Investment Priority for Poverty Reduction and Higher Agricultural

Report
Investment Priority for Poverty Reduction and Higher
Agricultural Productivity
Seema Bathla
Centre for the Study of Regional Development
Jawaharlal Nehru University, New Delhi
Project Team
P K Joshi, S K Thorat and Bingxin Yu
Rationale for Public Investment
•Social benefits from agricultural expenditure are far greater than
the private producer benefits
•Amount spent by private sector tends to be lower than the
socially optimal level, and this under-provision creates a rationale
for the public provision of such goods
•Literature abounds on the positive impact of public expenditures
on agricultural output and poverty in developing economies (Fan
and Rao (2008), Fan and Brzeska (2010), Mogues et al. (2012)
Public spending “in agriculture” and “for agriculture”
Rapid growth of productivity drives agricultural growth and is key to
poverty alleviation, mostly through technology adoption & increased spending
on R&D (Fan et al. 2006; Nin-Pratt, Yu and Fan 2009; Fuglie 2010).
The Prescription:
If government has an additional billion rupees to spend on
agriculture, it should be spent on investments rather than
subsidies to achieve higher agricultural growth and rural poverty
reduction. In other words, input subsidies “crowds out” public
investments in agriculture.
Literature (IFPRI 1999; 2002 & many others) Suggested :
•A reversal in the declining trend in public investments in
agriculture witnessed during the 1980s and 1990s
• Subsidies tend to be unproductive due to low marginal returns A check on the rising level of farm input subsidies, no matter how
important they have been for agricultural growth
The Scenario during the 2000s
•A phenomenal increase in the public investment (capital
formation) in agriculture
•Investment on private account has grown at a much higher rate
•Input subsidies continue to rise at nearly 3% per annum; a
higher growth in credit subsidy is identified in recent period
•Magnitude of input subsidies in 2000s three times higher than
public investment in agriculture compared to four times during
the 1990s
Annual Rate of Growth in Agriculture Investment and Input
Subsidies from 1981-2012 at 2004-05 price (Rs. Billion)
Average
Public
GCFA
Private
GCFA
GCFA
Input
Subsidy Irrigation
Credit Power Fertilizer
Input
Subsidy/
Capita
(Rs.)
1981-1989
105
231
336
347
36
19
54
235
618
1990-1999
92
329
422
476
62
14
201
195
744
2000-2011
175
769
945
639
86
47
257
248
825
1981-2011
127
461
588
502
64
29
180
227
739
Annual Rate of Growth
(Percent)
1981-1989
-3.70
1.79
0.14
8.04
8.69
12.36
21.09
4.99
6.13
1990-1999
2.78
3.11
3.06
2.49
2.49
-10.24
11.87
-4.24
1.43
2000-2011
6.38
8.75
8.25
3.18
3.19
9.99
2.33
2.98
1.52
1981-2011
3.27
6.39
5.64
2.60
3.21
3.96
6.65
0.12
1.00
Source: Finance Accounts, GOI, Agriculture Statistics at a Glance, Statistical Abstract of India, Fertilizer Statistics.
Note: Irrigation subsidy is estimated from 1983-84 onwards . Public and private GCFA is up to 2010-11.
1400
1100
1200
1000
900
1000
800
800
700
600
600
500
400
400
Public Investment Agriculture
Input Subsidy
2011
2009
2007
2005
2003
2001
1999
1997
1995
1993
1991
1989
200
1987
0
1985
300
1983
200
1981
Public and Private GCFA
Fig.1 : Public and Private Investment in Agriculture and Input Subsidy,
Rs. Billion at 2004-05 Price
Private Investment Agriculture
Fig. 1b: Tradeoff between Public Investment and Farm Input
Subsidy, Rs. Billion at 2004-05 Prices
250
1200
200
1000
800
150
600
100
400
50
200
0
0
Public Investment Agriculture
Input Subsidy
Fig.2a Agricultural Expenditure and Input Subsidies, in billion Rupees
Note: Fertilizer subsidy taken from Fert. Stat. is apportioned to states as per consumption;
Other subsidies based on the difference between suppliers’ costs and revenues received from farmers.
