Economic Impact of Drought on Balanced Development of ...wci.t.u-tokyo.ac.jp/events/awcs2016/presen/5-3.pdf · Pakistan (ROP) 6 Pakistan Fig. Map of Pakistan Table Recorded drought

Economic Impact of Drought on

Balanced Development of Pakistan:

Inter-disciplinary Approach of

Drought Risk Assessment

Asian Water Cycle Symposium 2016,

University of Tokyo,

March 1-2, 2016

Muneta Yokomatsu, Yushi Suzuki, Hiroaki Ishiwata, Yohei Sawada,

Asif Naseer, Muhammad Jehanzeb Masud Cheema, Toshio Koike

Impact on Drought Stress

� Drought is one of meteorological disasters that result in

serious damage to the economic activity around the world.

� Serious damage occurs in the agricultural sector, the

impact will spread to households and other sectors.

� Lack of water and food caused by drought, and

deterioration of sanitary condition lead to social unrest

such as immigration and conflict. Human lives are lost due

to large-scale famine and disease in the worst case.

1

Introduction

Reference: REUTERS (Feb. 4, 2011)

Fig. Agricultural land during drought

Table Classification and definition of drought

Classification of drought Definition

Meteorological drought Periods that precipitation is extremely

small continue.

Hydrological drought Periods that the water levels such as

streamflow, water reservoir, and

groundwater are extremely low continue.

Agricultural drought Agricultural production is decreased

mainly by lack of soil moisture.

Economic Aspect of the Drought Risk

�While the integration of knowledge about drought is

progressing in the field of meteorology, hydrology, agriculture,

etc., socio-economic risks are not recognized correctly.

� Since a part of the drought damage is mainly indirect, the total

damage should be simulated with consideration of market

transaction.

• E.g.) Data of total damage such as crops and livestock in EM-

DAT database does not include indirect effects.

Purpose

�Quantitative analysis of economic impact of drought, by

integrating a hydrological model and an economic model

Purpose

Hydrological Model: Hydrological-Dynamic Vegetation Model

� Express the process that meteorological drought causes

agricultural drought.

Economic Model: Multi-Sector and Multi-Region Model

� Evaluate the impact that agricultural drought affects overall

economy.

3

Framework

Fig. Research framework

Reference:

Basic Framework of General Equilibrium Model

Households Firms

Capital

Product Markets

Land, etc.

Factor Markets

Labor

Government

Rest of the world

Public goods,

Transfer

Public

goods

Taxes

Intermediate

demand

ProvisionConsumption

Savings

Imports Imports

Exports

Government

savings & debts

Investment

demand

Transaction of bonds

Government

consumption

4

Reference:

Disaster, Mitigation Policy, and Economic Growth

5

GDP

without disaster

Disaster

Event

With DRR Investment

（with disaster）

Without DRR

Investment

（with disaster）

Time

Effect of DRR

Gap expanding!

What is the macroeconomic benefit of the investment in disaster risk

reduction (DRR)?

Focus of the study

� Frequent occurrence of draught.

� Three regions in the model: Punjab, Sindh, the Rest of

Pakistan (ROP)

6

Pakistan

Fig. Map of Pakistan

Table Recorded drought in Pakistan (1871-2000)

Reference: Ahmad et al. (2004)

Table Overview of main provinces of Pakistan

Name of Province Drought Occurrence Year

Punjab 1899, 1920, 1935

Khyber Pakhtunkhwa 1902, 1951

Sindh 1871, 1881, 1899, 1931, 1947, 1999

Service

Three-Sector-Three-Region Closed-Economy Model

7

Pakistan

Rest of Pakistan (ROP)

Service

Punjab

Service

Manufacturing

Sindh

AgricultureAgriculture

Manufacturing

ManufacturingAgriculture

Labor Capital

Market of Service and Land is closed in each region.

Land,

Water

Land,

Water

Land,

Water

Irrigable Land

�Agricultural land which is provided water by irrigation

facilities

Rainfed Land

� Agricultural land which depends only on precipitation

8

Agricultural Land

Photo. Irrigable land

Source: Pakistan Today (Sept. 21, 2012)

Leaf Area Index (LAI)

� The ratio of the total area of one side of leaves to the unit

land area

� Higher LAI is, more vegetation exists, namely, more

preferable agricultural land is. (A positive correlation

between annual maximum LAI and agricultural production

is verified).

