You are on page 1of 16

IMPACT OF MECHANISATION ON THE

EMPLOYMENT OF PERMANENT FARM


SERVANTS IN INDIAN PUNJAB: A
COMPARISON OF COTTON-WHEAT AND
PADDY-WHEAT REGION

Dr. Varinder Sharma

Development Studies Unit,


Institute for Development & Communication (IDC),
Chandigarh.
Email: Varinder_10@hotmail.com
Impact of Mechanisation on the Employment of Permanent Farm

Servants in Indian Punjab: A Comparison of Cotton-Wheat and

Paddy-Wheat Region

Abstract

The new agricultural technology completely changed the agricultural


labour market in Punjab. With the mechanisation of farm operations,
demand for skilled and trained permanent farm servants, hired for the
whole year, increased on Punjab farms. But in literature not a single
study exists which analysed the impact of mechanization on the
employment of permanent farm servants in post green revolution
Punjab. In the present study we have attempted that and found that
tractorisation is the most important factor in increasing the
employment of permanent farm servants, followed by the use of power
operated tubewells for irrigation. In addition to these two factors, in
cotton-wheat region, a positive association exists between the number
of adult male workers of employer’s (farmer’s) family and number of
permanent farm servants working on the farms.

I. Introduction:

The new agricultural technology raised the production and


productivity in Punjab agriculture in a big way since the mid 1960’s.
The main factors behind this big rise in production and productivity
were HYV seeds, tractorisation and the use of tubewells for irrigation.

Dr. Varinder Sharma


Development Studies Unit, Institute for Development & Communication (IDC),
Chandigarh
Email: Varinder_10@hotmail.com

1
Most of the scholars focused on analysis of growth of agricultural
production and productivity under the impact of Green Revolution
Technology. But some scholars also analyzed its impact on the
employment of agricultural labourers.

No doubt with this new agriculture technology, the mixture of


traditional agricultural labour markets changed markedly in Punjab,
and a new class of cash wage agricultural labourers emerged over the
years1.

At present, after about four decades of beginning of the Green


Revolution, two types of agricultural labourers are found on Punjab
farms: casual agricultural labourers and cash wage permanent farm
servants. A casual agricultural labourer is hired for a day whenever
the need arises of his services on the farm. A permanent farm servant,
on the other hand, is hired for the whole year for various skilled and
non-skilled farm operations. His wages, perks and terms of
employment are mutually settled by employer and worker in the
beginning of the year in a written or oral contract. The nature of work,
working hours, man days of both types of agricultural labourers and
more over their demand on the farms completely changed with the
mechanisation of farm operations. In literature most of the scholars
remained solicitous only on capturing the impact of mechanisation on
the employment of casual agricultural labourers. There is not even a
single empirical study on the impact of mechanization on the
employment of permanent farm servants in Punjab, despite their
importance in Punjab agriculture; may be due to non-availability of
data on permanent farm servants in the secondary data sources. A
permanent farm servant is generally hired on big farms 2, for
performing more skilled and responsible farm tasks 3 e.g. for spraying
chemicals on crops and manuring etc4. Moreover, tractorisation

2
increased the demand of skilled permanent farm servants on Punjab
farms who can drive tractors5 and can supervise and arrange casual
agricultural labourers. Inspite of their importance in Punjab
agriculture no study has empirically studied the impact of
mechanisation on the employment of permanent farm servants. The
present study has tried to analyse the impact of mechanisation on the
employment of permanent farm servants in the post green revolution
period.

The matter is arranged into the following sequence: First, the


data sources and methodology have been discussed. Second, selection
and definition of variables has been given. Third, data analysis and
model used are discussed. Fourth, the characteristics of surveyed
farms have been discussed. Fifth, the empirical results estimated with
the logit model are presented and explained. Sixth, the estimated
coefficients are compared between the two cropping patterns i.e.
cotton-wheat and paddy-wheat. Lastly, the conclusions have been
given.

II. Data and Methodology:

The main secondary data sources like census of India and the
NSSO (Employment and Unemployment surveys) do not provide us
any information and data on the permanent farm servants.6. Even at
the village level no official record is maintained regarding the farms
which hire permanent farm servants. The data collected by individual
researchers are not easily available to other researchers. That is why
the present study is based on a primary survey conducted by the
author. In Punjab mainly three crops are grown i.e. wheat, paddy and
cotton. In this study we selected two villages at random on the basis of
cropping pattern. One village named Shergarh from district Bhatinda
in which cotton and wheat cropping pattern is dominant. The second

3
village is Dasonda Singhwalla, from district Sangrur, in which paddy
and wheat cropping pattern mainly prevails.

