DP Template - IFPRI - Ebrary

IFPRI Discussion Paper 01905

January 2020

Changing Composition of Private Investment in Indian Agriculture

and its Relationship with Public Investment and Input Subsidies

Anjani Kumar

Seema Bathla

Smriti Verma

South Asia Regional Office

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE The International Food Policy Research Institute (IFPRI), established in 1975, provides research-based policy solutions to sustainably reduce poverty and end hunger and malnutrition. IFPRI’s strategic research aims to foster a climate-resilient and sustainable food supply; promote healthy diets and nutrition for all; build inclusive and efficient markets, trade systems, and food industries; transform agricultural and rural economies; and strengthen institutions and governance. Gender is integrated in all the Institute’s work. Partnerships, communications, capacity strengthening, and data and knowledge management are essential components to translate IFPRI’s research from action to impact. The Institute’s regional and country programs play a critical role in responding to demand for food policy research and in delivering holistic support for country-led development. IFPRI collaborates with partners around the world.

AUTHORS Anjani Kumar ([email protected]) is a research fellow in the South Asia Regional Office of the International Food Policy Research Institute (IFPRI), New Delhi, India.

Seema Bathla ([email protected]) is a professor at Jawaharlal Nehru University, New Delhi, India.

Smriti Verma ([email protected]) is a research analyst at the South Asia Regional Office of the International Food Policy Research Institute (IFPRI), New Delhi, India.

Notices 1 IFPRI Discussion Papers contain preliminary material and research results and are circulated in order to stimulate discussion and critical comment. They have not been subject to a formal external review via IFPRI’s Publications Review Committee. Any opinions stated herein are those of the author(s) and are not necessarily representative of or endorsed by IFPRI. 2 The boundaries and names shown and the designations used on the map(s) herein do not imply official endorsement or acceptance by the International Food Policy Research Institute (IFPRI) or its partners and contributors. 3 Copyright remains with the authors. The authors are free to proceed, without further IFPRI permission, to publish this paper, or any revised version of it, in outlets such as journals, books, and other publications.

Contents ABSTRACT iii

ACKNOWLEDGMENTS iv

ACRONYMS v

1 Introduction 6

2 Database and Methodological Approach 9

3 Structure and Trends in Fixed Capital Expenditure among Rural Households 12

4 Results from Econometric Analysis 23

5 Summary and Conclusions 28

REFERENCES 30

Tables Table 1 Average values for broad heads of fixed capital expenditure (rupees / rural household

at 2004–5 prices), 1981–82 to 2012–13 Table 2 Trends in share of broad heads of fixed capital expenditure (FCE) per rural household

in major states (%), 1981–82 to 2012–13 Table 3 Impact of public investment in agriculture and irrigation on private investment in

agriculture: Regression results, 1981–82 to 2013–14 (IV2SLS) Table A.1 Annual rate of growth in broad heads of fixed capital expenditure per rural household

at 2004–5 prices Table A.2 Fixed capital expenditure in farm business (rupees / rural household) at 2004–5

prices, 2012–13 Table A.3 Share of different items of farm business expenditure per rural household (%), 2012–13 Table A.4 Annual rate of growth in fixed capital expenditure in farm business (rupees / rural

household) at 2004–5 prices, 2002–3 to 2012–13 Table A.5 Share of FCFA financed through borrowings and share of institutional borrowings in

total borrowings, 2012–13

Figures

Figure 1 Relative importance of the broad heads of fixed capital expenditure among rural

households, 1981–82 to 2012–13 Figure 2 Share of fixed capital formation in agriculture (FCFA) in gross capital expenditure

(GCE) per rural household (%), 1981–82 to 2012–13 Figure 3 Share of major states in all-India private fixed capital formation in agriculture among

rural households (%), 1981–82 to 2012–13 Figure 4 Trends in composition of fixed capital expenditure in farm business among rural

households, 1981–82 to 2012–13 Figure A.1 Relationships among concepts of capital expenditure in the NSS-AIDIS Figure A.2 Trends in composition of fixed capital expenditure in farm business among rural

households, excluding livestock, 1981–82 to 2012–13

iii

ABSTRACT

Using the decennial All-India Debt and Investment Survey from 1981-82 to 2012-13, this paper delves

into the spatial and temporal trends in private fixed capital expenditure and its composition, among rural

households in India. We also assess its relationship with public investment in agriculture. Amidst

sizeable ups and downs, the magnitude and rate of growth in private investment in agriculture has gained

momentum from 2000s except in Odisha, Himachal Pradesh, Jammu and Kashmir. An increasing

preference of farmers to invest in residential land and buildings, and that at the cost of asset formation

in farm business, is evident in agriculturally advanced states. Within agriculture, relatively higher

investments in land improvement, machinery-implements, tractors, and livestock are identified over the

period. Importantly, such investments are positively influenced by public investments in agriculture

and irrigation in the high and low income states and also by public spending on input subsidy in the

middle and low income states. An increase in public expenditure that is well targeted and is

commensurate with farmers’ investment portfolio would reinforce a complementary relation between

the two across-the-board. The impact of terms of trade on private investment though positive turns out

to be statistically insignificant. Land acts as a constraint, indicating need for policy interventions that

augment crop yield and can bring remunerative prices to farmers. A continued effort to improve the

outreach of formal financial institutions for credit is warranted for higher private capital formation.

Keywords: gross capital expenditure, private investment in agriculture, farm business, public

investment in agriculture, borrowings

JEL Classification: Q14, Q18

iv

ACKNOWLEDGMENTS

This work was undertaken as part of the ICAR-IFPRI workplan. We are thankful to the Indian Council

of Agricultural Research (ICAR) for its financial support to undertake this study. The inputs provided

by the participants in the ICAR-IFPRI Annual Work Plan Review meeting are duly acknowledged. This

study is aligned with CGIAR Research Program on Policies, Institutions, and Markets (PIM), led by

the International Food Policy Research Institute (IFPRI).

v

ACRONYMS

AIBP Accelerated Irrigation Benefit Programme AIDIS All India Debt and Investment Survey AY Agricultural Year CSO Central Statistical Office FAO Food and Agricultural Organisation FB Farm Business FCE Fixed Capital Expenditure FCE-FB Fixed Capital Expenditure in Farm Business FCFA Fixed Capital Formation in Agriculture GCE Gross Capital Expenditure GDCF Gross Domestic Capital Formation GDP Gross Domestic Product GSDP Gross State Domestic Product HIS High-Income States LIS Low-Income States MIS Middle-Income States NFB Non-Farm Business NSS National Sample Survey NSSO National Sample Survey Office OLS Ordinary Least Square RPB Residential Plots and Buildings SE Standard Error

6

1. INTRODUCTION

Agriculture and allied activities are practised in heterogenous agroclimatic conditions and geographies

in India. Farmers make investments in farm machinery, irrigation, livestock and other fixed assets, such

as construction of barns and sheds, which are financed through either their own savings or borrowed

funds. The role of government stands central given the small resource base of majority of farmers for

asset formation and various weather and market related risks that they generally face1. Government

intervention may manifest as large investments in major and medium irrigation systems, roads,

transport, agriculture markets, flood control, and dam management. Other crucial avenues for public

expenditure in agriculture are in the form of enhanced flow of institutional credit to farmers at

subsidized rates or indirect support through fixation of lower prices of inputs and direct subsidy on

seeds, machinery or facilitation of extension services and crop insurance to farmers to cover risks.

Government’s active support to agriculture is expected to induce farmers to make private investment

and together, public and private investments contribute to accelerate agricultural growth. A positive

relationship that is expected between public and private investments in agriculture has its roots in

macroeconomics which contextualises higher public investment to ‘crowd-in’ private investment.

Concomitantly, higher government expenditure on infrastructure, health, and education may have a

‘crowding in’ or ‘inducement’ effect on private investment by raising its marginal productivity. The

evidence on this relationship in the Indian context is found mixed at the economy-wide level (Mitra

1991; Bahal et al. 2018) but is very well established at the sectoral level. Dantwala (1986), Rath (1989),

Gandhi (1990), Shetty (1990), Dhawan (1998), Chand (2000), Gulati and Bathla (2001) quantified

strong interlinkages between the two sources of investments in agriculture during the sixties to mid-

nineties. Public investments in the development of canal networks in tandem with private investment

in minor irrigation and adoption of high yielding varieties contributed to higher crop output.

1 The natural risks, pest and insect attacks, crop destruction by stray and wild animals, and fires result in loss of yield while other risks concern market failure often arising out of low price of produce, collusion among traders, and inefficiency in the input markets.

