cash crops

advertisement
Drivers of commercialisation in
agriculture in Vietnam
Andy McKay and Chiara Cazzuffi
University of Sussex, UK
Paper in progress as part of a DANIDA/BSPS
project with CIEM and UoC
Joint with Luu Duc Khai and Lien Huong Do
Introduction
♦ Part of in depth analysis of VARHS data, follow up to
descriptive report which looks only briefly to the issue
of commercialisation.
♦ Aim is to understand what factors influence market
participation behaviour, with respect to:
(a) the decision to grow cash crops: coffee, tea, cocoa,
cashew nuts, sugar cane, pepper, rubber.
(b) the decision to sell rice: surplus (temporary or
permanent); or to meet immediate expenditure needs
♦ Many actors involved, but we focus on household end
of commercialisation process;
♦ Focus on crop cultivation only.
Introduction (ctd.)
♦ Focus so far on cross sections 2006 & 2008; only
preliminary insights from panel
♦ Analysis without sample weights in order to have a
larger sample
♦ Focus on role of land titling, credit access, training &
extension, group membership, rural infrastructure,
socio-economic characteristics
♦ In context of sharp food price increases
♦ Very preliminary results to invite comments and
develop understanding
World market price for coffee and rice, 2002-2009, US cents/kg
Coffee
70
60
50
40
30
20
10
0
2002
2003
2004
2005
2006
2007
2008
2009
Rice
VARHS 2006 and 2008






89% of sampled households (hh) are engaged in
agriculture in 2006; 91% in 2008.
Of these, 19% produce cash crops (CC) in 2006; 21% in
2008. Majority also grow rice.
75% of hh doing agriculture produce rice in 2006; 88% in
2008.
Of these, 37% sell rice in 2006; 48% in 2008.
Within the panel, there is a net movement of people into
rice production and sales, while there is little change in
cash cropping.
Hh that produce cash crops and hh that sell rice are better
off than their counterparts than don't in terms of food
expenditure per capita
Outline
- Cash crops
- province profile of cash cropping
- characteristics of cultivating households
- econometric analysis (cross sections)
- mobility within the panel
- Rice sales
- characteristics of hh that sell
- summary of econometric results (cross sections)
- mobility within the panel
- Conclusions
Cash cropping 2006 & 2008
Province
Ha Tay
Lao Cai
Phu Tho
Lai Chau
Dien Bien
Nghe An
Quang Nam
Khan Hoa
Dak Lak
Dak Nong
Lam Dång
Long An
Total
HH doing agriculture Cash cropping hh
2006
2008
2006
2008
495
405
9
3
90
280
9
7
321
275
71
59
116
296
12
10
112
295
0
0
196
146
32
19
297
246
6
3
78
41
7
11
143
316
88
230
108
274
78
203
69
61
51
55
299
179
4
3
2324
2814
367
603
%
Main
2006
2.11
10.71
24.48
10.53
0
20.92
2.4
16.28
71.54
78.79
87.93
1.97
2008
0.74
2.5
21.45
3.38
0
13.01
1.22
26.83
72.78
74.09
90.16
1.68
18.79
21.43
Tea
Tea
Tea
Tea
Sugar cane
Pepper
Sugar cane
Coffee
Coffee
Coffee
Sugar cane
Characteristics – cross sections
Growing cash crops
2006
2008
Yes
No
Yes
No
Mean total income
Mean income from agriculture
27204.33
16768.91
22006.87
7008.77
49942.2
34862.46
30792.9
12860.35
Total cultivated land area
% of irrigated land
% of land with crop restrictions
16468.37
0.61
0.59
9752.86
0.62
0.52
18813.62
0.6
0.5
9682.29
0.48
0.31
0.9
0.91
0.81
0.78
11088.9
0.74
0.42
0.98
0.93
1440.83
5609.28
917.34
