Cost-Benefit Analysis: Advanced Training
Presentation by Dr. Benoit Laplante
Bangkok, Thailand
November 27-28, 2017
Session 3:
Estimating cost of floods and benefits of flood mitigation
Estimating cost of floods
What are the impacts of floods and what are the costs of these
impacts? For example:
Impacts Costs
Damages to infrastructure Repair expenditures
People cannot go to work Lost productivity
Health and sickness Cost of treatment
Agricultural production Value of lost agriculture
Loss of lives ???
We can estimate
these costs…even
loss of lives.
Without CC No
project
Return
period
1/20
1/50
1/100
D1
D3
D5
Estimating cost of floods
Samoa
Estimating cost of floods
1-in-1 1-in-5
1-in-20 1-in-100
Estimating cost of floods
Without CC No
project
Return
period
1/20
1/50
1/100
D1
D3
D5
Expected
cost of flood
1/20*(D1)= + 1/50*(D3)
+ 1/100*(D5)
What is the expected cost of flood?
Estimating cost of floods
Cost
Return period1/201/501/100
D5
D3
D1
Exceedance probability function
Estimating cost of floods
Cost
Return period1/201/501/100
D5
D3
D1 Expected cost of CC
Exceedance probability function
Estimating cost of floods
Without CC No
project
With CC
No project
Return
period
1/20
1/50
1/100
D1 D2
D3 D4
D5 D6
Impact (cost) of
CC on flood
D1D2 -
D3D4 -
D5D6 -
Estimating cost of floods
Cost
Return period1/201/501/100
D5
D3
D1
Exceedance probability
function without CC
Estimating cost of floods
Exceedance probability
function with CC
D6
D4
D2
Without CC No
project
With CC
No project
Return
period
1/20
1/50
1/100
D1 D2
D3 D4
D5 D6
Impact (cost) of
CC on flood
D1D2 -
D3D4 -
D5D6 -
What is the impact of CC on the expected cost of flood?
Estimating cost of floods
Without CC No
project
With CC
No project
Return
period
1/20
1/50
1/100
D1 D2
D3 D4
D5 D6
Impact (cost) of
CC on flood
D1D2 -
D3D4 -
D5D6 -
Expected
cost of CC
1/20*(D2-D1)= + 1/50*(D4-D3) + 1/100*(D6-D5)
Estimating cost of floods
Cost
Return period1/201/501/100
D5
D3
D1
Estimating cost of floods
D6
D4
D2
This
area
is
the
expected
cost
of CC
Please note:
These estimated damages depend (among other things) on:
 The number of people in the future living in the identified flood
areas; and
 The value of assets (public and private) in the future in the
identified flood areas.
In order to estimate damages of flood in the future, we need:
Hence:
 Population projections for the targeted area (we can start with
national population projections and assume that in the
urbanized targeted area, in fact population growth will be higher
than national population growth projection; and
 GDP projections (it may be assumed that property values (in real
terms) increase at the rate projected GDP growth rate.
Estimating cost of floods
Potential sources of information:
For population:
For GDP:
 United Nations, Department of Economic and Social Affairs,
Population Division
 National statistics
 World Bank GDP projections
 IMF GDP projections
 Ministry of Finance
Estimating cost of floods
Potential sources of information:
For estimates of damages:
 Hazard mapping
 History of flood events and impacts
Wish list: For every flood event in the last many years:
 Date of the event
 Location of the flood
 Duration of the flood (number of days)
 Area covered by the flood (in km2 or other measures)
 Number of people impacted (exposed, deaths, sickness)
 List of infrastructure damaged (number of houses, public
buildings, etc.)
 Repair costs
 List of other losses (agriculture, etc.)
Estimating cost of floods
Estimating benefits of flood mitigation
With CC
No project
D2
D4
D6
Return
period
1/20
1/50
1/100
With CC
With project
D7
D8
D9
Benefit of the
project
D2D7 -
D4D8 -
D6D9 -
What are the estimated benefits of flood mitigation?