In these two selected villages we conducted a complete census of


cultivator households. During the census operation we collected
detailed information of farms hiring and not hiring permanent farm
servants. Besides this, we also gathered information of surveyed farms
on the cultivated area, mechanisation, cropping pattern and family
labour of cultivators etc. The data were collected during a year 2009-
10.

III. Selection and Definition of Variables:

On the basis of some earlier related studies7 we have selected the


following variables. The dependant and some of the independent
variables are in the binary form [‘0’ and ‘1’] in this study.

Dependant Variable: A farm hiring a permanent farm servant (=1)


and a farm not hiring a permanent servant (=0).

Independent Variables:

Variables Definition
Tractor ownership with employer This variable is in a binary form
[owns=1; otherwise=0] and is also a good proxy for size of
the farm;
Electric motor ownership This is again a binary variable and
[owns=1; otherwise=0] a good proxy for the irrigated
farms;
Adult male members of employer’s This is a continuous variable and
family working on the farms denotes actual number of male
members of employer’s (farmer’s)
family working on the farms;
Region: This is a binary variable for the
Cotton-Wheat=1,Paddy-Wheat=0 location of a farm.

4
IV. Data Analysis and Model Used:

There are a number of different probability models to carry out


the analysis when the dependant variable is in a binary form. Each of
these models has its own merits and demerits [Amemiya (1981)]. In
this study we used the binary logit model which is given by:

Pr(Yi =1/Xi) =1/(1+exp(-βXi) -(i)

Where ‘’ represents a row vector of coefficients in the logit form model
is written:

 Pr (Yi =1/Xi ) 
In   = βxi (ii)
 Pr(Yi = 0/Xi ) 

The log likelihood function for this model may be given as:

i 1+ exp(βxi) - (iii)


L(β) = βΣi XiYi -ΣIn

By differentiating ‘’ in eqt (iii) we get the maximum likelihood


estimation (ML). The first derivative of the log-likelihood function is

 L(β) Ù
= Σ Xi Yi -Σ Xi Yi - (iv)
β i i

By equating this derivative equal to zero we get equation (v)

Ù
Σ XiYi -Σ XiYi = 0 - (v)
i i

For most of the data and models the equation (v) does not give
explicit solution. In such cases the most common method is the
Newton-Raphson algorithm.8

For this purpose let d () be the vector of first derivatives of log-
likelihood with respect to ‘’ and L() be the matrix of second
derivatives

5
i.e.

δL(β) 
U[β] = = Σ XiYi - Σ X Yi
δβ i i

δ2L(β)  
ln[β] = = -Σ XiXi' Yi(1- Yi)
δφ i

The vector of first derivative U() is the gradient while the matrix of
second derivatives I() is called the Hessian. The Newton-Raphson
algorithm is then:

β j+1 = β j.I-1 (β j ).U(βv )

Where I-1 is the inverse of I. In this study we relied upon the Newton-
Raphson method for the estimation of coefficients.

V. Characteristics of Surveyed Farms in the Two Villages:

In this section we describe the characteristics of surveyed 368


farms. Out of these farms, 176 (47.8%) farms which hired 40
permanent farm servants located in a cotton-wheat region’s village. On
other hand, 192 farms (52.2%) employed 60 permanent farm servants
were situated in a paddy-wheat region’s village (Table-1). If we look at
table-1, it depicts the average holding size is bigger by seven acres in
the village of cotton-wheat region than the paddy-wheat region village.
This difference of seven acres is highly significant at 1 per cent level.
When we compare the degree of tractorisation in these two villages
then we find that in the village of paddy- wheat region, the number of
tractors is higher only by 5 percent than the cotton-wheat region
village and this difference is not significant even at 10 percent level.

6
Table-1

Characteristics of Surveyed Farms in Cotton-Wheat & Paddy-


Wheat Regions’ Villages

Variables Cotton- Paddy- Difference


Wheat Wheat β tvalues Intercept R2
Region Region ( )
Average holding 9.42 2.63 2.63 6.79 0.03
size (Acres) 3.22***
Percentage of 33.52 38.54 (-) 0.03 0.75 0.39 0.00
farms having
tractors
Percentage of 19.89 51.04 (-) 0.31 0.51 0.10
farms having 6.60***
electric motors
Average number 0.23 0.31 0.31 (-) 0.08 0.01
of permanent 1.79*
farm servants
hired on farms
Average number 1.59 1.41 1.41 0.18 1.36 1.41
of adult male
members of
employer’s
family working
on farms
Independent
Cotton- Wheat Region = 1
Variable
[Dummy] for
testing Paddy- Wheat Region = 0
difference
Regionwise
number of
Cotton-Wheat Region = 40 (40%)
permanent farm
Servants
[N]
Paddy- Wheat Region = 60 (60%)