Fig. 2b:Trends in Input Subsidy in Agriculture, Rs. Billion at 2004-05 Price
700
600
500
400
300
200
100
0
Irrigation
Credit
Power
Fertilizer
Fig. 3: Percentage Share of Input Subsidy in Agriculture Income
25
20
15
10
5
0
Input Subsidy Per NSDPA
Fig. 5: Regional Disparities in Input Subsidy Per Ha Rs. at 2004-05 Prices
(Average 1981 to 2011)
6500
5500
4500
3500
2500
1500
Input Subsidy/ Crop Area
TN
Punjab
Haryana
AP
Gujarat
UPUttra
Karnataka
JK
Kerala
UP
Maharashtra
BihJhar
WB
HP
MPChat
Rajasthan
Odisha
Assam
500
Fig. 6: Regional Disparities in Per Capita Public Expenditure
7500
6500
5500
4500
3500
2500
1500
500
Per Capita Development Expenditure
Structure of Social and Economic Expenditure and Share, Average 2000-11
Social welfare & nutrition
Welfare of SC, ST &
backward classes
Medical & public health
Rural development Irrigation & flood control
Economic services
Agriculture & allied
services
Rural energy
Education & sports
Rural road & transport
Village & small industries
Percentage Share of Various Heads of Expenditure in Total (Av. 1981 to 2011)
Share of
capital exp.
(%)
1981-1989
19901999
2000- Growth
2011 rate (%)
Total exp. (billion rupees)
1412.2
2427.3
5245.0
6.6
12.3
Development exp. (billion rupees)
1034.9
1589.8
3283.4
5.8
17.9
Social exp. (billion rupees)
516.9
827.2
1763.1
6.3
5.56
Economic exp. (billion rupees)
517.9
762.6
1519.7
5.4
31.0
100
100
100
5.4
Agriculture & allied services
21.8
22.0
17.3
4.2
9.1
Rural development
15.0
14.9
12.9
4.6
6.7
Irrigation and flood control
33.1
26.8
23.7
3.9
55.7
Rural energy
1.3
4.3
5.4
12.2
28.9
Rural road and transport
9.9
8.6
11.4
1.8
9.3
Village and small industry
2.84
2.35
1.37
1.76
45.6
Average share within head (%)
Economic services
Average share within
Agriculture (%)
Share of
2000- Growth
capital exp.
2011 rate (%)
(%)
19811989
19901999
Agriculture & allied services
100
100
100
4.2
Crop husbandry
23.1
27.9
27.8
5.8
2.5
5.9
6.9
5.7
3.9
20.9
Animal husbandry
11.6
11.4
10.9
3.9
2.3
Dairy development
13.6
8.5
4.6
-1.9
2.1
2.3
2.5
2.5
4.7
17.7
22.3
19.7
19.1
3.8
10.6
Plantations
0.4
0.1
0.9
-
71.2
Food, storage warehousing
2.5
5.4
10.2
-
-
Agri. research and education
6.3
6.2
7.3
4.8
1.8
Agri. financial institutions
0.8
0.7
0.02
--
--
11.1
10.4
11.1
4.0
21.7
0.8
0.9
0.7
3.7
17.1
Soil and water conservation
Fisheries
Forestry and wildlife
Cooperation
Other agri. Programs
Development and Non-development Expenditure (Rs. Billion)
2011-2012
2009-2010
2007-2008
2005-2006
2003-2004
2001-2002
1999-2000
1997-1998
1995-1996
1993-1994
1991-1992
1989-1990
1987-1988
1985-1986
1983-1984
1981-1982
0
1000
2000
3000
Development
4000
5000
Non-development
6000
7000
8000
Estimating the Relative Impact of various
Subsidies and Investments on Agricultural
Growth and Poverty reduction
Conceptual and Analytical Framework
Public Spending
Investments
Subsidies
Food Sec &
Empl Progs
Other exog. var.. Pop.
growth, agro-eco factors,
urban growth, macro and
trade policy and world price
AgriInputs
Soc Serv: Edn,
Health, Welfare
Infra: Irrigation,
Power, Road
Agriculture Productivity and
Growth
Non-agri Empt:
(MNREGA, Vill
Ind etc.)
