use of cost: benefit analysis to inform risk management decisions in pra

41
Use of cost: benefit analysis to inform risk management decisions in PRA Alan MacLeod Plant Health Group, Central Science Laboratory, York YO41 1LZ, UK International Plant Health Risk Analysis Workshop,

Upload: darrin

Post on 31-Jan-2016

49 views

Category:

Documents


2 download

DESCRIPTION

Use of cost: benefit analysis to inform risk management decisions in PRA. Alan MacLeod Plant Health Group, Central Science Laboratory, York YO41 1LZ, UK International Plant Health Risk Analysis Workshop, Niagra Falls, Canada Oct. 24-28 th 2005. Outline of presentation. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Use of cost: benefit analysis to inform risk management decisions

in PRA

Alan MacLeod

Plant Health Group, Central Science Laboratory, York

YO41 1LZ, UK

International Plant Health Risk Analysis Workshop, Niagra Falls, Canada Oct. 24-28th 2005

Page 2: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Outline of presentation

• Costs and benefits in SPS and the IPPC

• Two examples – Thrips palmi– Diabrotica virgifera virgifera

• Strengths and weaknesses

• Key challenges

Page 3: Use of cost: benefit analysis  to inform risk management decisions  in PRA

SPS Agreement (Art. 5 Assessment of risk)

• In assessing the risk to plant health, Members shall take into account economic factors: – potential loss of production or loss of sales

resulting from the entry, establishment or spread of a pest or disease;

– the costs of control or eradication

• SPS agreement does not mention benefits– cost-effectiveness of approaches to limit risks

Page 4: Use of cost: benefit analysis  to inform risk management decisions  in PRA

IPPC ISPM 11 (PRA)

• Stage 3: Pest risk management

• Point 3.4– Those measures with an acceptable

benefit-to-cost ratio should be considered

– measures chosen due to effectiveness of reducing probability of introduction (entry and establishment)

Page 5: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Example 1: Thrips palmi

• EU Quarantine pest• Wide range of

commercial hosts– Aubergines (egg plant)– Cucumbers – Sweet peppers– many ornamentals

• Vector of plant viruses– Melon spotted wilt virus– Watermelon silver mottle

virus

Page 6: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Thrips palmi – global distribution

Page 7: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Example 1: Thrips palmi

• Pest risk assessment shows could establish in glasshouses in northern Europe

• Previous outbreak in NL glasshouses

Page 8: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Thrips palmi – 1st UK outbreak

• Chrysanthemum glasshouse • Although not damaging to crop many

other glasshouses nearby with cucumbers, aubergines and peppers

• Measures aimed to eradicate to prevent establishment

• What were the extra costs to the grower?

Page 9: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Estimated expenditure on invertebrate pest management in one glasshouse at Thrips palmi

outbreak site over one year (Nov. 1999 to Oct. 2000)

0

5,000

10,000

15,000

20,000

25,000

Nov Dec Jan Feb Mar Apr May June July Aug Sept Oct

Month

Ex

pe

nd

itu

re (

£)

Monthly costs Forecast without PHS

Page 10: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Estimated expenditure on invertebrate pest management in one glasshouse at Thrips palmi

outbreak site over one year (Nov. 1999 to Oct. 2000)

0

5,000

10,000

15,000

20,000

25,000

Nov Dec Jan Feb Mar Apr May June July Aug Sept Oct

Month

Ex

pe

nd

itu

re (

£)

Monthly costs Forecast without PHS

Thrips palmi diagnosed late April 2000

Page 11: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Estimated cumulative expenditure on invertebrate pest management in one glasshouse at Thrips palmi outbreak site over one year (Nov. 1999 to Oct. 2000)

0

5,000

10,000

15,000

20,000

25,000

Nov Dec Jan Feb Mar Apr May June July Aug Sept Oct

Month

Ac

cu

mu

late

d e

xp

en

dit

ure

)

Monthly costs Cumulative costs Forecast without PHS

Thrips palmi diagnosed late April 2000

Page 12: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Changes to producers profits (partial budgeting)

• Sales were unaffected• Extra production costs

• Pesticide spray costs • Soil fumigation (methyl bromide)• Treated compost• Plastic sheeting - additional labour• Costs up by approx. £15,000 (US$ 25,000)

• Margin fell by between 13 and 18%

Page 13: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Cost: benefit analysisCosts = costs of eradication• Industry costs

• determined from additional costs at outbreak site

• Govenment costs• staff costs during campaign

Page 14: Use of cost: benefit analysis  to inform risk management decisions  in PRA

What are the benefits ?

• Losses avoided

• Estimated by modelling

Page 15: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Modelling spread from the outbreak site

• Use Monte Carlo technique to simulate uncontrolled spread from outbreak site

• Consider two rates • Fast - similar to previous spread of

Frankliniella occidentalis • nationwide in 3 years

• Slow - based on T. palmi in Japan, • 10 years to occupy 2/3rds of the endangered

area

Page 16: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Crops at risk from Thrips palmi

Crop Area (ha)

Value HPM £’000

Potential loss %

Potential loss £’000

Cucumbers 172 38,539 10 3,854

Sweet peppers 48 7,799 8 624

Aubergine 11 2,548 15 382

Protected ornamentals

99 14,705 1 147

Total 330 63,191 5,007

Mean annual data, 2000-2001, Defra stats

US$ 107,000 US$ 8,500

Page 17: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Modelling impacts during spread

• 10 years of uncontrolled spread

• Susceptible crops incur losses • High impact (as in previous table)

• Low impact (1/10th of previous)

• Discount value of losses to present day

Page 18: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Cost: benefit ratios (eradication: losses avoided)

Fast

Rate of Spread

Slow

High impact 1: 19 1: 9

Low impact 1: 5 1: 4

• With fast spread, T. palmi reaches all susceptible glasshouse crops in 3 years (as with WFT).

