biosecurity kompas society risk analysis · tom kompas australian centre for biosecurity and...
TRANSCRIPT
Tom Kompas Australian Centre for Biosecurity and
Environmental Economics Crawford School of Economics and Government
Australian National University
www.acbee.anu.edu.au
.
The Problem Close proximity between countries and international trade and tourism increases the probability of an incursion and the spread of exotic diseases and pests; ones that can do great harm, and in some cases can be potentially devastating to local industry, animal and human health, and the environment.
Traditional Measures • Pre-border measures and border quarantine (i.e.,
preventing a potential incursion at the border). • Limits on imports • Airport inspections, and inspections of shipping
containers and contents • Local surveillance programs (preventing spread in
the local environment). • Screening and local awareness • Surveillance traps (e.g., insects) • Blood screening and visual inspection
• Containment and eradication programs.
The Economic Puzzle How much should be spent, or what costs should be incurred, for pre-border measures and border quarantine, surveillance and containment/eradication activities to protect human, plant and animal health as well as the environment? How to allocate resources across various threats?
• Ban imports and close airports? • Spend $0 on quarantine and surveillance? • Spend all of GDP on quarantine and surveillance? • Eradicate? Contain? Neither? • How to allocate resources across various threats? • Who Pays? Cost Sharing?
Rate of Discount? Level of rate, time dependent, hyperbolic?
Non-Market Valuation? Contingent value, stated preference?
Value of Information? Model selection, parameter estimates, value to
precision? Forward Projections?
Spatial and density spread? The economics of containment vs. eradication
A Simple Spread Model for an Invasive
Time
Q(N)
Infected (N)
Containment and Eradication The role of cost-benefit analysis (CBA) after an incursion,
or for a potential incursion
Time
Infected (N)
A
Note: The potential size of benefits depend on size of initial entry and the choice of early detection, for a given eradication/containment exercise.
FMD is endemic in Vietnam, recently as part of a pan-Asian epidemic beginning in the 1990s.
Presumed vectors: border trade with China and Cambodia (virus-strain specific).
Vietnam currently follows a containment campaign, with vaccination to protect non-infected animals, along with control on movements, surveillance and border quarantine. Many areas are in the process of being declared FMD free, and Vietnam cooperates with SEAFMD 2020 in an effort to eliminate FMD from the region.
Vietnam has recently committed roughly $34 million USD for the containment and eradication of FMD (from 2006 to 2010), spearheaded by an extensive vaccination program.
Newly Infected Cows and Buffaloes: The Effect of the Vaccination Program (data from January 2006, by month, and
September 2006, by month)
Key benefit (foregone loss) parameters
2006 2007 2008
Cattle unit price per kg (USD) 1.62 1.82 2.19 Milk price (farm gate) per litre (USD) 0.31 0.38 0.44 Weight and value loss due to FMD infected animal (%)
40% N(1,0.05)
40% N(1,0.05)
40% N(1,0.05)
Weight per cow/buffalo (kg) 250 250 250
Compulsory culled percentage (%) 2% 2% 2%
Ratio of dairy in the total herd (%) 0.63% 0.63% 0.70% Milk produced per dairy cow (litre/year) 3,816 3,816 4,027 Loss in milk production due to FMD infected animal (%)
50% N(1,0.05)
50% N(1,0.05)
50% N(1,0.05)
Meat production (tonnes) 224,746 273,651 298,739
Key cost parameters 2006 2007 2008 2009 2010
Management expenses 16,313 40,625 55,625 55,625 55,625 Training 34,188 34,188 34,125 Awareness campaign 69,688 69,688 104,500 104,500 104,500 International conference 4,375 4,375 4,375 Domestic conference 1,250 1,250 1,313 Outbreak reporting 8,375 8,438 8,438 Sending sample abroad for virus type separation 2,250 2,250 2,250 Research on causes of FMD 20,000 Development of epidemiology map 2,750 2,750 2,750 Blood test analysis and virus study 7,500 18,750 123,125 123,188 123,188 Sterilization chemical 15,000 15,000 118,125 118,125 118,125 Labour cost for vaccination 1,093,750 1,093,750 1,100,500 1,100,500 1,100,500 Control zone vaccine for buffaloes & cows 1,789,992 2,563,130 2,416,376 2,668,688 2,668,688 Buffer zone vaccine for buffaloes & cows 2,986,853 1,523,375 2,715,037 2,852,750 2,852,750 Contingency fund 1,158,323 1,193,384 4,041,523 4,041,523 4,041,523 Vaccine under Decision 738 for buffalos & cows 190,646 504,225 Total Cost 7,328,064 7,021,926 10,747,998 11,118,148 11,118,148
CBA for vaccination against FMD in Vietnam Forward projections of new annual infections (based on
stratified Monte Carlo draws, conditioned by monthly 2006-2008 data on new infections, assuming ongoing and no further gains from vaccination).