Non-farm
Wages
Farm
Wages
Food
prices
Rural Poverty
Red line indicates indirect effect
Tech: Fert,
Seeds, R&D
1. RPoverty = f1(AY, TOT, AWage, NFEmpl , Pop Density, Rain)
2. Agri Income (AY) = f2 (Agri Research, Land, Labour, IRRI, ELECT, LITE,
FERT, ROAD, Rain, NFSM-Dummy)
3. AWAGE = f3(GDPGNA, AY, ELECT, ROAD, LITE, HealthExp, MNREGAdummy)
4. NFEmpl= f4(GDPGNA, NA Wage, ROAD, LITE, ELECT, RurDev Exp, Vill
Ind Exp, MNREGA )
5. TOT = f5(AY, World price, GDPGNA)
6. FERT = f6(Input subsidies , TOT(-1), IRRI, Agri Research, ROAD, ELECT)
7. IRRI = f7(Irrigation Exp., TOT, Power Subsidy, Rainfall)
8. ELECT = f8(Energy Exp.)
9. ROAD = f9(Road-transport Exp.)
10.LITE= f10(Education Exp.)
Changes Introduced based on IFPRI (2007)
•Data base based on Finance Accounts ; consistent from 1981-82 to 2011-12
based on 1987 budgetary classification
•Agriculture income taken to be land productivity instead of labour
productivity
•Literacy replaced with the number of years of schooling
•No. of villages electrified replaced with power consumption in agriculture
•Inclusion of non-agricultural wage rate
•Dummy to capture the impact of MNREGA and National Food Security
Mission during 2000
•Impact of Change in Policy Regime (election dummy to capture change in the
government )
•Public capita public expenditure is specified in lag as well as in stock terms
•Expenditure on agriculture research & development broaden to include
expenditure on soil conservation and crop husbandry
-Methodology : Both Structural Econometric Model based on
Maximum Likelihood Estimation and 3SLS (SURE)
-Use of double log functional form
-Inclusion of state and time effects
-Time period : 1981-1989; 1990-1999 and 2000-2011 based on 17
states data. Newly formed states combined (Jharkhand,
Uttrakhand and Chattisgarh) with old ones
-Test of model specification: endogeneity, identification,
multicollinearity, specification and measurement errors
-Hausman test shows a few subsidies and expenditure variables to be endogenous. One
year lag is taken to address the problem.
-Sargan test used for identification
-Alternate equations to address multicollinearity
Estimated Rural Poverty Equation
Agricult
ure
Income
TT
Agricultu Non Farm
re
Employm
Wage
ent
(% Share)
Rain
Con R2
stant
Likeliho
od Ratio
-0.21**
-0.49*
-1.17*
-0.06
0.07
10.7 0.82
*
4781.6
1980s
N=136
-0.18*
-0.42*
0.19***
-0.25*
-0.01
7.9* 0.96
1899.98
1990s
N=170
-0.63*
-1.88*
--
0.001
0.12*
*
17.0 0.83
2*
2484.55
2000s
N=204
-0.47*
-2.65*
-0.29
-1.29*
0.35*
22.6 0.89
*
2928.09
19812011
N=510
*,**,*** indicate significance at 1, 5,10% level. State effects significant. TOT is 3 yrs moving average.
Agricultural wage and non-farm employment are correlated during 1990s.
Estimated Agriculture Income (Land Productivity) Equation
Fertilize Irrig Labour Rain
r
ation
Agri
cultu
re
R&
D
Agri. Educa Roa
d
Electr tion
icity
Consu
mptio
n
NFS
M
Cons R2
tant
19812011
0.12*
0.14
*
0.09*
0.04*
0.08
*
0.11*
0.23*
0.01
0.07*
7.8* 0.9
5
1980s
0.15*
--
0.10
0.22*
0.07
***
-0.03
0.17
0.22
***
--
5.46 0.9
7
*
1990s
0.11*
0.05
***
0.02
--
0.06
0.13*
0.49*
0.16
*
--
7.84 0.9
6
*
2000s
0.43*
0.09
*
0.26*
--
0.25 -0.12* 0.35* -0.03
*
--
4.59 0.8
6
*
Variables specified on per ha basis; public expenditure is per capita; GCA/worker dropped due to
multicollinearity; NFSM dummy from 2006. extended agri-research variable correlated so the original variable
(agri R&D is taken.
Estimated Agriculture Wage Equation
Agricultur Agri.