• With slow spread 2/3rds of glasshouses infested after 10 years.

Page 19: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Example 2: Implementation of EU measures

against Diabrotica virgifera virgifera in the UK

• First reported in Europe near Belgrade airport 1992

• Spreading across Europe

• Damaging in permanent maize

• Listed quarantine pest

Page 20: Use of cost: benefit analysis  to inform risk management decisions  in PRA
Page 21: Use of cost: benefit analysis  to inform risk management decisions  in PRA

5 km2 cells with accumulatedtemperature > 670 = 4852

Fig. A3(vi) Fig. A3(vii)

Red, pink and purple cells show where Dvv can establish

• Single generation requires 670 DD above 11°C• Climate is critical

Area suitable for establishment

Cool (1996) Typical (1997) Hot (1995)

Page 22: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Fig. A3(viii) Area suitable for establishment: sufficient temperature and maize (1996, cool)

Page 23: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Fig. A3(ix) Area suitable for establishment: sufficient temperature and maize (1997, typical)

Page 24: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Model details (1 -3)

With EC control measures– 3 Spread scenarios

• 0 km year-1 (costs of surveys only)• 0 to 1.5 km year-1

• 0 to 4 km year-1

– Monte Carlo simulation (10,000 iterations)– Area suitable for establishment (<100 - 120k ha)– Insecticide sprays = £46 ha-1 (US$ 78 ha-1)– No maize in field for 2 years – 80% of area rotated at no extra cost– Costs from rotation (£182 ha-1 to £243 ha-1)– Costs from rotation (US$ 300 ha-1 to US$410 ha-1)

Page 25: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Model details (4 -6)

Without EC control measures– 3 rates of spread

• 0 to 15 km year-1

• 5 to 25 km year-1

• 10 to 40 km year-1

– Monte Carlo simulation (10,000 iterations)– Area suitable for establishment (<100 - 120k ha)– 20% not rotated– Impacts after 5 years– Maize worth £375 to £450 ha-1 (US$ 630 to US$ 760)– Yield losses 2 to 5% (6.5% in NE USA)– Model looks ahead 10 years

Page 26: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Area suitable for development:Example 1

0

20

40

60

80

100

120

1 2 3 4 5 6 7 8 9 10

Year

Are

a su

itab

le (

'000

ha)

Page 27: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Area suitable for development:Example 2

0

20

40

60

80

100

120

1 2 3 4 5 6 7 8 9 10

Year

Are

a su

itab

le (

'000

ha)

Page 28: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Area suitable for development:Example 3

0

20

40

60

80

100

120

1 2 3 4 5 6 7 8 9 10

Year

Are

a su

itab

le (

'000

ha)

Page 29: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Annual spread no control: Example 1

0

10

20

30

40

1 2 3 4 5 6 7 8 9 10

Year

An

nu

al s

pre

ad (

km/y

ear)

Page 30: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Annual spread no control: Example 2

0

10

20

30

40

1 2 3 4 5 6 7 8 9 10

Year

An

nu

al s

pre

ad (

km/y

ear)

Page 31: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Annual spread no control: Example 3

0

10

20

30

40

1 2 3 4 5 6 7 8 9 10

Year

An

nu

al s

pre

ad (

km/y

ear)

Page 32: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Area occupied no control: Example 1

0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Year

Are

a o

ccu

pie

d (

'000 h

a)

Example Mean (10,000 runs)

Page 33: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Area occupied no control: Example 2

0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Year

Are

a o

ccu

pie

d (

'000 h

a)

Example Mean (10,000 runs)

Page 34: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Area occupied no control: Example 3

0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Year

Are

a o

ccu

pie

d (

'000 h

a)

Example Mean (10,000 runs)

Page 35: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Costs of measures

Spread (1)

0

Spread (2)

0 - 1.5

Spread (3)

0 – 4

Govt + industry costs (£’000)

3,477 6,559 – 7,431 12,713-15,591

• Govt. costs (surveys, implementing measures) - based on staff costs on Dvv to date

• Industry costs (treatments on infested fields, forced rotation)

US $ (‘000) 5,900 11,100 – 12,500 21,500 – 26,300

Page 36: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Yield losses that are avoided

Impact Spread (4)

0 –15

Spread (5)

5 –25

Spread (6)

10 –40

Value of 2% yield loss (£’000)

193 -232 573 – 688 887 -1,065

Value of 5% yield loss (£’000)

482 -579 1,433 -1,720 2,218 -2,661

US $ (‘000) 820 – 980 2,400-2,900 3,750 – 4,500

Page 37: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Cost : benefit ratios (e.g. Scenario 2 vs 4, 5 & 6)

Spread rate

5% loss in yield 2% loss in yield

0 –15 14:1(6,559: 579)

(7,431: 482)

34:1

5 –25 5:1 11:1

10 -40 3:1 7:1

Page 38: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Conclusions

• In UK no economic justification for EC measures

• However consider assumptions– Faster spread– Greater yield losses– Climate change

Page 39: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Strengths & weaknesses of cost: benefit analysis

• Uses a single metric ($, £, €) • Is easy to understand

– Easy to communicate

• Takes account of many aspects • However, lack of data means assumptions

are necessary• Difficulty in assessing non-market goods• In simplifying results to a ratio, details are

lost

Page 40: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Challenges

• PRA workers often reacting to events– interceptions, incursions, little information

• Key in assessing risk is identifying uncertainty – Research may reduce uncertainty (time

constraints) – Probabalistic risk assessment may quantify

uncertainty

• Key to risk management is managing the uncertainty

Page 41: Use of cost: benefit analysis  to inform risk management decisions  in PRA

Thank you