Estimating benefits of flood mitigation
With CC
No project
D2
D4
D6
Return
period
1/20
1/50
1/100
With CC
With project
D7
D8
D9
Benefit of the
project
D2D7 -
D4D8 -
D6D9 -
Expected
benefit of
the project
1/20*(D7-D2)= + 1/50*(D8-D4) + 1/100*(D9-D6)
Cost
Return period1/201/501/100
Exceedance probability function
with CC without project
Exceedance probability
function with CC with project
Estimating benefits of flood mitigation
20
Let’s go back here
Sukhothai Uttaradit Phitsanulok
Return period Estimated
damages without
project
Estimated
damages with
project
Reduction in
damages
1-in-5 18,970,000 0 18,970,000
1-in-10 330,210,000 98,380,000 231,830,000
1-in-25 1,222,500,000 735,450,000 487,050,000
1-in-50 1,737,070,000 1,076,960,000 660,110,000
Estimated Reduction in Flood Damages (Thai Baht)
Estimating benefits of flood mitigation
Please calculate the expected benefits of the project on
flood mitigation.
Expected benefits: 59,661,200 Thai Baht
Return period Estimated
damages without
project
Estimated
damages with
project
Reduction in
damages
1-in-5 18,970,000 0 18,970,000
1-in-10 330,210,000 98,380,000 231,830,000
1-in-25 1,222,500,000 735,450,000 487,050,000
1-in-50 1,737,070,000 1,076,960,000 660,110,000
Estimated Reduction in Flood Damages (Thai Baht)
What happens if one benefit of the flood mitigation project is to
save lives? Do we calculate an economic benefit?
Estimating benefits of flood mitigation
Value of statistical life (VSL)
A flood mitigation project will reduce mortality risk in a given
population in Thailand from 1/100,000 to 1/200,000. What is the
benefit per statistical life saved?
Estimating benefits of flood mitigation
Meaning of VSL: It is the maximum willingness-to-pay that
society is willing to pay to reduce the risk of losing 1 life.
Key determinants of VSL: Income (or GDP per capita)
In the USA, the VSL is about $7.4 million
If risk preferences, discount rates, and survival probabilities
were the same in all countries, then the VSL should simply be
proportional to income:
VSL Thailand
= VSL USA *
YThailand
YUSA
Ɛ
Where Y is GDP per capita in PPP terms and Ɛ is income
elasticity.
Estimating benefits of flood mitigation
We use GDP per capita measured in PPP. The IMF’s World
Economic Outlook Database 2016 reports a GDP per capita of
$57,536 and $16,888 for the USA and Thailand respectively. This
provides a GDP per capita ratio of 0.294.
Step 3: Income elasticity. Assuming an income elasticity of 1,
then VSL in Thailand is: $2,234,640.
Estimating benefits of flood mitigation
Each time a flood mitigation project is expected to save 1
statistical live, then the economic benefit is $2.234 million per
statistical lives saved.
26
Let’s go back here
Sukhothai Uttaradit Phitsanulok
Estimating benefits of flood mitigation
Over the 9-year period between 2006 and 2014, a total of 44 lives were
lost to floods in Sukhothai, and 41 in Phitsanulok.
On an annualized basis, this would represent a total of approximately
4.9 lives lost in Sukhothai, and 4.6 in Sukhothai. Zero lives were lost
to floods in Uttaradit.
Using the population census of 2010, this would represent an
incidence of mortality of approximately 0.79 per 100,000 people in
Sukhothai, and 0.51 per 100,000 people in Phitsanulok.
What do we need to do next?
Scenario without project: How many lives in the future could be lost
as a result of floods in these two provinces?
Estimating benefits of flood mitigation
Assume mortality
incidence remains
the same
Population
projections
Year
Phitsanulok
Sukhothai
850,000
600,000
Number of
lost lives
Phitsanulok
Sukhothai
4.8
4.4
Estimating benefits of flood mitigation
What is the next question?
What could be the number of lives lost with the project?
Year
Number of
lost lives
With project
Without project
4.8
Year
Number of
lost lives
4.4
4.4
Sukhothai Phitsanulok
With project
Without project
From this step, we would have an estimate of the number of lives the
project could save in the future. Then we can apply VSL to these
estimates.