Road
Educati Health
MGNR GDP C R2
e Income Electricity Density on
Expenditu EGA
GNA
Consumptio
re
n
19810.34*
-0.01
0.15*
0.43*
0.29*
-0.01
-0.27
- 0.8
2011
3.2 8
*
1980s
0.23*
0.03**
-0.02
0.42*
-0.01
--
-0.25
0.7 0.4
9 7
1990s
0.24*
0.03**
0.04
0.34*
0.22*
--
-0.05
- 0.5
0.7 7
4
2000s
0.25*
0.02***
0.002
0.94*
0.17*
-0.03
-0.48
- 0.6
1.1 6
1**
Variables specified on per ha basis; public expenditure is per capita
Estimated Non-farm Employment Equation
Non
Farm
Wage
Agri.
Electricity
Consumpti
on
Road
Educa Rural MG
Density tion
Devel NRE
opme GA
nt
Villag GDP
e
GNA
Indust
ry
Co R2
nst
ant
0.17 -0.12* 0.43* 2.1 0.79
*
5
19812011
0.14*
0.09*
0.05
0.69*
0.09*
1980s
-0.32*
0.06*
0.42*
0.39*
0.02
--
-0.03
-0.07
1.2 0.96
0*
1990s
0.12**
*
-0.04
0.03
0.38*
**
0.01
--
--
0.4*
3.8 0.89
4*
2000
0.05**
*
0.11*
0.005
0.93*
-0.02
0.01
-0.02
0.21* 1.9 0.97
5*
Variables specified on per ha basis; public expenditure is per capita
Estimated Agriculture Price (TOT) Equation
GDPGN Trend Consta R2
nt
A
Agriculture
Income
World Price
1981-2011
-0.36*
0.17*
0.31*
0.03* -43.94*
0.72
1980s
-0.35*
0.11
0.29*
0.02* -37.61*
0.89
1990s
-0.11***
0.13***
-0.05
0.01* 26.93*
0.79
2000s
-0.21*
0.29*
0.17***
0.02* -26.26*
0.85
Estimated Agril. Production Equation (captured through
Fertilizer Demand )
Irriga Power Fertilize Cred Agricu TOT
tion
Subsid r
it
lture
(-1)
Subsi y
Subsidy Subs R&D
dy
idy
Roa Irriga Tren Const R2
ant
d
tion d
Den
sity
0.05** 0.44* 0.36 0.06*
*
--
-3.14 0.9
4
0.21*
--
-4.18 0.9
2
1981- 0.04*
2011
0.01
0.39*
--
1981- 0.11*
2011
0.02
--
0.01
1980
s
-0.02
0.07*
0.46*
0.43 0.08**
*
0.08
--
1990
s
0.05*
**
-0.06
0.07*
0.73*
0.12
*
0.07*
-0.18
--
-0.01
0.65*
0.04
*
0.07*
0.01
--
2000
s
0.47* 0.73 0.07*
*
*
0.12* - 72.51 0.8
* * 0.04 ** 9
**
0.05 0.03 - 0.9
* 77.25 4
*
0.20* - 100.4 0.8
0.05 *
8
*
Variables specified on per ha basis; public expenditure is per capita; Multicollinearity between fertilizer and credit subsidy during 19812011
Irrigation (Share of Irrigated Area in total) Equation
Public
Power
Expenditure Subsidy
on Major
and medium
irrigation
Terms of Trade
Constant R2
1981-2011
0.16*
0.05*
0.12**
1.54*
0.87
1980s
0.07*
-0.01
-0.10
3.65*
0.98
1990s
0.08
0.05
-0.36
4.09*
0.89
2000s
0.10*
-0.001
0.18**
2.16*
0.96
Public expenditure is in stock terms and specified on per capita basis; Rainfall is correlated with other
explanatory variables so dropped.