Total = 100

Cotton- Wheat Region = 176 (47.8%)


Total surveyed
Farm (N) Paddy- Wheat Region = 192 (52.2%)
Total = 368

tvalues significant at ‘***’ 1 % and‘*’10%

7
Further, the number of electric motors exceeds by 31 percent in
the paddy-wheat region’s village relatively to the cotton-wheat region’s
village. This difference of 31 percent is highly significant at 1 percent
level (Table-1). From this we may infer that the village falling in the
paddy-wheat region is using more tubewell than the village of cotton-
wheat region, but has smaller average size of holding. The
tractorisation in the two villages is almost similar.
Two villages are also different in case of employment of
permanent farm servants. The average number of permanent farm
servants working on farms is 0.31 in the village of paddy-wheat region,
where as in the village of cotton-wheat region it is 0.23. This difference
of 0.08 is significant at 10 percent level. It means in paddy-wheat
region more permanent farms servants are hired. Moreover, no
difference was found (Table-1) in terms of average number of adult
male members of cultivators’ households in the two cropping regions.
The estimated mean difference is not significant even at 10 percent
level.
From this discussion we may say that the selected villages are
quite non-homogenous in terms of average holding size (acres),
irrigation and average number of hired permanent farm servants on
farms. In the next section we have tried to explore how these
characteristics affect chances of employment of permanent farm
servants in two different cropping zones.
VI. Impact of Mechanisation and Availability of Family Labour on
the Employment of Permanent Farm Servants: Empirical Results:
In this section we have applied the logit model to analyse how
the previously discussed characteristics impact the probability of
employment of permanent farm servants. The results estimated with
the logit model are presented in tables 2(A) and 2(B).

8
From equation -1 given in table 2(A) and 2(B) we may conclude
that tractorisation in each village increases the chances of
employment of permanent farm servants. The estimated coefficient has
a positive sign and is highly significant at 1 percent level; moreover R2
is also quiet high. The AIC is quiet low in comparison to other single
variables given in equation-2 and equation-3.From the estimated
coefficients and slopes we may infer that in cotton-wheat region’s
village with the tractorisation of farms the chances of employment of
permanent farm servants are higher relatively to paddy-wheat region’s
village.
Table 2 (A)
Impact of Mechanisation and Family Labour Supply on the Employment of
Permanent Farm Servants in Cotton-Wheat Region: Logit model
Dependent Variable (Y) : A farm hired a permanent
farm servant =1
Explanatory otherwise = 0
Variables(x)
Eqt.1 Eqt.2 Eqt.3 Eqt. 4

2.59 2.09
Tractor ownership (0.43) (0.46)
- -
[Dummy] = 1 [5.96]*** [4.51]***
{0.45} {0.34}
1.74 0.92
Electric motor
(0.41) - (0.48)
ownership -
[4.25]*** [1.92]*
[Dummy] = 1
{0.35} {0.06}
Adult male 0.60 0.38
members of (0.16) (1.82)
- -
employer’s family [3.91]*** [2.10]**
working on farms {0.09} {0.06}
Intercept -2.48 -1.69 -2.31 -3.17
McFadden R2 0.23 0.09 0.09 0.28
Adj R2 0.21 0.07 0.08 0.24
AIC 149.09 174.60 174.32 143.55
Likelihood Ratio
43.56*** 18.05*** 18.34*** 53.10***
Test
Total Farm
176
(N)
Note: *,**, *** significant at 10%, 5% & 1% [Two Tailed]
‘( )’ contains standard Error ‘ [ ]’ contains Z values.
‘{ }’ contains slopes at mean.

9
This may be due to the greater use of tractors on farms in
cotton-wheat region to extract ground water for irrigation, spraying of
chemicals on cotton crop and at the time of sowing cotton and wheat
crops.

