package. Rich peasants dominate
the institutions which supply and distribute the new inputs. While these
biochemical inputs may be scale neutral, their adoption steps up pressures
for mechanisation which has associated scale economies. This leads to the
breakdown of the inverse relation with large capitalist farmers achieving
higher yields.
The empirical evidence supports the proposition that following the
introduction of the green revolution technology, the inverse relationship
breaks down, markedly so in those areas where the new technology has
penetrated most deeply: Chadha [1978] and Roy [1979] examine evidence
from the Punjab to show a significant transition to modern capitalist
agriculture. They divide Punjab into three heterogenous zones: the
relatively backward eastern Punjab, the more recently advanced central
districts, and the advanced western Punjab. The district regressions of yield
on farm size show an inverse relationship still significant in the eastern
zone, insignificant for the central zone, with the western Punjab exhibiting
a statistically significant positive relation between output per acre and net
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 125

cropped area. The fact that the inverse relationship is no longer in evidence
in the latter region despite the still significant inverse relationship between
cropping intensity and farm size suggests that large fanners have attained
important scale advantages.
The hypothesis supported by Roy [1979; 1981] is that during the early
period of transition, the institutional bias in favour of large farmers ensures
that they are the first adopters of the new green revolution technology. Scale
advantages become more important in the post adoption period, with large
farmers maintaining high investment and growth rates. Thus, rather than
small farmers being able to catch up, the initial advantages captured by large
farmers are further strengthened over time due to intrinsic scale advantages.
Downloaded by [New York University] at 07:22 18 October 2014

THE INVERSE RELATIONSHIP IN THE EGYPTIAN CONTEXT

In an earlier contribution [Dyer, 1991], I presented a preliminary analysis of
my fieldwork on the inverse relationship in the context of Egyptian
agriculture. This work and the subsequent Ph.D. thesis [Dyer, 1995] were
directly inspired by the class analytical framework suggested by T.J. Byres.
Space here prevents me presenting the results in any great detail, but the
broad outlines of my findings are as follows.
First, I showed that the evolution of cross-class land productivity in
Egyptian agriculture was regionally heterogeneous [Dyer, 1989; 1991]. In a
disaggregated study of ILO data for 18 villages in 1976 [Radwan and Lee,
1986], I was able to show that while some villages continued to exhibit an
inverse relationship, others seemed to have experienced a breakdown and
disappearance of that relationship. There was a high degree of correlation
between the latter villages and areas of the Egyptian countryside which had
experienced the development of capitalist relations and forces of
production.
Later, I was able to identify in much greater detail the mechanisms
adduced by political economy that lay behind the evolution of cross-class
productivity indicated [Dyer, 1995; 1996]. At the heart of the analysis were
processes of peasant differentiation and the important role played by the
Egyptian state in the institutional development of the agrarian sector, in
particular the relationship between rich peasants and the co-operative
system.
After 1952, all land reform recipients had to join the co-operative
system, at first confined to land reform areas. By the 1980s, there were over
5,000 agricultural co-operatives covering approximately three million farm
families [Rochin and Grossman, 1985]. Rich peasants were able to
dominate the co-operatives, especially those set up in non-land reform areas
after 1963. The Egyptian state simply did not have the cadre to carry out
 126 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

such a massive extension without relying extensively on the local dominant

classes [Richards, 1982]. Even where poor peasants were represented on co-
operative boards, they were highly vulnerable to rich peasant pressure and
bribery. Poor peasants are dependent on rich peasants for land, labour, input
purchases, crop sales, cash loans, and intercession with state authorities
[Adams, 1986]. These ties of dependence mean that the bulk of poor
peasants have been 'captured' - not by the state - but by the rich peasants.
Rich peasants were able to use their direct or indirect control of co-
operative boards to help themselves to co-operative supplies and
monopolise tractors and other mechanised inputs, excluding the poor as
systematically as a price system in an environment of unequal resource
Downloaded by [New York University] at 07:22 18 October 2014