9

Leaf Area Index (LAI)

Fig. Comparison of LAI observed value by GLASSLAI and LAI simulation results by AgriCLVDAS (winter)

Indicator of LAI composite

� Weighted sum of LAI of several representative crops across several

periods

� Applied to calibrate parameters of the land-water composite

� The total amount of available water in the irrigable land is assumed

by the sum of precipitation, irrigation water, and ground water.

� Performance level of irrigation facilities is treated as a policy variable.

10

The Model (1): Leaf Area Index (LAI)

Available waterAvailable land

Precipitation

River water Groundwater

Performance

Level of canal

irrigation

Performance level

of underground

irrigation

Available water in irrigable land:

Available water in rainfed land:

Land-water composite

Indicator of LAI composite

Production Technology of Agricultural Sector

� Value added of agricultural production is composed of labor, capital and

land-water composite.

� Two types of the land-water composite for irrigation land and rainfed land.

Production Technology of Manufacturing and Service Sector� Factor of production of value-added function of manufacturing sector: labor, capital

and land

� Factor of production of value-added function of service sector: labor, capital

� Both value-added functions are specified by Cobb-Douglas form.

� Intermediate goods are input in a way of Leontief function. 11

The Model (2): Production Technology

Intermediate

goods

(Agriculture)

Intermediate

goods

(Manufacturing)

Intermediate

goods

(Service)

Value added (Agriculture)

Labor Capital L-W composite

of irrigable landL-W composite

of rainfed land

1. Precipitation rank as random variable is determined.

Event Sequence and Equilibrium

Period t Period t+1

①Determination ofprecipitation rank

②Household activity(Supply of production

factors, consumption, etc.)

③Firms activity(Demand of production

factors, production, etc.)

④Market equilibrium ⑤Increase of total population

and total amount

of capital

⑥The same cycle as period t

Time

Precipitation

rank

Probability

(%)

Precipitation (km3/year)

Punjab Sindh ROP

1 5.3 37.1 1.9 62.8

2 22.8 48.7 8.9 81.5

3 42.1 60.3 15.8 100.3

4 28.1 76.7 31.8 125.0

5 1.8 93.1 47.8 149.8

In each period,

2. Households supply their labor and capital to firms,

consume final goods, and save fixed percentage of income.

13

Event Sequence and Equilibrium

Period t Period t+1

①Determination ofprecipitation rank

②Household activity(Supply of production

factors, consumption, etc.)

③Firms activity(Demand of production

factors, production, etc.)

④Market equilibrium ⑤Increase of total population

and total amount

of capital

⑥The same cycle as period t

Time

max �(�)

�� = (1 − �){ + �� + ∑ �}subject to

� : consumption bundle (vector), �: price vector,

: wage, �: interest rate, �: capital, �: rent income

�: Utility

IncomeConsumption

rate

Budget constraint

3. Firms demand the factors of production and the

intermediate goods, and produce the final goods.

Event Sequence and Equilibrium

Period t Period t+1

①Determination ofprecipitation rank

②Household activity(Supply of production

factors, consumption, etc.)

③Firms activity(Demand of production

factors, production, etc.)

④Market equilibrium ⑤Increase of total population

and total amount

of capital

⑥The same cycle as period t

Time

Y=min ! ", $, % , Intermediate inputssubject to

max Π(* ", $) = +,* − " − �-K

/ : Production function, *: products, +,: value-added price, : wage, �-: interest rate, ": labor, $: capital, X: Land-water composite

*, ", $

Profit

Production technology

4. All the markets (products markets and factor markets)

clear at the equilibrium: demands and supplies are

balanced at the equilibrium price .

Event Sequence and Equilibrium

Period t Period t+1

①Determination ofprecipitation rank

②Household activity(Supply of production

factors, consumption, etc.)

③Firms activity(Demand of production

factors, production, etc.)