Cost-Benefit Analysis: Advanced Training
Presentation by Dr. Benoit Laplante
Bangkok, Thailand
November 27-28, 2017
Session 3:
Estimating cost of floods and benefits of flood mitigation

Thailand UNDP-GIZ workshop on CBA - Effective water management and sustainable agriculture in Thailand

  • 1.
    Cost-Benefit Analysis: AdvancedTraining Presentation by Dr. Benoit Laplante Bangkok, Thailand November 27-28, 2017 Session 3: Estimating cost of floods and benefits of flood mitigation
  • 2.
    Estimating cost offloods What are the impacts of floods and what are the costs of these impacts? For example: Impacts Costs Damages to infrastructure Repair expenditures People cannot go to work Lost productivity Health and sickness Cost of treatment Agricultural production Value of lost agriculture Loss of lives ??? We can estimate these costs…even loss of lives.
  • 3.
  • 4.
  • 5.
  • 6.
    Without CC No project Return period 1/20 1/50 1/100 D1 D3 D5 Expected costof flood 1/20*(D1)= + 1/50*(D3) + 1/100*(D5) What is the expected cost of flood? Estimating cost of floods
  • 7.
  • 8.
    Cost Return period1/201/501/100 D5 D3 D1 Expectedcost of CC Exceedance probability function Estimating cost of floods
  • 9.
    Without CC No project WithCC No project Return period 1/20 1/50 1/100 D1 D2 D3 D4 D5 D6 Impact (cost) of CC on flood D1D2 - D3D4 - D5D6 - Estimating cost of floods
  • 10.
    Cost Return period1/201/501/100 D5 D3 D1 Exceedance probability functionwithout CC Estimating cost of floods Exceedance probability function with CC D6 D4 D2
  • 11.
    Without CC No project WithCC No project Return period 1/20 1/50 1/100 D1 D2 D3 D4 D5 D6 Impact (cost) of CC on flood D1D2 - D3D4 - D5D6 - What is the impact of CC on the expected cost of flood? Estimating cost of floods
  • 12.
    Without CC No project WithCC No project Return period 1/20 1/50 1/100 D1 D2 D3 D4 D5 D6 Impact (cost) of CC on flood D1D2 - D3D4 - D5D6 - Expected cost of CC 1/20*(D2-D1)= + 1/50*(D4-D3) + 1/100*(D6-D5) Estimating cost of floods
  • 13.
    Cost Return period1/201/501/100 D5 D3 D1 Estimating costof floods D6 D4 D2 This area is the expected cost of CC
  • 14.
    Please note: These estimateddamages depend (among other things) on:  The number of people in the future living in the identified flood areas; and  The value of assets (public and private) in the future in the identified flood areas. In order to estimate damages of flood in the future, we need: Hence:  Population projections for the targeted area (we can start with national population projections and assume that in the urbanized targeted area, in fact population growth will be higher than national population growth projection; and  GDP projections (it may be assumed that property values (in real terms) increase at the rate projected GDP growth rate. Estimating cost of floods
  • 15.
    Potential sources ofinformation: For population: For GDP:  United Nations, Department of Economic and Social Affairs, Population Division  National statistics  World Bank GDP projections  IMF GDP projections  Ministry of Finance Estimating cost of floods
  • 16.
    Potential sources ofinformation: For estimates of damages:  Hazard mapping  History of flood events and impacts Wish list: For every flood event in the last many years:  Date of the event  Location of the flood  Duration of the flood (number of days)  Area covered by the flood (in km2 or other measures)  Number of people impacted (exposed, deaths, sickness)  List of infrastructure damaged (number of houses, public buildings, etc.)  Repair costs  List of other losses (agriculture, etc.) Estimating cost of floods
  • 17.
    Estimating benefits offlood mitigation With CC No project D2 D4 D6 Return period 1/20 1/50 1/100 With CC With project D7 D8 D9 Benefit of the project D2D7 - D4D8 - D6D9 - What are the estimated benefits of flood mitigation?
  • 18.
    Estimating benefits offlood mitigation With CC No project D2 D4 D6 Return period 1/20 1/50 1/100 With CC With project D7 D8 D9 Benefit of the project D2D7 - D4D8 - D6D9 - Expected benefit of the project 1/20*(D7-D2)= + 1/50*(D8-D4) + 1/100*(D9-D6)
  • 19.
