,

I (

I 1
I CHAPTER

Farm size-Productivity Relationship:

The Ongoing Debate

2. 1 Introduction

2.2 Farm-size and Productivity

2.3 Summary of a Debate

2.4 Need of the Study

2.5 Stylised Facts of Indian Agriculture

•

, \ = == = = = = = = = =

1 (35)
 CHAPTER - 2

FARM SIZE - PRODUCTIVITY RELATIONSHIP:

THE ONGOING DEBATE

2.1 INTRODUCTION

The relationship between farrr;foi ze and productivity is not very

clear. Different econom ists hold different views . While some argue

that, there is an inverse relationsh ip in that productivity declines as

ta rm size increases, other economists have argued that as farm

size increases, productivity increases while, some other econom ists

argue that productivity is neutral as between farm sizes. There are

also economists who feel that inverse relationship holds over

"certain ranges of size" and not over "ali ranges of size". Th us,

instead of ciear cut conclusions on the issue, what we have is a

debate . This debate was initiated by A.K.Sen (1962,1964) with the

conclusion that an inverse relationship existed between farm si7e

and productivity in Indian agriculture, which is the principle

hypotilesi3 of this present study, to be examined for Gujarat

Agriculture.

21

1 (36)
2.2 FARMSI Z E AND PRODUCTIVITY

Du ring the pre HYV pe riod , the ques:ion of farm size wa s

us ual ly rai sed in conn ection with the foll owing three issues:

1. T h8 rela tive effici en cy of fa rm s of different size groups ;

efficiency being defi ned vd riou sly as :

a. Prod uctivity per unit of land ,

b. Productivity per unit of labour inputs,

c. Net return per unit of ca pital employed; and

d. A mount of su rplus marketed per un it of lan d.

z
2. The conce pt of vi abl e farm size which oplirni,jes the use of

bullocks, the major indivi sible input in pre-HYV agriculture, and

the stock of family la bour and provides the filmi ly with a

subsiste nce income.

3. T he issue of land re form i.e . to what extent a policy of

redistributing land in favou r of smalle r peasan ts and land less

labou rers would further the twin objectives of attaining so cia l

justice as w ell as improving the efficien cy of Indian

ag ricultu re.

Apart from these iss ues for ove r two decades, economis ts

1 (37)
 -
have bee n engaged in a debCltE on the relationship between fa rm

size, productivity and farm effi ciency. While farm produ ctivity

implie s output per un it of land , farm effi cien cy or farm profitilbili ty

refers to th e surplus of the val ue of output ove r all costs (incl udi ng

the impu ted va lue of inputs contributed by th e farm er or his fa mi ly.)

Since 1954- 55, the stu dies in the economics of farm

management un de rtaken by the Governrnent of India orovided a

statistical base for the economists to work out the relationsh ip

between farm size on the one sid6 and pr0ductivity and efficiency

on the other. The debate on farm size and productivity in India was

initiated by A.K.Sen (1962) and was later on joined by A.M .Khus ro

(1964 ), AP. Rao (1967), Hanumantha Rao (1966, 1968), Ashok

Rudra (1968) and others, but the con troversy has remained largely

inco nclusive . Most of these stud ies observed an inverse

relatio nship between farm size and output per hectare in the Farm

Management Studies Sen(1962), Khusro (1964), Sain i (1969) an d

•

others] i.e. small farms in India were far more productive than I"'ge

holdings because of intensive cul;ivation.

One explanation for this relationship was offered in terms of

the superior quality of land under smaller holdings; mo re fertile 13nd

provides 9reate r oppo rtun ities for earn ing income , so that the

family size expands at a faster rate which leads to quicker

sub-division of more fertile land over time [A.K.Sen;1964]. Thrs

23

1 (38)
expla nation was la rge ly supported by B. Oasgupta (19T!) in

studying village socio-econo mic sys tem s, who fo und that areas char-

acterize d by land scarcity we re usually mo re advanced in terms of

agricul tural inputs and yield , whe reas land is relatively , nore

abundant in low - yield backward areas. Whe n the fert ility

differenti al is fully accounted for, as Khusro ( 1964) shows , the

negative relat io nshi p between size and p rod uctivity d isap pe ars.

