Overview
Factors associated with the
technical, allocative and economic
efficiency of Irish dairy farms
Background
Objectives
Efficiency measurement
E. Kelly 1,2, L. Shalloo 1, U. Geary 1 , A. Kinsella 3,
F. Thorne 3, M. Wallace 2
•
•
•
•
Technical, allocative and economic efficiency
Data Envelopment Analysis (DEA)
Models
Results
1Teagasc,
2School
Livestock Systems Department, Animal & Grassland and Innovation Centre,
Moorepark, Fermoy, Co. Cork,
of Agriculture Food Science and Veterinary Medicine, NUI Dublin, Belfield, Dublin 4,
3Teagasc, Rural Economy Research Centre, Athenry, Co. Galway, Ireland
Background
Conclusion
Background
Rising costs of inputs
Volatile prices
Policy changes
• Quota removal
• Subsidies
Need to increase efficiency
Management
• Widely regarded as associated with differences in efficiency
Objectives
Objectives
 Determine the levels of technical, allocative and economic
efficiency of Irish dairy farms
 Investigate the associations of key management, qualitative
and demographic characteristics on efficiency
Efficiency measurement
Definitions of Efficiency (Farrell 1957)
• Technical: ability to maximise output from the current level
of inputs
• Allocative: ability to react to market changes when
choosing inputs
• Economic: technical x allocative efficiency
Data Envelopment Analysis
(DEA)
Data Envelopment Analysis (DEA)
• Developed by Charnes et al., (1978)
• Non-parametric methods
• Deterministic
• Linear programming
• Efficiency scores range between 0 and 1
B
C
Efficient
C
D
0
Inefficient
x2/q
x2/q
(Input 2)
DEA Frontier
(Input 2)
DEA Frontier
Efficient
D
x1 /q
(Input 1)
0
x1 /q
(Input 1)
DEA Frontier Economic Efficiency
θ = 0B0/0B
θ = 0B’’/0B
x2/q
x2/q
Projected Point
(Input 2)
(Input 2)
DEA Frontier
Inefficient
B
Inefficient
B0
B
C
C
Efficient
Isocost Line
D
D
x1 /q
0
(Input 1)
x2/q
Projected Point
θ = 0B’’/0B
Inefficient
B
C
(Input 1)
DEA Frontier – Economic Efficiency
Projected Point
x2/q
(Input 2)
DEA Frontier – Economic Efficiency
x1 /q
(Input 2)
0
B’’
θ = 0B’’/0B
Inefficient
B
C
B’’
Technical and
economic efficient
Isocost Line
Isocost Line
D
0
D
x1 /q
(Input 1)
0
x1 /q
(Input 1)
Efficiency Models
•
Dataset
• National Farm Survey (NFS), 2008
• Annual survey of Irish dairy farms
• Over 300 farms
•
Input Variables
• Land, cows, labour, concentrate, fertiliser, other direct and
overhead costs (Physical and financial)
•
Output Variables
• Milk Solids (MS) and other farm output (Physical and financial)
• Models
• Variable Returns to Scale (VRS)
Results
Second Stage Analysis
• Tobit Regression: Suitable as DEA efficiency scores
bound between 0 and 1
• Management Factors
• Qualitative Factors
• Demographic Factors
Technical efficiency
90
80
70
60
50
No of producers 40
30
20
10
0
100%
8089%
6069%
4049%
Efficiency (%)
2029%
0-9%
Average
0.771
Minimum
0.106
Maximum
1.000
St Dev
0.182
Allocative efficiency
Economic efficiency
80
70
60
No of 50
40
producers
30
20
10
0
Average
100%
8089%
6069%
4049%
2029%
0-9%
0.740
Minimum
0.213
Maximum
1.000
St Dev
0.159
Efficiency (%)
70
60
50
40
No of
producers 30
20
10
0
100% 80- 6089% 69%
Average
0.571
Minimum
0.089
Maximum
1.000
St Dev
0.195
40- 20- 0-9%
49% 29%
Efficiency (%)
Results summary
•
Potential to increase technical, allocative and economic efficiency
•
More technically efficient than economically efficient producers
•
85% of sample had technical efficiency > 60% with 13% fully technically efficient
•
83% had allocative efficiency>60% with 3% fully allocative efficient
•
42% had economic efficiency>60% with 3% fully economic efficient
•
Need to focus on key factors associated with efficiency
Factors associated with
technical efficiency
Association of key factors with technical efficiency
Model 1
Model 2
Model 3
Model 4
Model 5
Variable
Estimate
Estimate
Estimate
Estimate
Estimate
Intercept
-0.4504***
-0.3563***
-0.3214***
-0.3331**
-0.2857*
Grazing Days
0.0013***
0.0010**
0.0007*
0.0002
0.0002
Breeding Days
-0.0005***
-0.0004**
-0.0003*
-0.0003*
-0.0003*
Dairy% Output
0.0564
-0.0451
-0.0507
-0.0200
-0.0279
Milk Penalties €
-0.000*
Milk Bonuses €
0.000***
-0.000*
-0.000*
-0.000**
-0.000**
0.000**
0.000***
0.000**
0.000**
Milk Recording
0.0450
0.0398
0.0327
0.0308
Discussion†
0.0330
0.0472*
0.0323
0.0426*
Calving‡ ≤Feb14th
0.0624
0.0761*
Calving ≤March1st
0.0997**
Soil§ 1
0.0833*
0.1033**
0.1076**
0.1142**
0.1050*
Age
Factors associated with
allocative efficiency
0.0173*
* Significance PROC LifeReg SAS *** <0.001, **0.001-0.01, *0.01-0.05
† Discussion Group Member.