0
500
1000
1500
2000
2500
3000
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034
Year Net Present Value (million USD)
Net Present Benefit (million USD)
Net Present Cost (million USD) Benefit Cost Ratio
Mean Standard Error Mean Standard
error Mean Standard error Mean Standard
Error
2040 1,443 324 1,719 320 276 44 6.38 1.57
2027 1,290 263 1,463 260 173 19 8.56 1.81
4.17 9.11 5.0% 90.0% 5.0%
0
0.05
0.1
0.15
0.2
0.25
0.3
2 4 6 8 10
12
14
16
Benefit-Cost Ratio (BCR), 2006-2040
0
1
2
3
4
5
6
7
8
9
10
11
12
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
Surveillance Expenditures S(N)
Earliest Detection
S(N)
Natural Detection
Local Surveillance Measures
N
Time
Q(N)
Infected (N)
Earliest detection
Optimal ‘Early Detection’
S(N) Natural Detection Point
A
Research Design: Optimal Surveillance • Benefit: Surveillance ensures ‘early detection’, lowering economic and
environmental losses and pest/disease management costs.
• Tradeoff: The more early the detection the more expensive the surveillance measure.
• Objective: Find optimal surveillance expenditures to minimize: • Economic losses (e.g., plant and animal losses, damage to the
environment, biodiversity losses, trade bans, etc.) • Eradication and management costs of any pest/disease
incursion • Surveillance expenditures (e.g. monitoring, the cost of setting
and monitoring traps, etc.)
• Method: Stochastic (Optimal Control) Bioeconomic Model with a Jump-Diffusion Process.
Hawkweed Distribution Map
Current infestation 500 ha - 1000 ha (Williams, Hans, Morgan and Holland, 2007)
Hawkweed spread
Time
Are
a in
hec
tare
s
Model Parameters Annual growth rate: 8% - 8.6% (Scott 1993; Johnstone, Wilson and
Bremner, 1999); discount rate 5% Maximum Carrying Capacity: 26,996,700 ha (DAFF) Production loss percentage: 15% on improved land and 5% on
unimproved land (Grundy 1989) Production loss value: $60-$100 ha;$959 million for grazing; $289
million for perennial horticulture (1996 land use commodity) (Brinkley and Bomford 2002)
Eradication cost: 65 – 83 AUD / ha with multiple treatments (Espie, 2001; Ministry of Agriculture, Food and Fisheries, British Columbia, Canada)
Natural detection: 100 ha Maximum annual surveillance cost to detect Hawkweed when the
spread is equal or greater than 1 ha: 200,000AUD per year, or approx 4 million NPV over time.
Costs of controlling Hawkweed
Hectare
Cos
ts in
mill
ion
AU
D
Hectare
Cos
ts in
mill
ion
AU
D
Sensitivity with the discount rate
Hectare
Cos
ts in
mill
ion
AU
D
Sensitivity with the damage per hectare
Hectare
Cos
ts in
mill
ion
AU
D
Sensitivity with the incursion probability
Time
Q(N)
Infected (N)
Border Quarantine Measures Quarantine
Expenditures E(N)
Zero entry
Zero entry
E(N)
N
Optimal Border Quarantine Benefit: Less initial entry gives smaller damages (or more benefits)
over time, prior to full saturation. Trade-off: The smaller is initial entry the more expensive is the border
quarantine program Objective: Minimize all expenditures: damages (e.g., losses in plant
and animal health, damage to the environment, trade restrictions, containment and eradication costs) plus the cost of the quarantine program itself. (Equivalent to equating the marginal benefits of having less damages to the marginal costs of quarantine expenditures.)
Conditional on: Probability of entry Spatial and density spread, and E(N) function Prices and costs that may vary over time, discount rate