Krishna Bh aradwaj (1964 ) fi nds , usi ng the sa me data base th at tnere

is virtually no e'Jide nce of th is negative relationship when in dividual

crops are considered; this, however, does not necessarily

contradict the above hypothes is as the prevailing cropping patte rn

usuallv ta kes account of fertility diffe rences in soil. The major criti-

cism of this explanation is its neglect of possibilities of non-farm

employment of members of large famili es [K.B ha radwaj (1964)J.

A more important expl an ation, perhaps, is the relatively much

higher labour input in smallar holdings. The Sn13i1er a farm, the greate r
.
is usually its reliance on famil y labou r and the larger a farm, the

higher is usually the proportion of hired labour in its total labour

input. As Sen argued, on the basi s of follow ing figure 2. 1 giving a

margina l product curve M P apr ,lied to a given area of land and a

wage rate OW for hired wo rk ers ; a large fam', would employ OC of

labo ur and receive a profit of MWB .But a farm based on fam ily

labou r would employ the fa m ily me mbe rs upto the point where the

margina l prod uct of labour is zero ; that is, he would employ OP of

2-1

1 (39)
la bour If 0\\1 IS Imputed as til e w age for fami ly labou r, the total

IJI'L'lll cost 101 tilt' fam ily becomes OWAP, whil e tile lola I ou tpul

L'C'CL'nk'S O~ (P, and tile profi t O MP - OWAP = MWB - BAP. In other

\",",~ s , til e famll\ farm will produce BC P more , and earn BAP less

orott. than tile farm based on wilge la bour. T ll ere is nothing ill ogical

about thi S decI sion by the fam ily farrner to increa se the output (which

s ,ar9811 for subS istence ) and employmen t of fam ily llle lll Ders

t3SSJIllIng no alt ernat 've opportunity for em ployment outs ide) al tile

cost of profit [AKSe n; 1962 ).

I

\
I
l

I
Not C , I do smaller farms apply mo re labou r per unit of land ,

E \ c! so Lu tl, ate II rlore If'tens :ely 10 te rms of o{her inputs . In

tl-" d tc'-3:"e exp.a~a: ors 9' . en for higher produclivity on

1 'arr's v. ere mall' i related to the follol'lin g aspects:

Irttlnsl.e use 0; '3rrl" labour on small farms,

f) Jal 'at. e d,fferences In land and labou r Inputs; and

,fOp" nl) pdttern Crop Intensity and Tech nological

,-.,-

1 (40)
 -

It is also to be noted that, macro-study 'Jf 1961 census data

shows the positive relationship between the proportion of smal l

holdin gs on the one hand, and proportion of irrigated land and

doubl e-cropping on the other [B.Dasgupta; 1971]. Given the hig hly

aggreg ative nature of the data used, from this finding alone one

cannot infer that at the farm level, small er farmers resort to more

in tens ive cultivation; but statistica l analysi s of ferm man agem ent

su rv ey data, as also data at still lower level s of agg reg ation, points

to the same conclusion [K. Bharadwaj; 1974].

A related is sue is the qU8stion of management : whereas the

small er farms are managed with the support of family members for

the common good of the family (freq uently under the ste rn but

benevolent direction of the family head ), the larger fa rms , wh ich

rely mo re on hired labour, find the ma nagement of farms a much

harder occupation [C.H.Hanumantha Rao; 1966] In pre-HYV agri-

culture, the smaller farms usually apply more of al most every ir.~uts

- including fertilizers and bullock power, although the Imter is more

an indic2tion of non-optimum use of an indivisible resource per unit

of land and even more of land itself in lerms of the proportion left

fallow or used for erecting building and othe r purposes .

By the mid-sixties, land reform and the role of small farmers

in indi an agriculture became explosive political issues; thus, when

-"
'J '

1 (41)
the new technology associ3ted with the high-yielding seed varieties

was intreduced , its implicatiens fer fa rms of different size

categeries became a major issue of debate. The propenents of the

new techn elogy contended that it was 'scale neutral'; new inputs

li ke fe rt ilizers , inse cticides/pesticides and seed s were divis ible and

could be used in the sa me propertion by both the large and the

small farmers al ike [6.Sen ;1 974J.