endowments would have done. Similar efforts characterised the subsidised

credit programmes as well: most peasants could not gain access to medium
and long term loans under the borrowing criteria set. The critical investment
loans for livestock and machinery remained predicated on property,
producing a large farm bias in the provision of credit.
By advancing loans to all farmers on the basis of the size of their
cultivated crop, the Egyptian state clearly engaged in the unequal
subsidisation of large farms. Those with less than five feddans, the vast
majority of farms, received only half the credit advanced by the co-
operative system, while the small minority of farms with over five feddans
collected some 44 per cent of the total [Sadowski, 1991].
This rich peasant bias in access to co-operative and village bank credit
resources also has important implications for the pattern of diffusion of
modern technology in Egyptian agriculture. On the basis of their
institutional control, large farms were able to obtain inputs before other
farm households, the timely supply of such inputs constituting a major
bottleneck and a condition of high productivity. Ownership of tractors, and
other modern farm equipment such as irrigation technology, is strongly
correlated with farm size, as Abdel-Fadil [1995] shows, with tractors being
too large and indivisible an investment for poor and middle peasants.
Tractor loans require collateral of at least five feddans and a 25 per cent
down payment, and water pumps a collateral of three feddans. However, use
of tractors via the private rental market was more widespread: in the 1981
agricultural census for Qena govemorate in Upper Egypt, while only 2,500
farms owned or shared tractors, some 2,500 farms rented tractors from the
public sector, and over 98,000 farms (66 per cent) rented tractor time from
the private sector [Ministry of Agriculture, 1985].
As Abdel-Fadil [1995] notes, machine rental has become increasingly
important as a mode of surplus extraction: those who own machinery rent
to others on a piecemeal basis in return for cash or crop share. The
profitability of such operations is enhanced by subsidised fuel, oil and
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 127

machine purchase. Further, rental prices are set fairly high through tacit
collusion between machine owners. Over the last 15 years or so, the
emphasis has been on machine accumulation rather than land accumulation,
reinforcing and accentuating the distinction between large and small farms.
Such mechanisation reinforces the power and position of the rich peasants
and the choice of technology reflects their perceived interests. In the past,
reciprocal labour exchange between small peasant farms solved labour
availability problems at peak periods, but mechanisation and monetisation
have produced a pattern of hired labour with a high correlation between
degree of mechanisation and hired labour use [Hopkins et ah, 1982]. The
benefits of tractorisation can be seen in terms of increased cropping
Downloaded by [New York University] at 07:22 18 October 2014

intensity and yields, providing more flexibility in the cropping pattern,

timeliness, and cost savings. In particular, mechanisation has increased the
cropping intensity of rich peasants. The pattern of improved yields
corresponds to the pattern of highly mechanised villages. Of course, what is
important are the factors which generate differences in mechanisation levels
in the first place [El Sahrigi and Shepley, 1984; 1985].
None the less, the diffusion of mechanisation throughout Egypt is very
uneven. Levels of mechanisation vary considerably within and between
villages according to the degree of social differentiation, and the unevenly
distributed power of the rich peasantry in the Egyptian countryside, itself
unevenly distributed. This heterogeneity has had profound implications for
agrarian transition in Egypt. Dyer [1995; 1996] delineates two paths of
agrarian transition from semi-feudalism in Egypt: a capitalist path (using
hired wage labour and machinery) and a path dominated by petty
commodity production with small farms producing for the market. In these
studies, I provide a typology of villages reflecting these different paths of
agrarian transition in Egypt: (1) land reform or resettlement villages which
have moved in the direction of intensification of petty commodity
production; and (2) villages in which capitalist agriculture and the
emergence of capitalist relations of production around wage labour have
appeared (see also Hopkins [1987] who identifies four possible paths for the
Egyptian case).
Land reform villages are dominated by family farms of three to five
feddans dependent on the use of intensive family labour. The effect of land
reform here is to maintain the 'traditional' family farm as the unit of
production and link it to the market through co-operatives. Larger farmers
are not qualitatively different from smaller farmers, just quantitatively
differentiated. These villages, characterised by small scattered land
possessions, prohibit the application of modern technology and lead to the
fragile formation of capitalism. Land fragmentation weakens their ability to
adopt new agricultural methods, and intensify capital utilisation. Thus, land
 128 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

reform had the effect of slowing down capital accumulation, dampening the
rate at which farm units are able to save and invest.
Direct state influence was certainly much stronger in land reform areas
than elsewhere, where co-operatives were more easily dominated by rich
peasants. Villages in which rich peasants have been able to accumulate land
and other means of production such as machinery have exhibited a different
outcome: increased social differentiation and the potential disappearance of
small farmers rather than their survival. The capitalist villages have
relatively greater concentration of land, larger area and population, and
higher levels of mechanisation. Such villages also benefit from
development efforts in the form of loans for machinery and other modern
Downloaded by [New York University] at 07:22 18 October 2014