7

A weakening of this relationship was, however, observed during the mid-eighties and the nineties

consequent upon a stagnation in public irrigation investment. In fact, Mishra and Chand (1995) refuted

the complementarity hypothesis between public and private capital formation in Indian agriculture in

the long run. Their argument was largely based on the fact that complementarity arises due to limitation

of official data on public investment, which are available only for major and medium irrigation. The

share of public capital formation in agriculture and allied activities in total capital formation has

consistently declined from almost 50 percent during the sixties to 15 percent in 2010s. But the

significance of public investments in agriculture, irrigation, and rural infrastructure for agricultural

productivity, income, and rural poverty remains irrefutable (Fan et al. 1999; Fan 2008; Bathla 2017).

The debate on the relationship between public and private investment saw a renewal recently due to a

revival of public investment along with higher financial outlays toward provision of input subsidies,

direct income as well as investment support to farmers in most states. The ‘crowding in’ effect of public

investment on private investment in agriculture, which is firmly established at the national level is less

researched at the disaggregate level. Similarly, whether input subsidy given directly or indirectly to

farmers impacts their investment decisions is hardly known, though Dhawan and Yadav (1995), Roy

and Pal (2001), Gulati and Narayanan (2003), Sharma (2012), Kannan (2014), and Bathla (2014) have

tried to delineate these issues. However, the findings remain inconclusive as only a few studies

supported the role of input subsidies in accelerating private investment in certain regions. The

concurrence is mainly on the determinants of private investment in agriculture, broadly taken to be

terms of trade, technology, returns from farming, and availability of land and credit (Mishra and Hazell

1996; Chand and Kumar 2004; Bisaliah and Dev 2010; Bathla and Kumari 2017).

We take up these issues for investigation in greater detail using the decennial All-India Debt and

Investment Survey (AIDIS) of the National Sample Survey Office (NSSO) and the Finance Accounts

that provide state level quantum of private investment and public expenditure, respectively. We

hypothesize a positive and significant effect of public investment and input subsidy on private

investment in agriculture at the disaggregate state level having varied geographical conditions and

8

levels of development. This hypothesis is tested over a longer period of time from 1981-82 to 2012-13

to cover different phases of policy reforms. Besides, the study uses a two-stage least squares (2SLS)

model which has an advantage over ordinary least squares as it addresses the problem of endogeneity

in the explanatory variables. In all, 17 major states are taken up for analysis. The states are grouped into

high, middle, and low income states (HISs, MISs, and LISs, respectively) based on the average per

capita income in each from 2000-01 to 2013-14. The low income states include Bihar, Uttar Pradesh,

Assam, Jammu & Kashmir, and Madhya Pradesh; the medium income states are Odisha, Rajasthan,

West Bengal, Andhra Pradesh, and Karnataka; and the high income states include Punjab, Himachal

Pradesh, Tamil Nadu, Kerala, Gujarat, Haryana, and Maharashtra. Panel data is created for empirical

analysis of four groups (all-India, HISs, MISs, LISs) for the 33-year period.

Before conducting an empirical analysis, we delineate the investment behavior of farmers, particularly

their asset preferences, changes in the composition of investments, and the sources of finance for such

investments. The NSS collects data from households on gross and fixed capital expenditure in farm

business (synonymous with fixed capital formation in agriculture or private investment in agriculture),

non-farm business, and residential plots and buildings. A detailed study of private investment in

agriculture and allied activities among rural households will demonstrate clear trends and inferences

that will be useful to inform related policy decisions. Given that agriculture and irrigation are state-

specific subjects, this analysis may also carry implications for financial allocations toward investment

and subsidy to encourage private investment and accelerate agricultural growth.

The paper is structured as follows: section 2 explains the database and methodological approach; section

3 delineates the structure of FCE and FCE-FB, its share in GCE, and trends thereof. Section 3 also

describes the sources of financing of farm business investment in rural households. Section 4 provides

an econometric estimation of the determinants of private investment in agriculture and their relationship

with public investment and input subsidy in agriculture. Finally, section 5 summarizes the main

findings.

9

2. DATABASE AND METHODOLOGICAL APPROACH

The NSSO data was analysed for agricultural year (AY) 1981–82 (37th NSS Round), AY 1991–92

(48th NSS Round), AY 2002–03 (59th NSS Round), and AY 2012–13 (70th NSS Round) across 17

major states, tracking changes in various aspects of investment among rural households over a period

of three decades. This includes private investment by small as well as large farmers, but not any private

corporate investment. The survey was spread over 3,000 villages across India, and data on a wide array

of characteristics are collected from more than 30,000 households. In each year, the same households

are visited twice. The first visit was done in January–July, and the second visit in August–December.

The NSS collects data on private household investment for both rural and urban sectors. We restrict this

study only to the rural sector, as agricultural activities are mostly concentrated in rural areas.

Approximately 80 percent of rural households have farming as their main occupation. Hence, all

discussion about investment henceforth is for rural households, even when unspecified. Three states

that came into existence in 2001—Uttarakhand, Jharkhand, and Chhattisgarh—have been merged with

their respective parent states (Uttar Pradesh, Bihar, and Madhya Pradesh) to maintain comparability of

data across time. The gross and fixed capital expenditure data are interpolated based on the past rate of

growth in each state.

To analyse the relationship between private and public capital formation in agriculture and irrigation,

we use the two-stage least squares (2SLS) instrumental variable regression method with state fixed

effects. The 2SLS model is used to counter the problem of endogeneity of independent variables; that

is, when one or more of the independent variables are determined within the system of equations. We

find the problem of endogeneity in the public investment variable. Accordingly, per hectare private

investment in agriculture and allied activities is taken to be a function of per hectare public investment

in agriculture and irrigation, along with terms of trade, per hectare institutional credit, per hectare input

subsidy (on account of irrigation, fertilizer, power, and credit), gross cropped area, rainfall, and public

infrastructure (represented by road density). This relationship is defined in Equation 1 below:

10

PVTINVPHit = αit + β1PUBINVPHit + γX’it + t + uit , (1)

where PVTINVPHit is per hectare private investment in agriculture for the ith state in time period t;

PUBINVPHit is the per hectare public investment in agriculture and allied activities plus irrigation for

the ith state in time period t; and X’it collectively represents all the other determining variables

mentioned above. A trend variable t has been specified in the equation to capture the effects of variables

other than those explicitly mentioned in the equation. Since per hectare public investment in agriculture-

irrigation is an endogenous variable, it is instrumented for by a vector of variables Y’, which includes

per capita income, net loans per capita, net market borrowings per capita, per capita revenue deficit of

the government, and irrigation intensity (ratio of net irrigated area to net sown area)2. Equation 2

represents the determination of the endogenous variable through its proxies:

PUBINVPHit = δit +ηY’it + λit , (2)

where PUBINVPHit is per hectare public investment in agriculture and irrigation for the ith state in time

period t; and Y’ is the vector representing the proxy variables for per hectare public investment

mentioned above. Each of these variables has been normalised by the population of the respective state.

γ and η denote the coefficient vectors while β1 captures the impact of public investment in agriculture

and irrigation on private investment in agriculture. αi is the unobserved time-invariant individual effect

and uit is the disturbance in estimation. The robustness of the estimates is tested using various diagnostic

tests. The analysis has been done for the period from 1981–82 to 2012–13 for all 17 states taken

together, which represents the all-India picture, as well as for the three groups of states—high-income

states (HISs), middle-income states (MISs), and low-income states (LIS)—categorised on the basis of

their per capita incomes from 2000–01 to 2013–14. State wise series on public expenditure variables

are extracted from the Finance Accounts, Comptroller and Auditor General, Government of India. Data

on road density, electricity consumption in agriculture, and per capita income are noted from

Agriculture Statistics at a Glance, Ministry of Agriculture and Farmers’ Welfare, Government of India.

2 With the only exception of irrigation intensity – a demand side variable that influences the government to make

investments in agriculture – all other variables considered as instruments are from the supply side.

11

These series along with estimates on gross and fixed capital expenditure are converted into real price at

base 2004–05 using gross state domestic product (GSDP) deflators. The time series on the estimates of

input subsidy on account of fertilizer, irrigation, electricity, and credit at 2004-05 price is sourced from

Bathla et al. (2017).

12

3. STRUCTURE AND TRENDS IN FIXED CAPITAL EXPENDITURE AMONG

RURAL HOUSEHOLDS

At the outset, it is important to highlight the concept of capital expenditure and its structure as explained

by NSS. Capital expenditure is synonymously used with capital formation to include a household’s

expenses on new purchases, improvement, major repairs, and alterations. The survey classifies rural

household investment into gross capital expenditure (GCE) which comprises Fixed Capital Expenditure

(FCE), Expenditure on Purchase of Land (EPL), Normal Repairs & Maintenance Expenditure (NRME),

each divided into three sub categories – residential plots/land and buildings, farm business and non-

farm business (see Annex Figure A.1 for details). FCE in farm business (FCE-FB) represents private

capital formation in agriculture and encompasses eight components, the major ones being land

improvement, livestock, irrigation, transport, and machinery & implements. The nonfarm business

covers expenditures on manufacturing and other activities, land improvement, purchase of equipment

and other accessories. The survey considers expenditure on land improvement and land rights as

components of fixed capital formation and excludes value of land in estimation of capital stock.