Table 2 (B)
Impact of Mechanisation and Family Labour Supply on the
Employment of Permanent Farm Servants in Paddy-Wheat Region:
Logit model
Explanatory Dependent variable (Y) : A farm hired permanent farm
Variables (x) servant =1
otherwise = 0
Eqt.1 Eqt.2 Eqt.3 Eqt. 4

Tractor 1.64 - - 1.23


ownership (0.33) (0.39)
[Dummy] = 1 [4.86]*** [3.17]***
{0.34} } {0.25}
Electric Motor - 1.29 - 0.86
ownership (0.34) (0.37)
[Dummy] = 1 [3.79]*** [2.33]**
{0.26} {0.15}
Adult male - - 0.37 0.13
members of (0.13) (0.15)
employer’s family [2.83]*** [0.85]
working on farms {0.08} {0.02}
Intercept -1.59 -1.58 -1.39 -2.08
McFadden R2 0.11 0.07 0.04 0.13

Adj R2 0.09 0.05 0.02 0.10


AIC 213.93 223.62 230.62 211.59
Likelihood Ratio 25.32*** 15.63*** 8.91*** 31.65***
Test
Total Farm
192
(N)
Note:
***, ** significant at 1%& 5%
‘( )’ contains standard Errors
‘[ ]’ contains Z values
‘{ }’ contains slopes at mean

10
On the other hand in paddy-wheat region during paddy season
the tractor is used only for preparing fields for the transplantation of
paddy and at the time of sowing of wheat. Due to this difference in
use of tractors for various farm operations in two regions there may be
more chances of employment of permanent farm servants in cotton-
wheat region, the bigger size of holding may also be contributory to the
observed tendency as owners of big farms usually employ permanent
farm servants as tractor drivers. Similarly, the use of electric-motors
on farms increases the probability of employment of permanent farm
servants [Equation-2 in table- 2 (A) and 2 (B)]. In each equation-2, the
estimated co-efficient is highly significant at 1 percent level and R 2
reduces in equation-2 comparatively to previous equation-1 and the
AIC increases. The size of coefficient and slope also reduce in
comparison to tractorisation. In cotton-wheat region’s village the size
of estimated co-efficient and slope is higher than paddy-wheat region’s
village. It is quiet puzzling and difficult to explain because it was
expected to be high in paddy-wheat region where there is more use of
electric motors during paddy season and more labour is consumed for
managing water even during night times. This odd result may be
partly due to the tendency to hire temporary servant for three/ four
months to manage water on farms. This additional employment of a
semi-permanent farm servant may reduce the probability of
employment of permanent farm servants on more irrigated farms with
electric-motors in paddy-wheat region.

Further equation-3 given in tables 2 (A) and 2 (B) elaborates that


the probability of employment of permanent farm servants increases
with the additional number of adult male members of employer’s
(farmer’s) family. No doubt, the size of co-efficients and slope is much
lower in comparison to the variables given in equation-1 and equation-
2. In each table 2 (A) and 2 (B) the value of R2 of this variable is low in

11
equation-3. The slope is marginally high of this variable in cotton-
wheat region. It means there is more need of family labour and
permanent farm servants. It is true also because cotton crop is more
labour intensive because its harvesting is totally manual and very time
consuming. Moreover, this crop needs spray of chemicals against
insecticides frequently and takes long time to mature which needs
more labour for its care and supervision.

Finally in equation-4 [Table 2 (A) and 2 (B)], we included all the


three variables i.e. tractor-ownership, electric motor ownership and
number of adult male members of employer’s family working on the
farms. In table 2 (A) all these variables are significant as remained in
equation-1 to equation-3. The R2 is quite satisfactory and the value of
AIC reduces to 143. 55. On other hand in equation- 4 [Table 2 (B)] the
variable number of adult male members of employer’s family on the
farms is insignificant which shows it has no impact on the
employment of permanent farm servants. The size of the significant
coefficient and slopes are again high in cotton-wheat region.

From these empirical results we may conclude that tratorisation


and the electric motor operated tubewells increase the chances of
employment of permanent farm servants significantly. The
tractorisation has a stronger impact in each region. The differences in
regression co-efficient of two cropping regions also gave us interesting
results which have been compared in next section to analyse their
actual variations.

VII. Comparison of Regression Coefficients in Cotton-Wheat and


Paddy-Wheat Region:

The results of comparison of regression coefficients have given in


table-3.

12
Table-3
Comparison of Impact of Mechanisation and Family Labour
Supply on the Employment of Permanent Farm Servants in
Cotton-Wheat and Paddy-Wheat Region(Z test statistic)

Variables Equation-1 Equation-2 Equation-3 Equation-4

Tractor 1.76* - - 1.43


ownership

Electric - 0.83 - 0.09


motor
ownership
Adult Male - - 1.05 0.14
members of
employer’s
family
working on
the farms
Note: Formula used for the group comparison (βcw- βpw) where ‘CW’
stands for cotton-wheat and ‘PW’ for paddy-wheat is:
Z=βcw-βpw/√SEβ2cw+SEβ2PW.
‘*’Significant at 10% level.