inputs. The mode of production changes because capital has penetrated the
village and changes the system of production instead of being merely
externally imposed via market relations. The family farm gives way to
larger enterprises based on wage labour and the intensive use of machine
inputs.
Within the uneven development of the Egyptian countryside, rich
peasants are in many cases not 'technically more efficient' based on their
use of modern technology. On the contrary, in those areas of relatively
backward agriculture, rich peasants use essentially the same traditional
technology as the poor and middle peasants, thus generating the
circumstances under which an inverse relationship arises: with all farm
sizes using more or less the same techniques of production, the higher
cropping intensities of poor peasants (less than three feddans) which imply
higher labour use intensities and higher value yields, have generated an
inverse relationship between holding size and land productivity.
Only in areas where rich peasants have transformed themselves into a
class of capitalist farmers are they more efficient, producing the conditions
for a breakdown in the inverse relationship. In the capitalist regions, the
significantly advanced level of the productive forces has allowed rich
peasants to reap important scale advantages which more than outweigh the
still higher labour and cropping intensities of poor peasant farms. The
development of the forces of production here includes significantly different
cropping patterns as well as use of modern inputs.
Thus, in the dynamic context of the development of both the relations
and forces of production, in the shape of new technology, the inverse
relationship breaks down and disappears. Rich peasants are able to capture
the gains from the new technology, and with increased accumulation
develop into capitalist farmers [Dyer, 1991]. These findings would seem to
lay to rest the possibility of employing the inverse relationship evidence in
favour of redistributive land reform.
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 129

REFERENCES

Abdel-Fadil, M., 1995, 'Class Differentiation and the Evolution of Agrarian Relations in Rural
Egypt in the Liberal Era (1975-90)', Paper presented at the International Conference on
Agricultural Productivity and Economic Development, Trondheim University, June 1995.
Adams, R.H., 1986, Development and Social Change in Rural Egypt, New York: Syracuse
University Press.
Bardhan, P.K., 1973, 'Size, Productivity, and Returns to Scale: An Analysis of Farm-Level Data
in Indian Agriculture', Journal of Political Economy, Vol.18, Dec.
Berry, R.A. and W.R. Cline, 1979, Agrarian Structure and Productivity in Developing Countries,
Baltimore, MD: Johns Hopkins University Press.
Bhaduri, A., 1973, 'A Study in Agricultural Backwardness Under Semi-Feudalism', Economic
Journal, Vol.83, March.
Bhagwati, J.N. and S.J. Chakravarty, 1969, 'Contributions to Indian Economic Analysis: A
Downloaded by [New York University] at 07:22 18 October 2014

Survey', American Economic Review, Vol.59 No.4, Part 2, Supplement.

Bhalla, S.S., 1979, 'Farm Size, Productivity and Technical Change in Indian Agriculture', in
Berry and Cline [1979].
Bharadwaj, K., 1974a, Production Conditions in Indian Agriculture, Cambridge: Cambridge
University Press.
Bharadwaj, K., 1974b, 'Notes on Farm Size and Productivity', Economic and Political Weekly,
Review of Agriculture, Vol.9, No.13.
Bharadwaj, K., 1985, 'A View of Commercialisation in Indian Agriculture and the Development
of Capitalism', Journal of Peasant Studies, Vol.12, No.4.
Bhattacharya, N. and G.R. Saini, 1972, 'Farm Size and Productivity - A Fresh Look', Economic
and Political Weekly, Review of Agriculture, Vol.7, No.26.
Binswanger, H.P. and K. Deininger, 1993, 'South African Land Policy: The Legacy of History
and Current Options', World Development, Vol.21, No.9.
Byres, T.J., 1972, 'The Dialectic of India's Green Revolution', South Asian Review, Vol.5, No.2.
Byres, T.J., 1974, 'Land Reform, Industrialisation and the Marketed Surplus in India: An Essay
on the Power of Rural Bias', in David Lehmann (ed.), Agrarian Reform and Agrarian
Reformism, London: Faber & Faber.
Byres, T.J., 1977a, 'Output Per Acre and Size of Holding: The Logic of Peasant Agriculture
Under Semi-Feudalism'. Paper presented at the IDS Rupag Seminar Programme, and in
modified form at ISS, The Hague, April 1979.
Byres, T.J., 1977b, 'Agrarian Transition and the Agrarian Question', Journal of Peasant Studies,
Vol.4 No.3.
Byres, T.J., 1981, 'The New Technology, Class Formation and Class Action in the Indian
Countryside', Journal of Peasant Studies, Vol.8, No.4.
Byres, T.J., 1986a, 'The Agrarian Question and Differentiation of the Peasantry', Foreword to
Atiur Rahman, Peasants and Classes. A Study in Differentiation in Bangladesh, London and
New Jersey: Zed Books.
Byres, T.J., 1986b, 'The Agrarian Question, Forms of Capitalist Agrarian Transition and the
State: An Essay with Reference to Asia', Social Scientist, Vol.14, Nos.11-12.
Byres, T.J., 1990, 'The Agrarian Question and Differing Forms of Capitalist Agrarian Transition:
An Essay with Reference to Asia', in Jan Breman and Sudipto Mundle (eds.), Rural
Transformation in Asia, New Delhi and Oxford: Oxford University Press.
Byres, T.J., 1991, "Agrarian Question, The" and "Peasantry", in Tom Bottomore et al. (eds.), A
Dictionary of Marxist Thought (2nd edition), Oxford: Blackwell.
Byres, T.J. and G. Dyer, (forthcoming), The Logic of Peasant Agriculture Under Semi-Feudalism.
Chadha, G.K., 1978, 'Farm Size and Productivity Revisited - Some Notes from Recent
Experience of Punjab', Economic and Political Weekly, Review of Agriculture, Vol.13, No.9.
Chattopadhyay, M. and A. Rudra, 1976, 'Size-Productivity Revisited', Economic and Political
Weekly, Review of Agriculture, Vol.11, No.39.
Chayanov, A.V., 1966, The Theory of Peasant Economy, Homewood, IL: The American
Economic Association.
Cornia, G.A., 1985, 'Farm Size, Land Yields and the Agricultural Production Function: An
 130 AGRARIAN QUESTIONS: ESSAYS IN APPRECIATION OF T.J. BYRES