    Cost Return period1/201/501/100 Exceedance probabilityfunction with CC without project Exceedance probability function with CC with project Estimating benefits of flood mitigation
  • 20.
    20 Let’s go backhere Sukhothai Uttaradit Phitsanulok
  • 21.
    Return period Estimated damageswithout project Estimated damages with project Reduction in damages 1-in-5 18,970,000 0 18,970,000 1-in-10 330,210,000 98,380,000 231,830,000 1-in-25 1,222,500,000 735,450,000 487,050,000 1-in-50 1,737,070,000 1,076,960,000 660,110,000 Estimated Reduction in Flood Damages (Thai Baht) Estimating benefits of flood mitigation Please calculate the expected benefits of the project on flood mitigation. Expected benefits: 59,661,200 Thai Baht
  • 22.
    Return period Estimated damageswithout project Estimated damages with project Reduction in damages 1-in-5 18,970,000 0 18,970,000 1-in-10 330,210,000 98,380,000 231,830,000 1-in-25 1,222,500,000 735,450,000 487,050,000 1-in-50 1,737,070,000 1,076,960,000 660,110,000 Estimated Reduction in Flood Damages (Thai Baht) What happens if one benefit of the flood mitigation project is to save lives? Do we calculate an economic benefit? Estimating benefits of flood mitigation
  • 23.
    Value of statisticallife (VSL) A flood mitigation project will reduce mortality risk in a given population in Thailand from 1/100,000 to 1/200,000. What is the benefit per statistical life saved? Estimating benefits of flood mitigation Meaning of VSL: It is the maximum willingness-to-pay that society is willing to pay to reduce the risk of losing 1 life. Key determinants of VSL: Income (or GDP per capita)
  • 24.
    In the USA,the VSL is about $7.4 million If risk preferences, discount rates, and survival probabilities were the same in all countries, then the VSL should simply be proportional to income: VSL Thailand = VSL USA * YThailand YUSA Ɛ Where Y is GDP per capita in PPP terms and Ɛ is income elasticity. Estimating benefits of flood mitigation
  • 25.
    We use GDPper capita measured in PPP. The IMF’s World Economic Outlook Database 2016 reports a GDP per capita of $57,536 and $16,888 for the USA and Thailand respectively. This provides a GDP per capita ratio of 0.294. Step 3: Income elasticity. Assuming an income elasticity of 1, then VSL in Thailand is: $2,234,640. Estimating benefits of flood mitigation Each time a flood mitigation project is expected to save 1 statistical live, then the economic benefit is $2.234 million per statistical lives saved.
  • 26.
    26 Let’s go backhere Sukhothai Uttaradit Phitsanulok
  • 27.
    Estimating benefits offlood mitigation Over the 9-year period between 2006 and 2014, a total of 44 lives were lost to floods in Sukhothai, and 41 in Phitsanulok. On an annualized basis, this would represent a total of approximately 4.9 lives lost in Sukhothai, and 4.6 in Sukhothai. Zero lives were lost to floods in Uttaradit. Using the population census of 2010, this would represent an incidence of mortality of approximately 0.79 per 100,000 people in Sukhothai, and 0.51 per 100,000 people in Phitsanulok. What do we need to do next? Scenario without project: How many lives in the future could be lost as a result of floods in these two provinces?
  • 28.
    Estimating benefits offlood mitigation Assume mortality incidence remains the same Population projections Year Phitsanulok Sukhothai 850,000 600,000 Number of lost lives Phitsanulok Sukhothai 4.8 4.4
  • 29.
    Estimating benefits offlood mitigation What is the next question? What could be the number of lives lost with the project? Year Number of lost lives With project Without project 4.8 Year Number of lost lives 4.4 4.4 Sukhothai Phitsanulok With project Without project From this step, we would have an estimate of the number of lives the project could save in the future. Then we can apply VSL to these estimates.
  • 30.
    Cost-Benefit Analysis: AdvancedTraining Presentation by Dr. Benoit Laplante Bangkok, Thailand November 27-28, 2017 Session 3: Estimating cost of floods and benefits of flood mitigation