Electricity (Share of Power Consumption in
Agriculture in Total) Equation
Public
Expenditure on
Rural Energy
Constant
R2
1981-2011
0.24*
-1.49*
0.89
1980s
0.18*
-2.05*
0.94
1990s
0.11*
-0.72*
0.96
2000s
-0.04
0.58*
0.97
Road (Road Density Total) Equation
Public Expenditure on
Rural Road and
Transport
Constant
R2
1981-2011
0.25*
5.16*
0.92
1980s
0.18*
5.34*
0.98
1990s
0.29*
4.89*
0.97
2000s
0.10*
6.2*
0.95
Literacy (No. of Years of Schooling) Equation
Public Expenditure
on Education
Constant
R2
1981-2011
0.65*
-3.05*
0.91
1980s
0.43*
-1.98*
0.98
1990s
0.46*
-1.89*
0.96
2000s
0.34*
-0.91*
0.92
Broad Findings
• Investments in agriculture (both public and private) have increased at a much faster
rate during 2000 compared to subsidies; subsidies not crowded out public investment
•Investment & subsidy distribution across states continues to be extremely unequal
•Relative decline in expenditure on economic heads – Agriculture, irrigation and rural
development have born the maximum brunt
•Land productivity, non-farm employment and agriculture prices have significant
impact on poverty reduction
•Road & electricity appear to have a decreasing impact on land productivityFertilizer, irrigation, education, Agriculture R&D and NFSM dummy significant
•Rural wages are increasingly affected by agril. productivity and education but
decreasingly affected by road density
•Urban growth, education & non-farm wages have increasing impact on nonfarm
employment
•Agricultural prices are negatively affected by agricultural growth and positively by
world prices and urban growth
•Power and irrigation subsidies have relatively decreasing impact on production
(captured through fertilizer use).
Investments vs Input Subsidies based on Estimated Marginal Returns
Returns in Agricultural GDP (Rs per million
Rs Spent)
Ranking
1981- 19812011 1989
1990-1999 2000-2011 1981-2011 1981-1989 1990-1999 2000-2011
Agriculture R&D
21.36
30.38
24.71
45.27
1
2
2
1
Education Inv
3.39
3.59
3.58
4.29
4
5
4
5
Health Investment
6.89
-0.29
6.10
3.67
2
10
3
4
3.27
6.63
2.28
0.01
5
4
6
7
1.50
1.36
0.40
0.75
8
7
9
6
Rural Energy
Investment
3.54
6.60
2.86
0.11
3
3
5
9
Rural Development
1.37
0.75
0.35
-0.44
10
9
10
10
Irrigation Subsidy
2.67
-1.57
-3.26
-3.82
6
11
11
11
Fertiliser Subsidy
2.28
1.28
2.12
8.87
7
8
7
2
Power Subsidy
0.84
1.61
1.25
-0.11
11
6
8
8
Credit Subsidy
1.49
64.29
33.58
4.60
9
1
1
3
Rural Road &
Transport Inv
Irrigation
Investment
Returns in Rural Poverty (Decrease in number of poor per
million Rs spent)
Ranking
1990198119901981-20111981-1989 1999 2000-2011 2011 1981-1989 1999
2000-2011
Agriculture
R&D
982
2463
1278
1278
1
2
2
1
Education
Investment
156
291
185
121
4
5
4
5
Health
Investment
317
-24
315
104
2
10
3
4
Rural Road &
Transport Inv
Irrigation
Investment
Rural Energy
Investment
Rural
Development
Irrigation
Subsidy
Fertiliser
Subsidy
Power Subsidy
Credit Subsidy
150
538
118
0.4
5
4
6
7
69
110
21
21
8
7
9
6
163
535
148
3
3
3
5
9
63
61
18
-12
10
9
10
10
123
-127
-169
-108
6
11
11
11
105
39
104
131
109
65
251
-3
7
11
8
6
7
8
2
8
69
5213
1738
130
9
1
1
3
Implications
• Absence of tradeoff between productivity and poverty reduction from public
spending on various heads and subsidies
•Fertilizer and credit subsidies have large growth effects in 2000: No more
unproductive and compete with public investment. If macro fiscal policy allows
more funds towards more investments and subsidies, then no issue
•Returns are relatively higher from fertilizer and credit subsidies compared to rural
energy and irrigation subsidies –Focus on their rationalisation rather reduction
•Marginal impact of investment in rural energy on agril productivity has declined
•High returns to investments from road and irrigation have subsided; but
expenditure should be there for maintenance ; Regional impact may stand imp.
• Roads-transport investment holds importance for non-farm employment and
hence poverty reduction.
•Agriculture research continues to be the most effective public spending item
followed by health, education , irrigation & credit subsidy; to increase it and also
the investment intensity
•Non-farm opportunities to be increased through agro-processing (not much effect
of rural development and village industry)
Thank you!

similar documents