Analysis for Fifteen Developing Countries', World Development, Vol.13, No.4.

Dyer, G.D., 1989, 'Agrarian Transition in Egypt: Technological Change in Egyptian Agriculture
and Its Impact on the Relation between Farm Size and Productivity', unpublished M.Phil.
thesis, University of Cambridge.
Dyer, G.D., 1991, 'Farm Size - Farm Productivity Re-examined: Evidence from Rural Egypt',
Journal of Peasant Studies, Vol.19, No.1.
Dyer, G.D., 1995, 'Farm Size and Productivity in Egyptian Agriculture: An Analysis of Agrarian
Structure and Technical Change', unpublished Ph.D. thesis, London University.
Dyer, G.D., 1996, Class, State and Agricultural Productivity in Egypt, London: Frank Cass & Co.
El Sahrigi, A.F. and S.C. Shepley, 1984, 'Farm Equipment Manufacture in Egypt', Egypt
Agricultural Mechanization Project, Cairo: ARE Ministry of Agriculture.
El Sahrigi, A.F. and S.C. Shepley, 1985, 'Socio-Economic Evaluation of Farm Machinery
Introduced in Project Villages from 1980 to December 1984', Egypt Agricultural
Mechanization Project, Working Paper No.19, Cairo: ARE/USAID.
Downloaded by [New York University] at 07:22 18 October 2014