Hewever, ep penents ef the new technolegy pointed to. the

indivi sibility of machinary, which ceuld be me re optimally exploited

by the large r farms, to. which tile propenents, while admitting the

rele ef the tra ctors and tu bewell s unde r the new technelogy,

replied that their using time was divisible, and that it was pe3~; i b le

fer even a smal l farmer to. hire tractor tim e in accordance wi,n his

farm size . On the ether ha nd, the oppenents argued abeut

imperiectiens in the facter market and the small farmer's li mited

access to. it [Keith Griffin; 1972J . Since -nest ef the new inputs had

to. be purchased frem the market with cash 01' through the

ce-eperatives or banks and because ef the unequal access to

credit and co-operative resources by the farmers of different size-

greups, the opponents argued that the new technology had

widened the disparity between the smal l and large farmers in the

country1 [C .H. Hanumantha Rae ; 1975J.

1. While admitting the scale-neutralityoftlJe technology, C.H.Hawmantha

Rao, pOinted ouf that it was not resource neutral, and so urrered greater
opportunity to bigger farmers, with more initial capital and lalger access
to the credit market.

27

1 (42)
 In sho rt , the green revolution in Indian agriculture has been

characterised basically by a capital inten sive technoloqy in which

hybrid seeds, use of chemical fertilizers, irrigation etc., playa

significant role. Even thoug h the new agricultural technology is "f ize-

neutra l" the access to capital and use of inputs for small and large

farms, has not been the same and accordingly the dist ribution of

gains between them has been uneven. In other words, after the

green revolution in Indian agricu ltu re (by introducing new

agricultural technology) , the inverse relationship has yielded place

to pos itive relationship which means as farm size increases , the

output (income) increases more than proportionately [Saini; 1979).

Thus , the income gap between small and large farms has widened

[Hanumantha Rao; 1975, Saini;197S, Singh and Petel; 1973).

By the early seventies, most of the researchers and

commentators working in this field agreed ttmt the introduction of

th e• new technology had worsened the distribution of income in the

country side, largely because of differential access of various farm

size categories to new inputs and credit. The proponents of the

new technology now argued that this was unavoidable; without the

participation of the large, "progressive" knowledgeable, resource-

fu l and enterprising farmers, the technology would have had little

chance of succcess [S.S.johl, 1974J. Unlike the small farmers,

who were tradition-bound, risk-averting 2nd devoid of means to

undertake expansive investments, the larger farmers were capable

2H

1 (43)
of mode rn izing ag riculture and of operati ng with a time horizon which

spans seve ral agricu ltural years, in addi tion to being highly

effective in bring ing la rge surp luses to the market; while th e

distribution of income followi ng the adopti on of new technology

favoured the larger farmers, the surplus they ge nerated helped the

growth of output and employment, which benefited the poorer

sections of the village populatiol1. Moreover, with time spread of

knowledge and im proveme nts in the facilities for distributing inputs,

more small fa rm ers would adopt new se eds and the associated

inputs and practices ; in the long run, the income distribution in the

coun tryside would not remain as skewed as it was in the early years

of tho new technology.

Those, who opposed tile above formula tion contestee the

argument that the selective approach of the new agricu ltura l

strategy favouring the rich farmers was necessary and unavoidable;

a better strategy would have been, they contended , to carry out

fac-reaching institutional changes, including radical land reforms and

eliminati on or at least effective curbing of the influences of the rllra l

rich. Such a policy would have re ceived popular support and helped

in inc reasing production. In contrast, the new technology was

causing growing proletarization of the peasantry, alienation of the

large majorit y in the country side from the government programmes,

and strengthening of the vil lage power structure; this process of

growing inequality could nat be reversed without sweeping

institutio nal changes.