4416.23
0.7
0.32
0.96
0.94
411.25
1629.39
765.32
44301.06
2.73
0.64
0.97
0.94
2692.12
12604.98
1467.69
12674.49
1.92
0.38
0.86
0.94
515.2
2642.83
974.22
0.76
0.66
0.57
0.41
Prop of hh with a red book
Total value of inputs
Cost of cultivation per unit of land
Proportion of hh hiring labour
Prop of hh using chemical fertiliser
Proportion of hh using pesticides
Cost of hired labour
Cost of chemical fertiliser
Cost of pesticides
Prop of hh who borrowed something
Socio-economic characteristics
Growing cash crops
2006
2008
Yes
No
Yes
No
Prop of hh with male head
Prop of hh whose head is of Kinh ethnicity
Prop of hh whose head speaks Vietnamese
Prop of hh heads who can read & write
Prop of poor hh (MOLISA)
Household size
Prop of hh members of farmers union
Prop of hh members of women union
Prop of hh with access to extension or training
0.87
0.73
0.96
0.9
0.27
4.84
0.57
0.72
0.32
0.82
0.78
0.97
0.89
0.27
4.69
0.52
0.68
0.39
0.88
0.64
0.99
0.9
0.21
4.97
0.57
0.72
0.28
0.84
0.55
0.87
0.79
0.44
5.05
0.51
0.63
0.21
Cash crop growers compared to hh doing
agriculture but not producing cash crops:
1. More likely to be in top food expenditure quintile, and more so in 2008 than
in 2006
2. Mean income from agriculture is twice as high for cash cropping hh in 2006,
and almost 3 times higher in 2008
3. Land area is about double for cash cropping hh in both years
4. Input cost per unit of land is much higher in 2008, and it is higher for cash
cropping hh
5. More likely to hire labour, and much more so in 2008 than in 2006
6. More likely to borrow: 15% more likely in 2006, 39% more likely in 2008
7. Socio-economic characteristics of the hh head: more likely to be male,
speaking Vietnamese and able to read & write, and less likely to be
minority, in both years
8. More likely to have at least one hh member in farmer union and in women
union, in both years
9. Results are stronger in the most dominant coffee growing provinces.
Modelling strategy
We want to understand the correlates of growing
cash crops.
We look at whether some household and
production characteristics have a significant
association with growing cash crops.
We use a probit model with a 0-1 dependent
variable, where 1 means that the hh is
producing cash crops, and 0 that it is not.
At this stage we are not making assumptions
about causality.
Probit model for the probability of growing cash crops – cross sections
Coef
Proportion of restricted land
Log of operated land area
Proportion of irrigated land area
Daily market in commune (yes = 1)
Total input cost (000 VND)
Member of women union (yes = 1)
Hh owns mean of transport (yes = 1)
Hh head is Kinh (yes = 1)
Province fixed effects are included
Observations
Pseudo R2
2006
z-stat Marginal effect
0.314*
(1.908)
0.397*** (5.525)
-0.085 (- 0.422)
0.276*** (2.659)
0.000*
(1.696)
-0.038 (- 0.341)
0.103
(0.627)
0.165
(1.165)
1741
0.461
0.050*
0.063***
-0.013
0.044***
0.000*
-0.006
0.016
0.025
Coef
2008
z-stat Marginal effect
1.327*** (6.371)
0.395*** (6.174)
-0.654*** (- 3.286)
0.014
(0.128)
0.000** (1.980)
0.352*** (3.322)
0.418** (2.182)
0.307** (2.219)
0.243***
0.072***
-0.120***
0.003
0.000**
0.060***
0.062**
0.052**
1769
0.590
Robust z-statistics in parentheses; *** p<0.01, ** p<0.05, * p<0.1
Above showed selected results.