Ghose, A.K., 1979, 'Farm Size and Land Productivity in Indian Agriculture: A Reappraisal',
Journal of Development Studies, Vol.16, No.1.
Griffin, K., 1974, The Political Economy of Agrarian Change, London: Macmillan.
Hopkins, N.S., 1987, Agrarian Transformation in Egypt, Cairo: AUC Press.
Hopkins, N.S., Mehanna, S.R. and B.M. Abdel Maksoud, 1982, 'The State of Agricultural
Mechanization in Egypt: Results of a Survey 1982', Agricultural Mechanization Project,
Cairo: ARE Ministry of Agriculture/USAID.
Khan, M.H., 1979, 'Farm Size and Land Productivity Relationships in Pakistan', Pakistan
Development Review, Vol.18, No.l.
Khusro, A.M., 1964, 'Returns to Scale in Indian Agriculture', Indian Journal of Agricultural
Economics, Vol.19, Nos.3-4.
Lipton, M., 1993, 'Land Reform as Commenced Business: The Evidence Against Stopping',
World Development, Vol.21, No.4.
Lipton, M. and M. Lipton, 1993, 'Creating Rural Livelihoods: Some Lessons for South Africa
from Experience Elsewhere', World Development, Vol.21, No.9.
Mazumdar, D., 1963, 'On the Economics of Relative Efficiency of Small Farmers', Economic
Weekly, Vol.15, Nos.28-30.
Mellor, J.W., 1966, The Economics of Agricultural Development, Ithaca, NY: Cornell University
Press.
Mellor, J.W. and B.F. Johnston, 1984, 'The World Food Equation: Interrelations among
Development, Employment and Food Consumption', Journal of Economic Literature,
Vol.22, No.2.
Ministry of Agriculture, 1985, Nita'ig al-Ta'adad al-Zira'ai 'an al-Sinah al-Zira'aiah 1981-82,
Muhafidhah Qena (Agricultural Census 1981-2, Qena Governorate), Cairo.
Nolan, P., 1988, The Political Economy of Collective Farms, Cambridge: Basil Blackwell.
Patnaik, U., 1972, 'Economics of Farm Size and Farm Scale: Some Assumptions Re-Examined',
Economic and Political Weekly, Vol.7, Nos.31-33.
Patnaik, U., 1979, 'Neo-Populism and Marxism: The Chayanovian View of the Agrarian
Question and Its Fundamental Fallacy', Journal of Peasant Studies, Vol.6, No.4.
Patnaik, U., 1987, Peasant Class Differentiation. A Study in Method with Reference to Haryana,
Delhi: Oxford University Press.
Radwan, S. and E. Lee, 1986, Agrarian Change in Egypt: An Anatomy of Rural Poverty, London:
Croom Helm.
Rao, A.P., 1967, 'Size of Holding and Productivity', Economic and Political Weekly, Vol.2,
No.44.
Rao, C.H.H., 1963, 'Farm Size and the Economics of Scale', Economic Weekly, Vol.15, No.51.
Rao, C.H.H., 1966, 'Alternative Explanations of the Inverse Relationship between Farm Size and
Output Per Acre in India', Indian Economic Review, Vol.1, No.2.
Rao, C.H.H., 1968, 'Farm Size and Yield Per Acre - A Comment', Economic and Political
Weekly, Vol.3, No.37.
Richards, A.R., 1982, Egypt's Agricultural Development, 1800-1980, Boulder, CO: Westview
Press.
 THE LOGIC OF PEASANT AGRICULTURE UNDER SEMI-FEUDALISM 131

Rochin, R.I. and J.C. Grossman, 1985, 'Agricultural Cooperatives and Government Control in
Egypt: A Historical and Statistical Assessment', Department of Agricultural Economics,
University of California.
Roy, P.L., 1979, 'The Relation between Farm Size and Productivity in the Context of Alternative
Modes of Production in Indian Agriculture', Ph.D. thesis, Delhi School of Economics,
University of Delhi.
Roy, P.L., 1981, 'Transition in Agriculture: Empirical Indicators and Results (Evidence from
Punjab, India)', Journal of Peasant Studies, Vol.8, No.2.
Rudra, A., 1968a, 'Farm Size and Yield Per Acre', Economic and Political Weekly, Vol.3,
Nos.26-8.
Rudra, A., 1968b, 'More on Returns to Scale in Indian Agriculture', Economic and Political
Weekly, Review of Agriculture, Vol.3, No.43.
Rudra, A., 1973, 'Allocative Efficiency of Indian Farmers: Some Methodological Problems',
Economic and Political Weekly, Vol.8, No.3.
Downloaded by [New York University] at 07:22 18 October 2014

Rudra, A. and A.K. Sen, 1980, 'Farm Size and Labour Use: Analysis and Policy', Economic and
Political Weekly, Vol.15, Nos.4-6.
Sadowski, Y.M., 1991, Political Vegetables? Businessman and Bureaucrat in the Development of
Egyptian Agriculture, Washington DC: The Brookings Institution.
Saini, G.R., 1971, 'Holding Size, Productivity, and Some Related Aspects of Indian Agriculture',
Economic and Political Weekly, Review of Agriculture, Vol.6, No.26.
Sen, A., 1981, 'Market Failure and Control of Labour Power: Towards An Explanation of
"Structure" and Change in Indian Agriculture', Cambridge Journal of Economics, Vol.5,
Nos.3-4.
Sen, A.K., 1962, 'An Aspect of Indian Agriculture', Economic Weekly, Vol. 14, Nos.4-6.
Sen, A.K., 1964a, 'Size of Holdings and Productivity', Economic Weekly, Vol.16, No.18.
Sen, A.K., 1964b, 'Size of Holdings and Productivity - A Reply', Economic Weekly, Vol.16,
No.18.
Sen, A.K., 1966, 'Peasants and Dualism with or without Surplus Labour', Journal of Political
Economy, Vol.74, No.5.
Srinivasan, T.N., 1973, 'Farm Size and Productivity: Implications of Choice under Uncertainty',
Sankhya, Series B, Vol.34, No.4.
Taslim, M.A., 1989, 'Supervision Problems and the Size-Productivity Relation in Bangladesh
Agriculture', Oxford Bulletin of Economics and Statistics, Vol.51, No.1.