‡ Mean calving Date: Dummy variable, 1 if calving ≤14th February, 2 if calving ≤ 1st March, 3 if calving ≤17th March, 4 if calving ≤31st March and 5 if
calving ≥ April 1st
§ Soil quality ranges from 1-6 with 1 an indication of best soil with widest range of use and 6 the poorest soil quality with most limited range of use
Association of key factors with allocative efficiency
Model 1
Model 2
Model 3
Model 4
Model 5
Variable
Estimate
Estimate
Estimate
Estimate
Estimate
Intercept
-0.5981***
-0.4933***
-0.5034***
-0.5484***
-0.5004***
Grazing Days
0.0006*
0.0001
0.0003
0.0001
0.0004
Breeding Days
0.0001
0.0001
0.0000
0.0000
0.0000
Dairy% Output
0.3290***
0.2147**
0.2278**
0.2357**
0.2363**
Milk Penalties €
0.000
0.000
0.000
0.000
0.000
Milk Bonuses €
0.000*
0.000*
0.000*
0.000
0.000
Milk Recording
0.0612**
0.0631**
0.0657**
0.0648**
AI use
0.0431*
0.0450*
0.0497*
0.0451*
Discussion†
0.0362
0.0368
0.0347
0.0371
-0.0510*
-0.0472*
Calving‡≤March 17th
Soil§
1
Soil 2
Age
-0.0445
0.0736*
0.0730*
0.0956*
0.0984*
0.0014
* Significance PROC LifeReg SAS *** <0.001, **0.001-0.01, *0.01-0.05
† Discussion Group Member.
‡ Mean calving Date: Dummy variable, 1 if calving ≤14th February, 2 if calving ≤ 1st March, 3 if calving ≤17th March, 4 if calving ≤31st March and 5 if
calving ≥ April 1st
§ Soil quality ranges from 1-6 with 1 an indication of best soil with widest range of use and 6 the poorest soil quality with most limited range of use
Factors associated with
economic efficiency
Association of key factors with economic efficiency
Conclusions
Model 1
Model 2
Model 3
Model 4
Model 5
Variable
Estimate
Estimate
Estimate
Estimate
Estimate
Intercept
-1.3882***
-1.2421***
-1.2243***
-1.2366***
-1.0253***
Grazing Days
0.0025***
0.0017**
0.0015**
0.0009
0.0009
Breeding Days
-0.0004*
-0.0004*
-0.0003
-0.0003
-0.0003
Dairy% Output
0.6034***
0.4089**
0.3809**
0.4241***
0.3968**
Milk Penalties €
-0.000
-0.000
-0.000
-0.000
Milk Bonuses €
0.000***
-0.000
0.000***
0.000***
0.000***
0.000***
Milk Recording
0.1050**
0.1038**
0.0980**
0.0958**
AI use
0.1142**
0.1107**
0.1147**
0.1240**
Discussion†
0.0580
0.0577
Calving‡ ≤March1st
Soil§ 1
Age
0.0852*
0.0724*
0.0816*
0.0805
0.0953*
0.1430*
0.1269*
0.0274*
* Significance PROC LifeReg SAS *** <0.001, **0.001-0.01, *0.01-0.05
† Discussion Group Member.
‡ Mean calving Date: Dummy variable, 1 if calving ≤14th February, 2 if calving ≤ 1st March, 3 if calving ≤17th March, 4 if calving ≤31st March and 5 if
calving ≥ April 1st
§ Soil quality ranges from 1-6 with 1 an indication of best soil with widest range of use and 6 the poorest soil quality with most limited range of use
Conclusion
Conclusion
Greater efficiency associated with
•
•
•
•
•
•
•
•
•
•
Mean calving date
Grazing days
Breeding season length
Milk quality
Discussion group membership
Milk recording
AI use
Dairy specialisation
Land Quality
Age
• Potential to increase efficiency at farm level
• Identified KPI at farm level that are causing differences in efficiency
among producers
• These factors can be used for the benchmarking of producers
Future work
Acknowledgements
•
• Scale efficiency of Irish dairy producers
Walsh Fellowship
•
Supervisors
Brian Moran
• Measure productivity over a longer period of time
NFS Farmers
Sean Cummins
James O’Loughlin
Brian McCarthy
Thank you for your attention!