Additional variables used include: household characteristics,
access to extension services, access to credit, land titling
Main Results: strongest correlates
of cash cropping
Land area: strongest message, consistent across
food expenditure quintiles (FEQ) in both years
Restricted land
Inputs expenditure: consistent across FEQ in both
years
Market & own transport; market consistent across
FEQ
Ethnicity: minorities are less likely
Women's union
Mobility in cash cropping: panel 0608,
characteristics take 2006 values
Both years
Switch out
Switch in
Never
Mean total income
Mean income from agriculture
27258.22
16939.17
19996.49
9838.51
19806.22
7920.03
21527.24
7062.26
Total cultivated land area
% of irrigated land
% of land with crop restrictions
17803.3
0.61
0.59
12930.6
0.54
0.51
14432.84
0.44
0.44
9797.84
0.63
0.52
0.92
0.91
0.91
0.91
10726.75
0.67
0.44
0.99
0.95
1387.46
5610.97
888.46
3548
0.4
0.22
0.96
0.98
444.36
1661.66
217.69
3049.22
0.36
0.31
0.97
0.91
119.06
1464.25
216.09
4455.1
0.67
0.31
0.96
0.94
420.27
1645.45
779.03
Prop of hh who borrowed something
0.8
0.73
0.84
0.66
Total number of households
193
55
32
1379
Prop of hh with a red book
Total value of inputs (all crops)
Cost of cultivation per unit of land
Proportion of hh hiring labour
Prop of hh using chemical fertiliser
Proportion of hh using pesticides
Cost of hired labour
Cost of chemical fertiliser
Cost of pesticides
Mobility in cash cropping: panel 0608
Characteristics take 2006 values
Both years Switch out Switch in
Prop of hh with male head
Prop of hh whose head is of Kinh ethnicity
Prop of hh heads who can read & write
Prop of poor hh (MOLISA)
Household size
Prop of hh members of farmers union
Prop of hh members of women union
Prop of hh who experienced a shock
0.88
0.69
0.9
0.32
4.97
0.62
0.77
0.63
0.89
0.67
0.96
0.29
4.67
0.6
0.73
0.56
0.88
0.72
0.84
0.38
4.56
0.75
0.78
0.69
Never
0.82
0.76
0.89
0.27
4.75
0.53
0.69
0.4
Key points:
HH that stop compared to hh that continue: less
income from agriculture, less land, less inputs
expenditure;
HH that start compared to hh that never produce:
more land, less irrigated land, more likely to be
poor and to be member of women's union.
Rice production and sales
Cross sections:



Share of hh growing rice increases from 75% in 2006 to 82% in 2008.
Share of hh growing and selling increases from 37% in 2006 to 48%
in 2008.
Proportion of quantity sold (conditional on selling) is unchanged
between 2006 and 2008;
Panel:

92% of panel hh produce rice both years.

Of these, 35% sell some rice in both years.

18% of rice producers start selling in 2008.

14% of rice producers stop selling in 2008.
Rice production and sale, 2006 & 2008
0.9
0.8
0.7
0.6
0.5
2006
2008
0.4
0.3
0.2
0.1
0
Share of hh growing rice
Share of hh selling rice
Prop of qt sold
Rice sales by food expenditure quintiles



Bottom FEQ most likely to
grow rice in both years;
decline in % of rice growers
in top quintile in 2008.
Share of households
growing & selling rice is
relatively constant across
quintiles in 2006.
Richer households sell a
higher proportion of output.