29

1 (44)
 -
An important aspect of the debate on the new techno logy is

its implications fo r a policy of la nd reform. The empi rical findings -

as shown ea rlier - of inve rse relationshi p between the farm size and

productivity during the fi rst part of the sixties added forceful weapon

to the armoury of the proponen ts of land reform. A major empirical

question now is to what extent such relatio nsh ip holds u,lder the

new technology and if it does not, does it weaken the case for land

reform and associated measures for helping the small farmers? A

subsidiary question is, how has the new technology affected the

small farme rs in terms of theil rate of adoption and return '!

2 .3 SU M M A RY OF DEBATE

In sum, Sen argued that an invE,rse s i ~e - prod uvtlvity

relat ionship existed in Indian agriculture. Th is conclus ion was

reached at on the basis of the data preser,ted in the Farm

Mana~ement Services conducted in different parts of the country.

•
Many other economists, like D.Mazumdar, A.M.Khusro, G.R.S aini

and C. H.Hanumantha Rao, also analysed the data and reached

conclusions similar to Sen(1962, 1964).

Ashok Rudra (1968), joined the debate and argued that while

inverse relationship may hold in certain areas, it is not a universa l

phenomenon and can not be said to operate in all parts of the country.

His conclusions are best summed up in his own words as, "we may

.10

1 (45)
 '.

emphasize that we never expressed the view that the inverse

relationship was not to be observed in any circum stances in Indian

agriculture. Our view was that s:Jch a relationEhip could not be

regarded as a universally valid law operating in Indian agriculture;

and that there were indications in the Farm Management Survey

data themselve s that in certain areas yielcl per acre, instead of

declining with increasing size might actual y be inr.reasing . Also,

that in certain cases where on an inverse relati onship could be

recogni sed to hold it might do so among the smallest size class of

farmE,rs but not among others. It was our view that in many areas ,

one could not possiblly observe any systematic patte rn of

dependance between yield per acre and farm 's size. "z

In a study based on disagg regated data (belonging to the

same village), A. P. Rao concluded t lat productivity remained

constant over all the holding sizes, i.e. holding size had no effect on

productivity. Krishna Bharadwaj also reached similar conclusion .

•

Thus, what emerges from the debate on the relationship

between farm size and productivity is a totally confusing picture ;

studying the same data, different economists have reached dif:erent

conclusions by eompolying different statistical techniques. Howe "er,

the two main rivals in the debate, AKSen and Ashok Rudra, later on

attem;:>ted a synthesis of their views a~d restated the ~osition, thus:

2. Ashok Rudra, Indian Agricultural Econol'Jics ; New Delhi; 1982.

.1 I

1 (46)
 - ,

"The total ity of empirical research on the relationship betv/een

farm size and productivity has yielded a far from uniform pir :ure .

Even those who have emphasised confirmation of the inverse

rel atio n on the basis of individu al household data have Iloted failure

to see such a pattern in several region,. The general conclusion to

emerge is the diversity of Indian agriculture, regarding the exictL) nce

of the negative re lation between size and productivity;' the negative

re lation may hold in certain parts of the country, at certain times but

not everywhe re and not at all ti mes'. It also appears that even whe n

the inverse relation is more freq uently confirmed tha n rejected, it

wou ld be a mi stake to take it to be an empiri cal ge ~era lisa tio n for

Indian Ag ricultu re as a whol e.,,3

Thus, both a~thors have come to ag~ee that (i) the inve rse

relat ion is not a "univers al" phenomenon in Indian agricultu re, and

(ii) in the various regions where tho inverse relation has been teste d

it is "more frequent ly confirmed than rejected".

It is thus clear from the above discussion that, even thoug h

most of the studies asserted an inverse relationship between farm

size and productivity per hectare, the co ntro versy has rem ained

largely inconclusive. The present study, therefo re, is an attempt to

esti mate the productivity differential between small and large farms

3. Ashok Rudra and Sen, A K "Farm size and Labou r use : Analysis and
Policy" Economic .nd Political Weeki;, Annual Number; Feb ., 1980 .

.11

1 (47)
 '.

and then to decompose this differen ce into (a) neutral techno- •

logical difference4 (b) non-neutral techni cal differen ces, and

(c) input - use contribution (efficien cy) in Guj arat state, to answer
the questions rai sed .