Table 1 presents average expenditure per household on each of the three broad heads of FCE—

residential plots and buildings (RPBs), farm business (FB), and nonfarm business (non-FB)—during

1981–82, 1991–92, 2002–03 and 2012–13 at constant 2004–5 prices. As expected, the FCE per rural

household shows a more than three-fold increase at the all-India level. Within FCE, residential plots

and buildings increased four times, whereas farm and nonfarm businesses increased only 2.2 and 2.7

times, respectively, during this entire period.

Table 1: Average values for broad heads of fixed capital expenditure (in rupees/rural household at 2004–05 prices), 1981–82 to 2012–13

Residential plots and buildings

Farm business Nonfarm business

Fixed capital expenditure

1981–82 1,159 753 223 2,133 1991–92 1,477 815 194 2,484 2002–3 1,967 670 434 3,070 2012–13 4,751 1,631 611 6,993

Source: NSS-AIDIS 37th, 48th, 59th, and 70th Rounds.

13

It is also important to note the shift in relative importance of the three broad heads of expenditure within

FCE at the all-India level over time. As shown in Figure 1, the share of investment in farm business in

total FCE has hovered around 10 percent in total expenditure, revealing a larger and growing share of

residential plots and buildings.

Figure 1: Relative importance of the broad heads of fixed capital expenditure among rural households, 1981–82 to 2012–13

Source: NSS-AIDIS 37th, 48th, 59th and 70th Rounds.

Table 2 shows trends in state-wise shares of the broad heads of FCE—residential plots and buildings,

farm business, and nonfarm business among rural households. As a general trend, the share of

residential plots and buildings has remained much higher than the share of farm and nonfarm business

in total fixed capital expenditure. This share has increased steadily at the all-India level from 54.3

percent in 1981–82 to 64.1 percent in 2002–03, and finally to 67.9 percent in 2012–13. FCE-FB has

decreased from 35.3 percent in 1981–82 to 23.3 percent in 2012–13. The share of nonfarm business

fluctuated during this three-decade long period, declining from 10.5 percent in 1981–82 to 7.8 percent

in 1991–92, then rising to 14.1 percent in 2002–03, before falling back to 8.7 percent in 2012–13. Only

a very few states experienced an increase in the share of FB in FCE incurred by rural households.

Gujarat, after seeing an initial decline in share of farm investment from 56.9 percent to 31.5 percent

during the first decade of the period under study, witnessed significant increases in this share over the

next two decades, reaching 50.9 percent in 2012–13.

01020304050607080

1981-82 1991-92 2002-03 2012-13

Perc

ent s

hare

Residential plots and buildings Farm business Non-farm business

14

In Himachal Pradesh, the share of farm business rose from 15.6 percent in 1991–92 to 21.4 percent in

2012–13; however, this is lower than its initial share of 23.3 percent in 1981–82. Karnataka saw a steep

rise in capital expenditure on farm business from 39 percent in 1981–82 to 54.1 percent in 1991–92,

after which it declined precipitously to 22.9 percent in 2012–13. In Uttar Pradesh, the share of farm

business in fixed capital expenditure per rural household declined slightly from 34.5 percent in 2002–

03 to 30.4 percent in 2012–13. In Bihar, it increased significantly from 21.2 percent in 1981–82 to 33

percent in 1991–92. However, thereafter it declined to 12.7 percent in 2002–03 and only slightly

increased to 13.8 percent in 2012–13. Similarly, in Rajasthan, after increasing from 34.8 percent in

1981–82 to 45.8 percent in 1991–92, it declined to 40.5 percent and ultimately 27 percent in 2012–13.

Punjab has seen a more recent improvement in the share of farm business in capital expenditure, from

37.1 percent in 2002–03 to 47.2 percent in 2012–13; but this is lower than the share (52 percent) it had

initially witnessed in 1981–82. Madhya Pradesh has also followed a similar trend: increasing from 49.4

percent (1981–82) to 58.1 percent (1991–92) and then declining steeply to 25.6 percent (2002–03) and

then increasing to 41.2 percent (2012–13) over this three-decade period.

It can be observed from the analysis that investment in residential plots/land, buildings, and construction

has increased over the given period mainly at the cost of investment in farm business. This represents a

shift in the preferences of farmers away from investment in farming activities to investment in

residential land and buildings. It may be indicative of the rising nuclear family trend resulting in higher

fragmentation of land and therefore, more demand for residential land and buildings. The relative

decline in farm investment is more evident in the relatively richer and more developed states, such as

Haryana, Maharashtra, and Tamil Nadu, where the trend toward nuclearization is likely to be more

pronounced, lending further support to the theory of a shift in preferences toward nuclear families and

the resulting fragmentation of landholdings.

15

Table 2: Trends in share of broad heads of fixed capital expenditure (FCE) per rural household in major states (%), 1981–82 to 2012–13

Note: RPB is residential plots and buildings; FB is farm business; NFB is nonfarm business. Source: NSS-AIDIS 37th, 48th, 59th and 70th Rounds.

In terms of the rate of growth in each head, we find in Annex Table A.1 that investment in residential

plots and buildings grew at a slow pace during the 1980s and 1990s, ranging between 2.5 and 3.0 percent

per year at the all-India level. However, investment in this component grew remarkably at over 9 percent

per year in the first decade of the 2000s. In the 1980s, growth in farm business investment remained

negative in most of the major states in India, barring a few, such as Madhya Pradesh, where growth in

farm business capital formation was positive and rather high at 9.1 percent per year. Karnataka (6.2

percent), Himachal Pradesh (4.7 percent) and Rajasthan (4 percent) were some of the other exceptions

that registered reasonable growth in farm business investment during this period. Annual growth in

farm business investment fell at the all-India level from a meagre 0.8 percent in the 1980s to a negative

1.9 percent during the 1990s. Interestingly, Madhya Pradesh, an exceptionally good performer in terms

of growth in farm business investments during the 1980s, showed exceptionally poor performance

1981–82 1991–92 2002–3 2012–13

RPB FB NFB RPB FB NFB RPB FB NFB RPB FB NFB Andhra Pradesh 47.6 40.6 11.6 58.6 37.1 2.0 52.3 19.0 28.7 64.2 20.2 15.6

Assam 61.3 25.2 13.3 77.3 11.7 10.7 75.7 17.7 6.8 68.7 22.2 9.1

Bihar 70.2 21.2 8.1 57.5 33.0 9.5 72.7 12.7 14.6 75.4 13.8 10.8

Gujarat 34.8 56.9 8.1 54.5 31.5 14.1 50.8 37.3 11.9 42.6 50.9 6.5

Haryana 39.0 53.4 7.5 83.4 15.5 1.1 36.7 27.4 35.9 85.6 11.9 2.5 Himachal Pradesh 74.7 23.3 1.8 68.9 15.6 15.5 81.5 8.3 10.3 74.8 21.4 3.8

Jammu and Kashmir 65.4 13.8 20.6 62.5 21.6 15.9 78.0 12.2 9.7 84.3 12.5 3.2

Karnataka 52.6 39.0 8.3 39.3 54.1 6.5 69.7 23.0 7.3 68.8 22.9 8.3

Kerala 79.6 11.9 8.3 71.2 9.2 19.7 81.6 6.6 11.8 79.4 9.3 11.3

Maharashtra 35.8 48.8 15.1 48.4 45.2 6.5 58.7 29.7 11.6 59.8 32.5 7.7 Madhya Pradesh 40.5 49.4 10.1 38.6 58.1 3.2 60.3 25.6 14.1 54.0 41.2 4.8

Odisha 67.1 19.2 13.3 76.9 15.5 7.3 53.5 13.5 33.0 76.4 16.1 7.5

Punjab 31.9 52.0 16.0 51.5 46.2 2.3 51.7 37.1 11.2 42.5 47.2 10.4

Rajasthan 58.3 34.8 6.8 50.3 45.8 3.8 52.1 40.5 7.5 61.4 27.0 11.6

Tamil Nadu 59.1 32.5 8.4 67.7 26.3 5.9 76.8 14.3 9.0 85.1 5.7 9.1

Uttar Pradesh 54.1 34.5 11.2 64.4 27.3 8.1 63.6 31.0 5.4 59.8 30.4 9.8

West Bengal 74.8 17.1 8.0 66.4 18.3 15.1 80.3 5.2 14.5 88.7 5.2 6.1

India 54.3 35.3 10.5 59.5 32.8 7.8 64.1 21.8 14.1 67.9 23.3 8.7

16

during the 1990s with a steeply negative average annual growth rate during this decade. Karnataka, too,

slipped from doing reasonably well in farm business investment in the 1980s to an average annual

growth rate of a negative 11.1 percent in the next decade. Odisha transformed its growth in farm

business investment to a high of 9.4 percent per year; so did Haryana (6.4 percent) and Gujarat (4.6

percent), whereas Himachal Pradesh continued to perform steadily.