0.6
0.5
0.4
0.3
2006
2008
0.2
0.1
0
Poorest
2nd poorest
Middle Second richest Richest
Characteristics – cross sections
Sample: hh producing rice
Selling rice
2006
2008
Yes
No
Yes
No
Mean total income
Mean income from agriculture
22204.02
9359.92
23595.66
8420.95
35687.33
17983.2
28620.48
12464.18
Total cultivated land area
% of irrigated land
% of land with crop restrictions
13333.7
0.67
0.54
9164.04
0.58
0.52
11873.47
0.6
0.39
9871.03
0.41
0.29
0.92
0.9
0.82
0.76
0.44
7107.56
0.68
0.42
0.96
0.95
770.72
2638.98
1331.57
0.28
4521.89
0.72
0.27
0.97
0.93
471.81
2164.9
367.2
0.46
23306.22
2.43
0.54
0.92
0.97
1030.91
5105.99
1776.15
0.55
8543.9
1.56
0.31
0.84
0.93
385.14
2093.43
298.05
0.72
0.65
0.45
0.41
Prop of hh with a red book
Prop of hh using hybrid seeds
Total value of inputs
Cost of cultivation per unit of land
Proportion of hh hiring labour
Prop of hh using chemical fertiliser
Proportion of hh using pesticides
Cost of hired labour
Cost of chemical fertiliser
Cost of pesticides
Prop of hh who borrowed something
Socio-economic characteristics - cross
sections; sample: hh producing rice
Selling rice
2006
Prop of hh with male head
Prop of hh whose head is of Kinh ethnicity
Prop of hh whose head speaks Vietnamese
Prop of hh heads who can read & write
Prop of poor hh (MOLISA)
Household size
Prop of hh members of farmers union
Prop of hh members of women union
Prop of hh with access to extension or training
2008
Yes
No
Yes
No
0.84
0.73
0.96
0.88
0.25
4.83
0.52
0.64
0.42
0.81
0.8
0.97
0.9
0.3
4.63
0.54
0.72
0.34
0.85
0.63
0.91
0.85
0.31
5.04
0.53
0.69
0.28
0.86
0.45
0.85
0.76
0.52
5.16
0.52
0.62
0.18
Compared to hh that produce but do
not sell rice, hh that produce and
sell rice:
Have more land, more irrigated land, more inputs,
and are and more likely to have borrowed.
Are less likely to be poor; more likely to have
access to extension or training.
Summary of econometric results
A similar economic model to the one estimated for cash
cropping was also estimated for the probability of
selling rice.
The main results confirm the importance of many of the
factors identified in the descriptive analysis.
Strongest correlates with the probability of selling rice:
More land, more irrigated land, more restricted land,
more likely to have hired labour, more likely to be an
ethnic minority, more likely to have access to extension
Land, training and hired labour consistent across FEQ
Market: consistent across FEQ in 2008.
Probit model for the probability of selling rice – cross sections
Coef
Proportion of restricted land
Log of operated land area
Proportion of irrigated land area
Credit (borrowed smth = 1)
Red book (red book for at least 1 plot = 1)
Distance from all-weather road (km)
Road passable by car in commune
Daily market in commune (yes = 1)
Hybrid seed (1 = yes)
Total input cost (000 VND)
Hired labour (yes = 1)
Member of farmers union (yes = 1)
Member of women union (yes = 1)
Access to training or extension (yes = 1)
Hh owns phone (yes = 1)
Hh owns mean of transport (yes = 1)
Household size
Hh head is Kinh (yes = 1)
Gender of hh head (male = 1)
Hh head speaks Vietnamese (yes = 1)
Hh head can read & write (yes = 1)
Constant
Observations
Pseudo R2
Chi2
0.253**
0.271***
0.752***
0.123*
0.009
0.032***
0.088
0.023
0.581***
0.000
0.296***
-0.102
-0.042
0.125*
-0.256***
0.054
-0.029
-0.366***
-0.042
-0.093
0.085
-2.896***
2006
z-stat Marginal effect
(2.321)
(5.277)
(5.089)
(1.691)
(0.077)
(2.904)
(0.900)
(0.314)
(6.617)
(0.983)
(3.698)
(-1.346)
(-0.512)
(1.736)
(-2.667)
(0.470)
(-1.267)
(-2.841)
(-0.453)
(-0.433)
(0.680)
(- 5.707)
1830
0.223
452.6
Robust z-statistics in parentheses; *** p<0.01, ** p<0.05, * p<0.1
0.100**
0.107***
0.296***
0.048*
0.004
0.013***
0.034
0.009
0.228***
0.000
0.117***
-0.040
-0.016
0.049*
-0.099***
0.021
-0.011