2.4 NEED

In ['last, th rough Farm Management Stu dies, a few attepmts

were made to quantify the contribution of the above pos,ible

factors to the overall productivity, productivitl' defference and / or

overall resource -use efficiency between small and large farms. In

present study, therefore, an attempt has been made to look at

the controversy relating to the relationship between size of farm

holding and productivity with a different approach, that is, using

decomposition analysis of the change in ratio . The decomposition

technique adopted here allows one to have a more systematic look

at the factors and their contributions to the crop-wise productivity

cl "
4. oera il:e xpositiol7 of thenconcepts of neutral and non~neutral techn%gica l
change can be fou nd in Hicks (1964) , Brown , (1968), Nadiri ( 1970) and
Ferguson ( 1970). But in the present study they are defined as:
Effect of technological change on per hectare productivity of small
farms over large farms is equal to the ratio of the scafe parameters of
the production function of the small farms to the large farms, it is called
'neutral technological difference", and
Effect of Technological change on per hectare projuctivify of small
farms over large farms is equal to the ratio of the elasticities 0 1 the
inputs of small farms to the large farms, when large farms use same
level of technology as small farms; this is called "non·ll eutral teclmo-
,ogical difference".

33

1 (48)
 -
differencials between the farms or different size group . Most of the

Farm Manilgement Studies have been conducted in the states other

than Gujarat which are relatively better endowed zones of Ind ia.

Thu~" the importance of this study is further enhanced beciluse ot

the paucity of studies for Gujarat State.

Further, a study of the d fferen ces in crop-wise productivi ty

betwee n farms of different size group have important implications

for :

(a) Land reform policies ,

(b) Alloca tion of public resources to help farmers,

particularly th e smal l ones to increase production

efficiency; and

(c) Development of technology.

2.5 "S TYLISED FACT3"s OF INDIAN AGRICULTURE

Some of the "stylised facts" of the Indian Agriculture as

observed by other Indian economists are to be examined in

relation to Guja rat, which form the bases of the hypothesis of the

present research work.

5. It may not be appropria te to calilhe following hypotheses as "stylised !'acls"

be ,ause except for the firsl one all olher hypotheses are not yel
empirically well esta bilshed and generally accepted by economists But
ior wanl of a beller lerm, Ihis lerminology is adopled from "Growlh Tlleory"
literature for fhe purposes of the present study.

1 (49)
 •

These are as under:

1. There exists an inverse relationship b0tween the land size of

holding and productivity per hectare [Farm Management

Survey ReportsJ i.e. to estimate the crop-wise productivity

differential between farms of different size groups.

2. W hen family labou r employed in agriculture is given an

im puted value in terms of the rJling wage rate, much of

Ind ian agriculture seems unremuner<ltive [A .K.Sen, 1962J.

3. By and large , the profitability (efficiency) of agriculture

increases with the size of holding , "profitabil ity be ing

meas ured by the surplus (or deficit) of output over C0StS

Including the imputed value of labour" [AKSen , 1962J.

4. Labour days spent per hectare 0n all crop production are

,

Inversely rela ted to the size of holding [Krishna Bharad' vaj,

1974 J '

Moreover, less discussed bu t meaningful hypothesis can

be ad ded :

5, The adoption rate of high yielding va rieties is generally higher

for ~ ig farmers [Hanumantha Rao; 1975J.

1 (50)
 -
The above mentioned hypotheses are neither mutually

exclusive nor collectively exchaustive categories. Some of the

hypotheses are interrelated with one another and the existence (or

valid ity) of one may follow the existence (or the validity) of t:1e other.

The main purpose of listing of these hypotheses is to examine later

on, how far Gujarat State exhibits these charasteristics which were

observed in other zones of India [Ashok Rud ra ; 1968, Rudra &

Sen; 1980J. It should be noted further in this ,;ontext thai the firs t of

th e above stated statements is the most fundamental and all others

are either its consequences or its explanation. In the prese nt study,

the re fore, it is inteded to discuss this" stylised fact"in some what

mo re detailed form in CHAPTER-6, following the discussions about

ag ro-economic profile of Gujarat State, relate d met hodology, data

base and concepts .

36

1 (51)