The decade from 2002–03 to 2012–13 witnessed the most remarkable improvement in investment in

agriculture at the all-India level as well as at the individual state level. Farm business expenditure grew

at an average rate of 9.3 percent per year during this period. Most of the major states experienced high

annual rates of growth between 7 and 20 percent during this period. The only exceptions were Uttar

Pradesh, Odisha, Tamil Nadu, and Haryana, which witnessed very low or even negative growth in farm

investment during the 2000s. Madhya Pradesh recorded an average growth of a staggering 20 percent

per year, Karnataka 15.3 percent per year, and Bihar 14.8 percent per year during this decade. In contrast

to FB, the non-farm business expenditure fell during the 1980s, picking up again during the 1990s at a

significant rate of 8.4 percent per year, and then slowing down over the next decade during the 2000s

to a modest growth at 3.5 percent per year. While growth in nonfarm business has shown an upward

trend at the state level from largely negative growth rates across some states to significantly positive

growth in most states from 2002–03 to 2012–13, large interstate variations remain. Lastly, FCE grew

at a much faster rate at 8.6 percent per annum from 2002–03 to 2012–13 compared with its growth in

the previous two decades—over four times faster.

Notably, during the recent decade, the emphasis has gradually shifted away from nonfarm business and

even residential plots and buildings to farm business, which might be explained by increased public

investment in agriculture, irrigation, and various rural programmes, provision of input subsidies, easy

flow of credit, and improved terms of trade. Some of these driving forces have enabled a remarkable

rate of growth in private investment in the low-income and developing states (Chand and Parappurathu

2012; Bathla 2016). The absolute expenditure per household in farm business also provides interesting

insights. A comparison across the four distinct periods of study shows real investment in agriculture

17

initially fluctuated for two decades—it went up from Rs. (Indian rupees) 753 in 1981–82 to Rs. 815 in

1991–92, then fell to Rs. 670 in 2002–03. But it witnessed a remarkable revival in the next decade and

increased over two-folds to Rs. 1,631 per rural household in 2012–13.

The disquieting feature, however, is a falling share of FCE-FB in GCE as shown in Figure 2. It has gone

down from 20.2 percent in 1980–81 to 15.6 percent in 2002–03, with a slight improvement to 18.8

percent in 2012–13, at the all-India level. Among the states, Gujarat and Punjab performed well

somewhat consistently in terms of this ratio, although both witnessed a sharp decline in the share of

FCFA in GCE between 1981–82 and 2012–13. Madhya Pradesh showed exceptional performance in

1991–92 when this ratio touched 40 percent, but soon after it declined to 12.5 percent in 2002–03 and

recovered somewhat to reach 26.7 percent in 2012–13. Uttar Pradesh made considerable progress in

consistently improving the share of real investment in agriculture per rural household as a share of gross

capital expenditure (like few others, such as Odisha and Jammu & Kashmir, though these were at

relatively much lower levels compared with Uttar Pradesh). Haryana, Tamil Nadu, and West Bengal

experienced a significant decline in the share of fixed capital formation in agriculture in gross capital

expenditure during this entire period. Andhra Pradesh and Bihar also experienced a decline in the ratio

of FCFA to GCE but to a lesser extent. This data bolsters the fact that an increased expenditure on

residential plots and buildings has been undertaken at the cost of fixed capital formation in agriculture,

which has implications for the future of agricultural growth.

18

Figure 2: Share of fixed capital expenditure in FB in gross capital expenditure (GCE) per rural household (%), 1981–82 to 2012–13

Source: NSS-AIDIS 37th, 48th, 59th and 70th Rounds.

The survey also provides interesting insights on changes in the composition of FCE in farm business.

As shown in Figure 3, the most important investment items in agriculture for rural households turn out

to be machinery, implements, and transport equipment; followed by irrigation resources; livestock; and

land improvement works. Together, these account for more than 85 percent of FCE in farm business

among rural households.

Figure 3: Trends in composition of fixed capital expenditure in farm business among rural households, 1981–82 to 2012–13

Source: NSS-AIDIS 37th, 48th, 59th and 70th Rounds.

0

10

20

30

40

50

Perc

ent s

hare

1981-82 1991-92 2002-03 2012-13

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1981-82 1991-92 2002-03 2012-13

Agricultural Machinery,Implements, TransportEquipment

Livestock

Irrigation Resources

Land Improvements

Farm Buildings

Orchards and Plantations

19

Livestock was included only in the last two rounds due to which agricultural machinery, implements,

and transport equipment shows a declining trend. Investment in irrigation resources gained substantially

in importance during the 1990s and the 2000s. Its share in capital formation in agriculture grew from

25.4 percent in 1981–82 to 31.8 percent in 1991–92 and 33.1 percent in 2002–03. Thereafter, it declined

steeply to 22.6 percent in 2012–13. Investments in orchards and plantations has declined consistently

during this entire period. The high weightage of agricultural machinery, implements, and transport

equipment; irrigation resources; and land improvement works shows the high propensity of farmers

toward asset creation. It is interesting to note that livestock is included as an asset from 2002-03 and

has captured a considerable share of fixed capital expenditure in farm business from 13.8 percent in

2002-03 to 23.1 percent in 2012–13. Annex Figure A.2 shows this figure when livestock is excluded

from the FCE in the years 2002-03 and 2012-13 for comparison across years.

Annex Tables A.2 and A.3 show that together the three major items of farm business, along with

livestock (from 2002–03 onwards), accounted for 86 percent of rural household investment in 1981–

82, 92 percent in 1991–92, 89 percent in 2002–03, and 94 percent in 2012–13. At the all-India level,

household investment in livestock (15.1 percent) was most important, closely followed by land

improvement works (13.1 percent), and then agricultural machinery, implements, and transport

equipment together (10.4 percent), orchards and plantations (9.4 percent), and irrigation resources (5.2

percent) during 2012-13.

Between 2002–03 and 2012–13, all major states witnessed positive and significantly high average

annual growth in investment in machinery, implements, and transport equipment, and in livestock.

Kerala (40.3 percent) and Madhya Pradesh (34.5 percent) were the exceptional performers in the former,

and Madhya Pradesh (25.8 percent), Jammu and Kashmir (24.1 percent), and Karnataka (23.8 percent)

were the high performers in the latter. Investment in land improvement works has also grown at

reasonably high average annual rates in most states except a few, such as West Bengal, Kerala, and

Madhya Pradesh. Tamil Nadu showed scant positive growth under this component during the 2000s.

20

Bihar showed the second-highest average annual growth in land improvement works (18.1 percent per

year) as well as in irrigation resources (21.7 percent per year) between 2002–03 and 2012–13, the

highest being recorded in Himachal Pradesh (19.9 percent per year) in land improvement and in

Karnataka (17.7 percent per year) in irrigation resources. Between 2002–03 and 2012–13, Maharashtra

experienced a staggering average annual growth of about 47 percent in orchards and plantations. Assam,

Tamil Nadu, West Bengal, and Jammu and Kashmir are other states where growth in FCE in orchards

and plantations has been impressive (14 to 15 percent per year) (Annex Table A.4).

As discussed above, besides increase in public investment in agriculture and irrigation, easy availability

of credit to farmers at subsidized rates of interest could have been the driving factors behind this surge

in FCE-FB. A closer look at the magnitude of public expenditure in agriculture and irrigation in Figure

4 shows a significant increase across the LISs, MISs, and HISs from the early 2000s. However, a

consistent increase is identified in the less developed agriculturally dependent states, which is an

encouraging policy initiative. An important concern, however, is that capital expenditure, which is

synonymous with investment has not increased in a significant manner. Taking all states together, the

share of capital expenditure in total expenditure on agriculture has increased from 6.45 percent to 9.57

percent and on irrigation from 37.01 percent to 42.26 percent over the period. A significant increase in

capital expenditure on irrigation is visible only in Gujarat, Karnataka, Madhya Pradesh, Maharashtra,

Tamil Nadu, and West Bengal. A lesser share of investment indicates that government expenditure is

skewed toward day-to-day administrative expenditures including subsidies despite an improvement

from 2003-04. Lower share of capital expenditure in total expenditure may also imply mounting

bureaucracy and inefficiency, especially on major and medium canal irrigation systems (Bathla 2016).