In all the equations except in equation-1 (Table-3) the calculated


Z is not significant even at 10% level, except in the case of
tractorisation.It means there is no significant difference as such in the
estimated coefficients between the two regions. It means power
operated tubewells and availability of family male farm workers impact
chances of employment of permanent farm servants in a similar
manner. In equation-1(Table-3), the tractorization variable is
significant; it means tractorisation in cotton-wheat region increases
employment of permanent farm servants higher than paddy wheat
region.

13
VIII. Conclusions:

The mechanisation of farm operations in Punjab completely


changed the agricultural labour markets in terms of hiring practices
and demand for different types of labourers. Most scholars
investigated the impact of mechanisation on the employment of casual
agricultural labourers, and permanent farm servants remained out of
their focus. In this study, which has been carried out in two villages
located in cotton-wheat and paddy-wheat regions, it is found that
tractorization is the most important factor in increasing the
employment of permanent farm servants. Second important factor is
the use of electric motor operated tubewell which has potential of
creating employment of permanent farm servants. In cotton-wheat
region, besides these two factors, a direct relationship exists between
number of adult male family workers on the farm, and number of
permanent farm servants working on the farms.

Acknowledgement: I am thankful to Dr. H.S. Shergill, Professor Emeritus,


Department of Economics, Panjab University, Chandigarh (India) for
comments & suggestions to complete this paper.

Notes
1. The traditional agricultural labour markets completely disappeared with
the green revolution [Franke] (1971)];
2. In Punjab, big farms only hire permanent farm servants [Chand(1985)];
3. For a whole year talented and skilled permanent farm servants are hired
even after green revolution [Aggarwal (1971, 1981)];
4. The importance of permanent farm servants for responsible farm tasks
discussed by Mukesh and Kotwal (1985);
5. Mechanisation not only increased the demand of labourers in agriculture
but demand of permanent farm servants who can drive tractors on farms
increased tremendously in Punjab [Shyamol & Melvin (1978)];
6. The categorization of agricultural labourers into casual labourers and
permanent farm servants made in Agricultural Labour Enquiry Reports
during early 1950s criticized by Raj (1962) aftermath no such
demarcation was made by statistical authorities;
7. The probability of employment of permanent farm servants in rural West
Bengal with the logit model carried out by Bardhan (1986). He took
irrigation level of village & rainfall in district etc. as independent
variables;
8. For details see Allison (2008);

14
REFERENCES

Aggarwal, Bina (1981): “Agricultural Mechanisation & Labour Use: A


Disaggregated Approach.” International Labour Review, Vol. 120, No. 1,
pp. 115-127.
Aggarwal, Pratap C (1971): “Impact of Green Revolution on Landless
Labour: A Note”. Economic & Political Weekly, Vol. 6, No. 47, pp. 2363-
2365.
Allison, Paul D (2008): “Convergence Failures in Logistic Regression.”
www.2.sas.com/proceedings/forum2008/360-2008.pdf
Amemiya,Takeshi(1981): “Qualitative Response Models:A Survey”.
Journal of Economic Literature,Vol.19,No.4,pp.1483-1536
Bardhan, Pranab K (1986): Land, Labour and Rural Poverty, Oxford
University Press, Delhi.
Chand, Ramesh & Sidhu, DS etal (1985): “Impact of Agricultural
Modernization on Labour use Pattern in Punjab with Special Reference
to Women Labour.” Indian Journal of Agricultural Economics, Vol. 40;
No.1, pp. 252-258
Eswaran, Mukesh and Kotwal, Ashok (1985): “A Theory of Two Tier Labor
Markets in Agrarian Economies.” American Economic Review, Vol. 75,
No. 1, pp. 162-177.
Frankel, Francine (1971): India’s Green Revolution: Economic Gains &
Political Costs, Princeton University, New Jersey
Raj, K.N. (1962): “Some Comments on the Second Agricultural Labour
Enquiry.” In V.K.R.V. Rao (Ed): Agricultural Labour in India, Asia
publishing House, New Delhi, pp. 152-169.
Roy, Shyamol and Blose, Melvin G (1978): “Farm Tractorisation,
Productivity and Labour Employment: A Case Study of Indian Punjab.”
Journal of Development Studies, Vol.14, No. 2, pp. 193-209.

15

You might also like