-0.145***
-0.017
-0.037
0.033
Coef
0.287**
0.279***
1.009***
-0.014
-0.003
0.000
0.000
0.300***
-0.131
0.000
0.216***
0.072
0.001
0.163**
-0.114
0.154
-0.003
-0.265**
0.023
-0.076
0.168
-3.217***
1677
0.156
257.5
2008
z-stat Marginal effect
(2.356)
(5.067)
(6.975)
(-0.208)
(-0.029)
(0.941)
(0.287)
(3.861)
(-1.551)
(0.650)
(2.911)
(0.998)
(0.020)
(2.043)
(-1.565)
(1.257)
(-0.151)
(-2.210)
(0.238)
(-0.462)
(1.376)
(- 6.146)
0.114**
0.111***
0.402***
-0.006
-0.001
0.000
0.000
0.119***
-0.052
0.000
0.086***
0.029
0.001
0.065**
-0.046
0.061
-0.001
-0.105**
0.009
-0.030
0.067
Mobility in rice sales: panel 0608,
characteristics take 2006 values
Both years
Switch out
Switch in
Never
Mean total income
Mean income from agriculture
21632.04
9552.76
20294.57
8260.19
20520.68
6921.63
23018.21
5867.99
Total cultivated land area
% of irrigated land
% of land with crop restrictions
14146.89
0.72
0.59
10631.38
0.53
0.47
7389.05
0.61
0.54
7774.83
0.55
0.48
0.92
0.9
0.92
0.93
8406.22
0.74
0.42
0.97
0.97
922.19
3032.83
1736.81
2911.1
0.54
0.34
0.91
0.92
238.03
1165.73
240.21
3111.25
0.66
0.26
0.96
0.94
227.45
1445.72
264.28
2170.48
0.61
0.2
0.95
0.93
79.94
972.8
155.93
Prop of hh who borrowed something
0.73
0.67
0.62
0.67
Number of households
547
224
287
499
Prop of hh with a red book
Total value of inputs
Cost of cultivation per unit of land
Proportion of hh hiring labour
Prop of hh using chemical fertiliser
Proportion of hh using pesticides
Cost of hired labour
Cost of chemical fertiliser
Cost of pesticides
Mobility in rice sales: panel 0608,
characteristics take 2006 values
Both years Switch out
Prop of hh with male head
Prop of hh whose head is of Kinh ethnicity
Prop of hh heads who can read & write
Prop of poor hh (MOLISA)
Household size
Prop of hh members of farmers union
Prop of hh members of women union
Prop of hh who experienced a shock
0.86
0.8
0.92
0.22
4.84
0.56
0.69
0.41
0.83
0.52
0.77
0.36
4.97
0.49
0.6
0.41
Switch in
Never
0.78
0.77
0.88
0.28
4.74
0.58
0.72
0.4
0.83
0.79
0.91
0.32
4.61
0.55
0.74
0.45
Key points
HH that stop compared to hh that continue: less land,
less irrigated land, less inputs expenditure, borrow
less;
more likely to be minority, more likely to be poor, less
likely to be member of women's or farmers union.
HH that start compared to hh that never sell rice: more
restricted land, more irrigated land, more inputs;
Less likely to be male-headed, less likely to be poor,
more likely to be member of a farmers union.
Conclusions
Significant increase in proportion of hh growing rice
and of rice sellers, perhaps as response to rice
price increase over the period.
Small increase in the proportion of hh growing cash
crops.
Important factors for both rice and cash crops: land
size, input expenses, market infrastructure
(presence of daily market in commune) and
possibly access to extension/training and hh
ownership of means of transport.
Important across quintiles.
Conclusions (ctd.)
In predominantly cash crops regions, having a red
book for cultivated land and having access to
credit are important determinants for production
of cash crops.
Irrigated land, restricted land as well as land area
are important determinants for rice sales.
Being a member of farmers union or women's
union seems to be an important factor for
starting cash crop production or rice sales.
Further work
Preliminary results for comment and development
Possible areas for development:
Correlates of rice quantities sold conditional on selling
More in depth analysis of the impacts of rice price change
Panel dynamics
Land area allocated to cash crops
Consider other definitions of cash crops, e.g. fruits
Consider livelihood compositions of hh producing cash
crops or selling rice
Determinants of channel of sales (e.g trader or market)
xin cám ơn!
Download