21

Figure 4: Public expenditure on agriculture and irrigation (Rs. billion at 2004-05 prices)

Source: NSS-AIDIS 37th, 48th, 59th and 70th Rounds.

In sum, between 2002 and 2013, the rural households overall investment pattern has shifted away from

farm business towards RPB. And within FB, the preference has changed from irrigation towards land

improvement, livestock, machinery-implements and transport (tractors). In contrast, the composition of

public expenditure on agriculture shows a shift toward food storage and warehousing, crop husbandry,

forestry and medium-major irrigation works. The inducement effect of various types of public

investments on the composition of investments undertaken by farmers is a researchable issue. Even in

the case of other developing countries, complementarity between the public and private investments is

studied using investment in agriculture and irrigation in aggregate rather than looking at investments in

these individually (Mogues et al. 2015).

Households’ sources of finance for FCE in Farm Business

We find that nearly 86 percent of expenditure in 2012–13 was financed through credit (that is,

borrowings). Of total borrowings, 63.4 percent was from institutional sources3. The role of credit is

3 Institutional credit refers to borrowings from sources in the formal sector such as banks, financial companies, post

offices, etc., whereas non-institutional credit refers to borrowings from sources in the informal sector, for example, local moneylenders, friends, family, neighbours, etc.

10

30

50

70

90

110

130

150

170

19019

81—

1982

1982

—19

8319

83—

1984

1984

—19

8519

85—

1986

1986

—19

8719

87—

1988

1988

—19

8919

89—

1990

1990

—19

9119

91—

1992

1992

—19

9319

93—

1994

1994

—19

9519

95—

1996

1996

—19

9719

97—

1998

1998

—19

9919

99—

2000

2000

—20

0120

01—

2002

2002

—20

0320

03—

2004

2004

—20

0520

05—

2006

2006

—20

0720

07—

2008

2008

—20

0920

09—

2010

2010

—20

1120

11—

2012

2012

—20

1320

13—

2014

LIS - Agriculture MIS - Agricultture HIS - Agriculture

LIS - Irrigation MIS - Irrigation HIS - Irrigation

22

critical as over 80 percent of landholdings are owned by marginal and small farmers who have very

little earnings and hence little savings and relatively much poorer access to credit. In fact credit acts as

a facilitator in allowing them to engage in agriculture. Availability of credit allows farmers access to

better seeds, fertilizers, insecticides, pesticides, modern agricultural machinery, and improved

postharvest facilities.

Annex Table A.5 shows the state-wise share of investment in farm business financed through

borrowings and the share of such borrowings from institutional sources. Farmers in almost all states

depended on credit for more than 65 percent of their farm investments in 2012–13. Andhra Pradesh,

Kerala, Tamil Nadu, Punjab, and Maharashtra have farmers financing more than 90 percent of their

fixed capital expenditure in farm business through borrowings, both from institutional and

noninstitutional sources. While this is a happy situation, it is also important to segregate the share of

total borrowings that have come from institutional sources. The national average of 63.4 percent is

moderately satisfactory. The states that are doing relatively better on this count are Kerala, Gujarat,

Punjab, Maharashtra, and Odisha. Farmers in the relatively more developed states seem to have better

access to institutional credit, and it is also possible that they better use this access. On the other hand,

Jammu and Kashmir, Andhra Pradesh, Tamil Nadu, Bihar, Rajasthan, Madhya Pradesh, and Karnataka

have the lowest shares of institutional borrowings in total borrowings. Efforts should be made to

improve the outreach of formal financial institutions for farm loans, particularly in the states that lag in

the shares of institutional sources in their total borrowings. Marginal and small farmers, while most in

need of credit, typically depend on noninstitutional sources for most of their credit needs in almost all

states.

23

4. RESULTS FROM ECONOMETRIC ANALYSIS

This section focuses on an empirical analysis of the relationship between private and public investments

in agriculture and irrigation at the state level. As mentioned earlier, private capital formation in

agriculture is mainly done by the farm households whose investment decisions are influenced by

multiple factors. Using economic theory of investment, Gandhi (1990) developed an empirical model

from the period 1951-52 to 1980-81 and found a significant positive impact of government investment

in agriculture on private investment. Extending this model to the post-1980 period, when public and

private investments in agriculture had started trending divergently, Gandhi (1996) found that between

1952–53 and 1992–93, rural savings and cooperative credit to agriculture were the strongest

determinants of private investment, followed by high-yielding variety coverage, agricultural wages, and

commercial bank credit. Binswanger et al. (1993) observed that education, infrastructure availability,

and rural banks played an overwhelming role in influencing farmers’ investment in agriculture while

Mishra and Hazell (1996) found favourable terms of trade and technology (measured as percent area

under high-yielding varieties) to be the key determinants. Dhawan (1998), Gulati and Bathla (2001),

and Chand and Kumar (2004) endorsed these findings and argued the role of both price and non-price

factors including public investment to induce private investment in agriculture. Chand and Kumar

(2004) also stressed on the growing number of holdings causing farmers to spend on assets. Roy and

Pal (2001) extended the analysis by taking input subsidy as a variable to explain private investment in

agriculture. Similarly, Bathla (2014) took input subsidy and the backward linkage of agriculture with

the food-beverage industry as the added factors that influence private investment.

The empirical results on input subsidy diverge whereas a consensus appeared in the literature on a

positive impact of public investment, terms of trade, technology, credit, and literacy on private

investment. The ‘complementarity hypothesis’ was challenged only in Mishra and Chand (1995) which

in subsequent study (Chand 2000) was accepted based on a broader series on public investment ‘in’ and

‘for’ agriculture, both at the national and state levels with appropriate lags. Only in one study – Bisaliah

and Dev (2010) – the analysis on private investment is carried out at a further disaggregate farm level

24

from 1994-95 to 2007-08 wherein the authors found that for marginal, small, and large farmers, land,

availability of credit, and literacy hold significance. In contrast, for semi-medium and medium farmers,

only land and credit had a positive influence on private investment in agriculture.

The empirical analysis that follows is, therefore, an effort to explore the impact of some of these factors

on private capital formation in agriculture at the all-India as well as the state level during the three-

decade period from 1981–82 to 2013–14. At the national level, this period has experienced considerable

highs and lows in both public and private capital formation and hence agricultural growth. To reiterate,

the 1980s began with higher rates of investment on public and private accounts and experienced a

slowdown in both by the end of the decade which continued till the 1990s, and rebounded from the

early 2000s. The state level picture is, however, very divergent, showing a reasonable rate of growth in

the magnitude of investments and agricultural income in some while an agrarian crisis in others between

the 1980s and the 2000s.

A two-stage least squares (2SLS) (instrumental variables) method with state fixed effects is used to

examine the determinants of private investment in agriculture and test the proposed hypotheses relating

to the impact of public investment in agriculture and irrigation and input subsidy at the state level. The

explanatory variables, namely, public investment in agriculture and irrigation were found to be

endogenous in the panel data. Appropriate instruments as suggested in the literature on public

expenditure were used to handle the problem of endogeneity. Accordingly, overidentification of

equations (Sargan and Basmann tests), endogeneity (Durbin chi-square and Wu-Hausman F tests) and

other diagnostics were done to check the robustness of the functional form used. Two explanatory

variables – institutional credit and electricity consumption are dropped due to their high correlation with

input subsidy and road density. Table 3 presents the results of the regression analysis for all-India, as

well as for high-income, middle-income, and low-income states.

25

Table 3: Impact of public investment in agriculture and irrigation on private investment in agriculture: Regression results, 1981–82 to 2013–14 (IV2SLS) Dependent variable: Private investment in agriculture and allied activities per ha

Variables All states HISs MISs LISs

Coefficient SE Elasticity Coefficient SE Elasticity Coefficient SE Elasticity Coefficient SE Elasticity Public investment in agriculture & irrigation per ha 0.41*** 0.07 0.54 0.70*** 0.15 0.64 0.02 0.03 0.04 0.27*** 0.04 0.51

Terms of trade 1.93* 1.16 0.20 1.27 2.87 0.08 0.20 0.88 0.03 0.84 0.74 0.14

Input subsidy per ha 0.01 0.02 0.04 -0.10** 0.05 -0.25 0.04** 0.02 0.19 0.08** 0.04 0.23

Gross cropped area -0.03 0.03 -0.28 -0.49*** 0.13 -2.6 0.01 0.01 0.13 0.03 0.03 0.73 Rainfall -0.01 0.09 -0.01 -0.37 0.26 -0.29 -0.01 0.08 -0.02 0.03 0.05 0.06

Road density 0.30*** 0.09 0.31 0.82*** 0.18 0.68 0.02 0.06 0.03 -0.02 0.06 -0.03

Trend 6.4 6.24 0.11 38.57*** 13.94 0.43 9.33** 3.7 0.29 6.8 5.21 0.19

Constant 98.71 302.21

3507.98 958.41 173.79 146.27 -517.91 340.04

No. of observations 559 231 163 165

R2 (within) 0.193 0.285 0.542 0.5448

R2 (between) 0.146 0.616 0.599 0.5381

R2 (overall) 0.211 0.296 0.511 0.5518

sigma_u 664 2867 172 305

sigma_e 666 911 175 307

Rho 0.5 0.91 0.49 0.5 F-statistic 27.48*** 9.34*** 10.68*** 19.34***

Tests of endogeneity (H0: variables are exogenous)

Durbin (score) 37.98*** 23.76*** 0.091 14.73***

Wu-Hausman 38.93*** 24.76*** 0.084 14.90***

Tests of overidentifying Restrictions: (H0: instrument set is valid, and model is correctly specified)

Sargan (score) 46.42*** 21.05*** 48.34*** 9.412

Basmann 48.09*** 21.36*** 61.98*** 9.014 Note: SE is standard error; terms of trade (TOT) is based on a three-year moving average; elasticity is estimated taking mean of the variables; state dummies specified to capture state-level unobservable specific effects. *,**, *** indicate statistical significance at the 10 percent, 5 percent, and 1 percent confidence levels.

26

We find that the ‘crowding in’ effect of public investment on private investment in agriculture is

confirmed but not all around. Per hectare public investment in agriculture and irrigation has a positive

and statistically significant impact on per hectare private investment in agriculture at the all-India level

as well as in the high-income and low-income states. At the national level, with a 10 percent increase

in public investment, private agricultural investment per hectare increased by an average of 5.4 percent.

In the high-income states, private agricultural investment increased by 6.4 percent, and in the low-

income states, by about 5.1 percent with a 10 percentage point increase in per hectare public investment

in agriculture and irrigation. Surprisingly, in the middle-income states, public investment has a

negligible and statistically insignificant impact. Same is the case with another public infrastructure

variable, road density, which shows statistically significant impact on private investment only at the

national level and in high-income states, perhaps due to adequate development of road network.

Another debatable aspect is whether provision of input subsidy by the government incentivizes farmers

to scale up investments. In an attempt to reinvigorate agricultural growth, the government has

enormously increased expenditure on provision of subsidised farm inputs since 2003-04. From the

analysis, per hectare input subsidy is found to have a strong bearing on private investment in agriculture

in the relatively poorer states. While in the middle- and low-income states, a 10 percent increase in per

hectare input subsidy increases per hectare private investment in agriculture by an average of 1.9 percent

and 2.3 percent, respectively, it actually decreases private investment by about 2.5 percent in high-

income states. It can appear counterintuitive at first glance, but it is possible that in high-income states,

which are also likely to have relatively richer farmers, subsidies on inputs can dissuade them from

spending their own money, possibly due to lower marginal returns. Nonetheless, the elasticity estimates

of input subsidy are way below those of public investment in agriculture and irrigation.

Other variables that are included to examine the investment behavior of farmers reveal that the terms

of trade (price incentive) has a statistically significant and positive impact at the all-India level having

value of elasticity at 0.20. This becomes clear, as farmers, like other rational economic entities, are

responsive to the movement of commodity prices for getting higher returns. The impact of terms of

27

trade, however, turns out to be positive but statistically insignificant at the disaggregate state level. This

can be explained by considerable price spread in the supply chain, inadequate development of agri-

markets and infrastructure, and market failures such as negligible support to farmers in a situation of

collapse in commodity prices, hoarding by wholesalers, and formation of cartels. Land, another variable

taken to influence investment decisions, is found to act as a constraint in high-income states, rainfall

tends to be insignificant, and the trend variable, taken to capture the effect of other factors, is positive

and significant only in high- and middle-income states.

Broadly, the results obtained are in accordance with the literature, mostly carried out at the all-India

level, that private investment in agriculture is significantly determined by public expenditure on

agriculture and irrigation and on input subsidies, as well as by the terms of trade. However, the impact

of public investment is found to be larger than that of input subsidy and hence should be accorded

appropriate fiscal space. The state level analysis provides mixed results, indicating a need for targeting

public investment to augment its influence on private investment. The impact of institutional credit on

private investment in agriculture couldn’t be captured in the empirical analysis due to the problem of

multicollinearity. But the descriptive statistics in each state underscore its growing usefulness as a

source of investment finance.

28

5. SUMMARY AND CONCLUSIONS

We have examined the behaviour of private investment among rural households across 17 major states

for a period of three decades from 1981–82 to 2012–13 with an aim to empirically examine its

relationship with public investment in agriculture and irrigation and spending on input subsidies. For

this purpose, data on fixed capital expenditure (FCE) by rural households from the decennial NSS All-

India Debt and Investment Survey (AIDIS) (schedule 18.2) for 1981–82, 1991–92, 2002–03, and 2012–

13 is used and interpolated. The composition of expenditure and sources of borrowings of households

are also studied. The FCE on different types of investments is juxtaposed with the pattern of public

expenditures on agriculture and irrigation and its various segments. The analysis is done for the period

between 1981-82 and 2013-14 at real prices with base 2004-05.

The analysis shows a perceptible increase in gross as well as fixed capital expenditure (FCE) over the

study period. The largest share (nearly 68 percent) is of residential plots and buildings (RPB), which

has also shown a sizeable increase of over four times in its real value from Rs. 1,159 in 1981–82 to Rs.

4,751 in 2012–13 per household over this period. Large investments in RPB by farmers in almost every

state is indicative of growing urbanization, shift of households toward nuclear families, and

fragmentation of landholdings. A greater preference for land assets among farmers may also be

explained by expectation of higher returns compared to returns from farming. Increasing land

acquisition by the government for infrastructural development might have also contributed to making

landed property appear more lucrative. Importantly, a much higher share of expenditure on RPB in total

FCE is observed in the agriculturally advanced states and has come at the expense of investment in

agriculture.

The magnitude of private investment in agriculture and allied activities gained momentum over the

decade from 2002–03 to 2012–13 compared to the 1980s and the 1990s and grew at an average annual

rate of 9.3 percent. The composition has changed towards agricultural machinery, implements, transport

(mainly tractors), irrigation resources, livestock, and land improvement works, these occupying 85 to

29

95 percent share in FCE. With a few exceptions such as Bihar, West Bengal, Kerala, Madhya Pradesh,

and Chhattisgarh, other states have witnessed significant growth in investment in these heads of

expenditure. During 2012-13, nearly 86 percent of fixed capital formation in agriculture was financed

through credit, of which 63.4 percent was borrowed from institutional sources. In almost all the states,

farmers depended on credit to varying degrees for making farm investments, underscoring not only the

expansion of credit but also its outreach across states and farm sizes.

To what extent do farmers’ investment decisions get influenced by public expenditure at the

disaggregate state level is studied empirically based on the instrumental variables two-stage least

squares method with state fixed effects. The econometric exercise is undertaken for the period from

1981–82 to 2013–14, at the all-India level as well as for high-income, middle-income, and low-income

states. Broad findings reveal private investment in agriculture and allied activities to be positively and

significantly influenced by public investment in agriculture and irrigation at the all-India level, and in

high-income and low-income states. The result holds for input subsidy as well, which seems to act as

an important incentive for farmers in the middle-income and low-income states. The other important

determinant of farmers’ investment decisions is terms of trade, which is positive across the states but

statistically significant only at the national level. Land, as another determinant, acts as a constraint,

implying a need to improve land productivity and ensure remunerative prices to farmers. The analysis

corroborates the ‘inducement effect’ of public expenditure on private investment in selected states but

the mean elasticity of public investment is found to be much higher (> 0.50) than that of input subsidy

at around 0.20. Geographical targeting of input subsidies and public investments in agriculture and

irrigation with a focus on the type of investment that would encourage farmers to invest in agriculture

is imperative.

30

REFERENCES

Bahal, G., Raissi, M., & Tulin, V. 2018. Crowding-out or Crowding-in? Public and Private

Investment in India. World Development, 109, 323-333. Bathla, S. (2014). Public and private capital formation and agricultural growth in India: State level

analysis of inter-linkages during pre-and post-reform periods. Agricultural Economics Research Review, 27(1).

———. (2016). Investments and subsidies in Indian agriculture: Recent trends and implications. In

Haque, T. (Ed.), Agrarian distress in India: causes and remedies. Delhi: Concept Publishing Company.

Bathla S. 2017 Public Investment in Agriculture and Growth: An Analysis of Relationship in the

Indian Context, Chapter 2 in Bathla, S., and A. Dubey, eds. 2017 Changing Contours of Indian Agriculture: Investment, Income and Nonfarm Employment. Singapore: Springer Nature.

Bathla, S., Joshi, P. K., & Kumar, A. (2017). Revisiting investments and subsidies to achieve higher

and equitable income and poverty alleviation in rural areas across Indian states. Report Unpublished), New Delhi: IFPRI.

Bathla, S., & Kumari, Y. (2017). Investment behaviour of farmers across Indian states:

Determinants and impact on agriculture income. In S. Bathla and A. Dubey (Eds.), Changing contours of Indian agriculture-investment, income and non-farm employment. Singapore: Springer Nature.

Binswanger, H. P., Khandker, S. R., & Rosenzweig, M. R. (1993). How infrastructure and financial institutions affect agricultural output and investment in India. Journal of Development Economics, 41(2), 337–366.

Bisaliah, S., & Mahendra, D. S. (2010). Private Capital Formation in Indian agriculture: An Analysis

of Farm Level Data. Consultancy Report to FAO, Rome. Delhi: Commission for Agricultural Costs and Prices.

Bisaliah, S., Mahendra, D., & Syed, S. (2013). Investment in Indian agriculture: Macro and micro

evidences. New Delhi: Academic Foundation. Chand, R. (2000). Emerging Trends and Regional Variations in Agricultural Investments and Their

Implications for Growth and Equity. Policy Paper 11. New Delhi: National Centre for Agricultural Economics and Policy Research.

———. (2001). Emerging trends and issues in public and private investments in India agriculture: A

statewise analysis. Indian Journal of Agricultural Economics, 56(2), 161. Chand, R., & Kumar, P. (2004). Determinants of capital formation and agriculture growth: Some new

explorations. Economic and Political Weekly, 39(52), 5611–5616. Chand, R., & Parappurathu, S. (2012). Temporal and spatial variations in agricultural growth and its

determinants. Economic and Political Weekly, 47(26/27), 55–64. Dantwala, M. L. (1986). Strategies of Agricultural Development since Independence. Indian

Agricultural Development since Independence: A Collection of Essays.

31

Dhawan, B. D. (1998). Studies in agricultural investments and rural savings. New Delhi: Commonwealth Publishers.

Dhawan, B. D., & Yadav, S. S. (1995). Private fixed capital formation in agriculture: Some aspects of

Indian farmers' investment behaviour. Economic and Political Weekly, 30(39), A103–A109. ———. (1997). Public investment in Indian agriculture: trends and determinants. Economic and

Political Weekly, 32(14), 710–714. Fan, S. (Ed.). (2008). Public expenditures, growth, and poverty: Lessons from developing countries.

Baltimore, MD: International Food Policy Research Institute and Johns Hopkins University Press.

Fan, S., & Hazell, P. B. (1999). Are returns to public investment lower in less-favored rural areas?: an

empirical analysis of India (No. 581-2016-39390). Gandhi, V. P. (1990). Investment behaviour in developing countries: The case of agriculture in

India. Food Research Institute Studies, 22(1), 45–82. ———. (1996). Investment behaviour in Indian agriculture. Indian Journal of Agricultural

Economics, 51(4), 543. Greene W.H. (2018) Econometric Analysis, 8th Edition, Pearson Publication. Gulati, A., & Bathla, S. (2001). Capital formation in Indian agriculture: Re-visiting the

debate. Economic and Political Weekly, 36(20), 1697–1708. Gulati, A., & Narayanan, S. (2003). Subsidy syndrome in Indian agriculture. New Delhi: Oxford

University Press. Kannan, E. 2014. “Farm Input Subsidies in India: Access, Efficiency and Equity.” Paper presented at

International Association of Agricultural Economists (IAAE) symposium on “Revisiting Agriculture Policies in the Light of Globalisation Experience: The Indian Context.” MANAGE, Hyderabad, India.

Mishra, S. N., & Chand, R. (1995). Public and private capital formation in Indian agriculture:

Comments on complementarity hypothesis and others. Economic and Political Weekly, 30(25), A64–A79.

Misra, V. N., & Hazell, P. B. (1996). Terms of trade, rural poverty, technology and investment: The

Indian experience, 1952–53 to 1990–91. Economic and Political Weekly, 31(13), A2–A13. Mitra, P. (2006). Has Government Investment Crowded Out Private Investment in India? American Economic Review 96(2), 337–341. Mogues, T., Fan, S., & Benin, S. (2015). Public investments in and for agriculture. The European

Journal of Development Research, 27(3), 337–352 Rao, C. H. H. (1997). Agricultural growth, sustainability and poverty alleviation in India: Recent trends

and major issues of reform. Economic and Political Weekly, 33(29/30), 1943–1948 Rath, N. (1989). Agricultural growth and investment in India. Journal of Indian School of political

economy, 1(1), 64–83.

32

Roy, B. C., & Pal, S. (2001). Incremental capital-output ratio in Indian agriculture. Agricultural Economics Research Review, 14(1), 34–46.

———. (2006). Investment, agricultural productivity and rural poverty in India. Indian Agriculture in

the New Millennium: Changing Perceptions and Development Policy, 2, 367. Sharma, V. (2012). Dismantling Fertilizer Subsidies in India: Some Issues and Concerns for Farm Sector Growth. W.P. No. 2012-09-01. Indian Institute of Management, Ahmedabad. Shetty, S. L. (1990). Investment in agriculture: Brief review of recent trends. Economic and Political

Weekly, 25(7/8), 389–398.

33

Annex Figure A.1: Concepts of fixed capital expenditure and its composition in the NSS-AIDIS (schedule 18.2)

Purchase (new) Addition

Major Repairs and Alterations

Improvement

Capital Expenditure

Fixed Capital Expenditure (FCE)

Expenditure on Purchase of Land

(EPL)

Normal Repairs & Maintenance

Expenditure (NRME)

• Residential plots/land and buildings

• Farm business• Nonfarm business

• Residential plots/land and buildings

• Farm business• Nonfarm business

• Residential plots/land and buildings

• Farm business• Nonfarm business

Concepts of Capital Expenditure (GCE)

Gro

ss C

apita

l Exp

endi

ture

(GC

E)

34

Residential Plots and Buildings•Plots, houses, buildings, and construction works

Farm Business•Land improvement works•Barns and animal sheds•Orchards and plantations•Wells, bore wells, tube wells, field distribution systems, other irrigation resources•Livestock: working/breeding cattle and buffaloes•Livestock: egg-laying ducks and hens•Agricultural machinery and implements•Transport equipment (tractor, trolleys) used for farm business•Others

Nonfarm Business•Improvement of land•Workplace, workshop, manufacturing unit, shop, and other constructions•Nonfarm business equipment and accessories•Transport equipment used for nonfarm business only•Others

35

Annex Figure A.2: Trends in composition of fixed capital expenditure in farm business among rural households, excluding livestock, 1981–82 to 2012–13

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1981-82 1991-92 2002-03 2012-13

3.3 2.1 1.6 1.73.4 1.9 1.8 1.9

74.2

9.64.7

14.8

14.3 6.810.8

25.431.7 38.4

29.3

46.1 45.5 41.751.6 Agricultural machinery,

implements, transport equipment

Irrigation resources

Land improvements

Farm buildings

Orchards and plantations

Others

36

Annex Table A.1: Annual rate of growth in broad heads of fixed capital expenditure per rural

household at 2004–5 prices

1981–82 to 1991–92 1991–92 to 2002–3 2002–3 to 2012–13 RPB FB NFB FCE RPB FB NFB FCE RPB FB NFB FCE

Andhra Pradesh 0.4 -2.5 -11.1 -1.6 4.7 -0.9 28.3 5.9 11.9 10.3 3.1 9.6

Assam -1.3 -10.7 -5.6 -3.6 -0.4 4.0 -4.7 -0.2 6.4 9.8 10.7 7.4 Bihar -8.7 -2.6 -5.3 -6.8 1.7 -6.4 9.2 0.8 14.2 14.8 10.5 13.8 Gujarat 4.6 -5.7 5.7 0.0 2.1 4.6 1.1 2.8 4.8 10.0 0.3 6.6 Haryana 15.6 -5.3 -12.0 7.1 -7.4 6.4 43.0 0.5 18.0 -0.2 -17.0 8.4 Himachal Pradesh 8.0 4.7 35.2 8.9 13.3 4.6 6.9 11.5 -0.1 10.8 -8.8 0.7

Jammu and Kashmir -5.1 -0.3 -7.1 -4.7 11.6 3.2 4.0 9.2 8.2 7.6 -3.9 7.3

Karnataka -0.2 6.2 0.2 2.8 2.6 -11.1 -2.0 -3.1 15.2 15.3 16.7 15.3 Kerala 1.1 -0.4 11.4 2.2 5.5 0.7 -1.1 4.0 7.9 12.0 7.8 8.2 Maharashtra 5.9 1.9 -5.6 2.7 3.2 -2.9 7.3 1.2 9.4 10.2 4.9 9.2 Madhya Pradesh 6.9 9.1 -2.0 7.4 -8.1 -14.0 8.7 -9.5 13.2 20.0 2.8 14.5

Odisha 0.5 -3.0 -6.7 -0.9 7.0 9.4 29.0 10.9 2.5 0.7 -14.7 -1.1 Punjab 0.8 -5.0 -20.8 -3.9 3.0 0.7 20.7 3.0 3.8 8.5 5.0 5.9 Rajasthan -0.3 4.0 -4.5 1.2 1.1 -0.4 7.8 0.8 14.3 7.9 17.4 12.4 Tamil Nadu 5.9 2.2 0.8 4.4 5.0 -2.4 8.2 3.7 10.8 0.1 9.9 9.7 Uttar Pradesh 3.2 -0.9 -1.8 1.4 0.5 1.7 -3.8 0.6 1.8 2.2 8.7 2.4

West Bengal -3.6 -1.8 3.9 -2.5 10.0 -4.8 7.5 8.0 9.5 8.3 -0.6 8.4 India 2.5 0.8 -1.4 1.5 2.9 -1.9 8.4 2.1 9.2 9.3 3.5 8.6

Note: RPB is residential plots and buildings; FB is farm business; NFB is nonfarm business. Source: NSS-AIDIS 37th, 48th, 59th and 70th Rounds.

37

Annex Table A.2: Fixed capital expenditure in farm business (rupees / rural household) at 2004–5

prices, 2012–13

Land

improvement works

Orchards and

plantations

Irrigation resources

Agricultural machinery, implements,

transport equipment

Farm buildings Livestock Others

Andhra Pradesh 136.5 1.3 123.6 354.0 59.2 612.8 -

Assam 10.3 9.7 26.9 122.0 61.8 66.6 5.8 Bihar 9.0 10.8 27.9 121.4 54.6 66.6 9.9 Gujarat 294.2 6.3 746.6 1,154.7 196.1 730.8 34.8 Haryana 85.6 - 596.4 1,130.6 18.2 759.8 2.6 Himachal Pradesh 580.1 150.1 1,112.3 208.1 795.0 566.4 3.4

Jammu and Kashmir 59.0 240.4 5.9 199.1 33.9 933.5 2.9

Karnataka 255.2 - 1,120.2 430.1 24.3 592.9 7.3 Kerala 96.3 72.2 266.9 1,328.1 35.0 266.9 122.5 Madhya Pradesh 37.1 23.6 57.3 1,191.2 25.3 343.7 8.4

Maharashtra 465.3 139.1 901.2 564.2 72.2 500.1 34.8 Odisha 68.2 1.4 70.3 94.4 40.9 68.6 5.9 Punjab 141.6 - 481.5 3,016.3 56.6 986.6 37.8 Rajasthan 185.9 17.2 1,039.5 877.7 206.5 1,104.9 17.2 Tamil Nadu 34.4 5.0 299.1 112.0 105.1 63.2 6.3 Uttar Pradesh 297.5 14.5 71.1 2,87.3 37.7 551.5 193.0 West Bengal 2.9 5.0 13.2 1,10.2 32.3 98.4 1.3 All-India 135.3 24.5 368.5 647.4 58.7 376.7 21.2

Source: NSS-AIDIS 70th Round.

38

Annex Table A.3: Share of different items of farm business expenditure per rural household (%),

2012–13

Land

improvement works

Orchards and

plantations

Irrigation resources

Agricultural machinery, implements,

transport equipment

Farm buildings Livestock Others

Andhra Pradesh 9.0 0.1 33.4 21.8 2.8 32.9 0.1

Assam 3.4 3.2 8.9 40.3 20.4 22.0 1.9 Bihar 3.0 3.6 9.3 40.5 18.2 22.2 3.3 Gujarat 9.3 0.2 23.6 36.5 6.2 23.1 1.1 Haryana 3.3 0.0 23.0 43.6 0.7 29.3 0.1 Himachal Pradesh 17.0 4.4 32.6 6.1 23.3 16.6 0.1

Jammu and Kashmir 4.0 16.3 0.4 13.5 2.3 63.3 0.2

Karnataka 10.5 0.0 46.1 17.7 1.0 24.4 0.3 Kerala 4.4 3.3 12.2 60.7 1.6 12.2 5.6 Madhya Pradesh 2.2 1.4 3.4 70.7 1.5 20.4 0.5

Maharashtra 17.4 5.2 33.7 21.1 2.7 18.7 1.3 Odisha 19.5 0.4 20.1 27.0 11.7 19.6 1.7 Punjab 3.0 0.0 10.2 63.9 1.2 20.9 0.8 Rajasthan 5.4 0.5 30.2 25.5 6.0 32.1 0.5 Tamil Nadu 5.5 0.8 47.8 17.9 16.8 10.1 1.0 Uttar Pradesh 20.5 1.0 4.9 19.8 2.6 38.0 13.3

West Bengal 1.1 1.9 5.0 41.9 12.3 37.4 0.5 All-India 8.3 1.5 22.6 39.7 3.6 23.1 1.3

Source: NSS-AIDIS 70th Round.

39

Annex Table A.4: Annual rate of growth in fixed capital expenditure in farm business (rupees / rural

household) at 2004–5 prices, 2002–3 to 2012–13

Land

improvement works

Orchards and

plantations

Irrigation resources

Agricultural machinery, implements,

transport equipment

Farm buildings Livestock Others

Andhra Pradesh 11.4 -9.2 -5.2 18.1 6.9 17.1 -100.0

Assam 2.7 15.5 16.8 23.6 10.6 1.9 -5.5 Bihar 18.1 -- 21.7 10.0 19.0 16.8 32.4 Gujarat 13.3 -- 11.0 5.1 11.5 22.6 15.1 Haryana -- -- -7.5 3.5 -17.3 6.0 -- Himachal Pradesh 19.9 11.3 -- -9.2 9.1 10.9 11.0

Jammu and Kashmir 10.7 14.7 -23.6 3.2 -13.8 24.1 -32.0

Karnataka 8.6 -100.0 17.7 15.2 -3.0 23.8 -6.7 Kerala -6.2 -9.5 4.7 40.3 -3.2 17.8 29.7 Madhya Pradesh -2.6 6.8 -5.1 34.5 6.2 25.8 4.5

Maharashtra 22.0 46.8 5.4 12.6 -5.2 15.7 27.8 Odisha 4.4 -- -- -8.3 9.2 10.2 -10.0 Punjab 12.4 -- -2.6 12.5 -11.8 13.2 24.6 Rajasthan 12.1 -- 2.1 5.7 29.1 20.9 -1.9 Tamil Nadu 0.9 15.3 -2.9 2.8 7.3 11.9 -13.6 Uttar Pradesh 6.2 -19.0 -12.2 -4.5 -5.9 20.0 33.8

West Bengal -15.6 15.3 14.6 4.8 8.1 -17.6 All-India 13.1 9.4 5.2 10.4 0.5 15.1 8.5

Source: NSS-AIDIS 59th and 70th Rounds.

40

Annex Table A.5: Share of FCE-FB financed through borrowings and share of institutional

borrowings in total borrowings, 2012–13

% of FCE-FB financed from borrowings % of borrowings from institutional agencies

Andhra Pradesh 98.9 44.6 Assam 49.9 61.4 Bihar 65.3 47.1 Gujarat 84.4 78.8 Haryana 65.7 61.1 Himachal Pradesh 42.6 64.2

Jammu and Kashmir 36.9 43.8

Karnataka 89.2 58.4 Kerala 95.5 82.7 Madhya Pradesh 76.9 58.3 Maharashtra 91.7 71.3 Odisha 66.1 70.9 Punjab 95.4 77.7 Rajasthan 86.6 55.9 Tamil Nadu 95.5 46.6 Uttar Pradesh 85.5 64.3 West Bengal 79.2 69.4 All-India 85.9 63.4

Source: NSS-AIDIS 70th Round.

ALL IFPRI DISCUSSION PAPERS

All discussion papers are available here

They can be downloaded free of charge INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE www.ifpri.org IFPRI HEADQUARTERS 1201 Eye Street, NW Washington, DC 20005 USA Tel.: +1-202-862-5600 Fax: +1-202-862-5606 Email: [email protected] IFPRI SOUTH ASIA REGIONAL OFFICE Block C, NASC Complex, DPS Marg Opposite Todapur, Pusa New Delhi 110012 India Phone: +91-11-66166565 Fax: +91-11-66781699 Email: [email protected] https://southasia.ifpri.info/