④Market equilibrium ⑤Increase of total population

and total amount

of capital

⑥The same cycle as period t

Time

price

quantity

Demand

Supply

�, , �, �

Event Sequence and Equilibrium

5. Population increases, and total amount of capital is

increased by investment (=households’ savings).

6. Move to the next period, and repeat the cycle from 1.

16

Period t Period t+1

①Determination ofprecipitation rank

②Household activity(Supply of production

factors, consumption, etc.)

③Firms activity(Demand of production

factors, production, etc.)

④Market equilibrium ⑤Increase of total population

and total amount

of capital

⑥The same cycle as period t

Time

Available water

�A sample case that most severe drought (the scale of

once-in-50 years) occurs in 3rd, 7th and 12th period

17

Simulation Results: Sample Path Analysis (1)

Severe droughtLarge precipitation

Table. Precipitation ranks, probabilities and precipitation

※ Base year (first period) is set at 2010. Exogenous technological development rate of labor is 0%.

Fig. Sample path of precipitation and water of canal irrigation

Precipi-

tation

rank

Probability

(%)

Precipitation (km3/year)

Punjab Sindh ROP

1 5.3 37.1 1.9 62.8

2 22.8 48.7 8.9 81.5

3 42.1 60.3 15.8 100.3

4 28.1 76.7 31.8 125.0

5 1.8 93.1 47.8 149.8

18

Production

�Agricultural production drops in drought period.

Fig. Transition of Agricultural production

Simulation Results: Sample Path Analysis (2)

19

Simulation Results: Sample Path Analysis (3)

Fig. Transition of goods of each sector

(Price of manufacturing goods is set at numéraire.)

Prices

�Decrease of Ag-production causes price increase of Ag-

goods.

Ag-price drops in

period of high

precipitation.

Sectoral GDP (= Value-added price x Production )

�GDP of agricultural sector does not decrease so much

in draught period because of the price increase.

�Manufacturing sector and service sector suffer indirect

damages through transaction of intermediate goods.

20

Simulation Results: Sample Path Analysis (4)

GRP (Regional GDP)

�Negative impact on GRP by drought is large in Punjab

and Sindh because of large indirect effect on Non-Ag

sectors.

21

Simulation Results: Sample Path Analysis (5)

Fig. Transition of regional GDP

GDP and utility

� GDP that includes both direct and indirect impacts of draught

drops in drought periods.

� Owing to the cancelling effect of price, fluctuation of GDP is mild,

while paths of GDP per capita and utility (function of absolute

amount of consumption) is more sensitive.

22

Simulation Results: Sample Path Analysis (6)

Fig. Transition of GDP

Fig. Transition of GDP per capita Fig. Transition of utility

Comparative dynamics with respect to irrigation efficiency

�Higher irrigation efficiency attains higher economic growth.

23

Policy Analysis 1: Improvement of Irrigation Facility

Fig. Transition of GDP Fig. Transition of GDP per capita

Reallocation of river water between Punjab and Sindh (-20% ~ +20%) by

dam control

24Fig. GRP growth of Punjab

Policy Analysis 2: Regional reallocation of river water

Fig. GRP growth of Sindh

v

Enlarged viewEnlarged view

v

� Although Punjab has already been allocated more water than

Sindh, further transfer from Sindh to Punjab increases GDP.

� Considering sizes of farm lands, current policy seems to allocate

too much water to Sindh.

25

Policy Analysis 2: Regional reallocation of river water

Fig. GDP per capita

v

Enlarged view

� Integration of a hydrological model and an economic growth

model

• Leaf Area Index (LAI), derived in the hydrological-vegetation model,

was incorporated into production function of the economic model.

Hydrological results provided a rationale for relationship between

water resources and land.

�Quantitative evaluation of economic impacts of drought and

irrigation facilities

• With stochastic precipitation process, economic growth paths, both

regional and sectoral, were derived as well as effects of improvement

of irrigation facilities and regional reallocation of river water.

26

Summary: Results

Summary: Future works

�Monte-Carlo simulation

�Dependency of productivity of land-water composite on

sequential drought

27

For example,

Dummy variable

1)

2)

If and ,

, otherwise 0.

Summary: Future works

� Stepwise improvement of irrigation facilities and

consideration of construction costs

2828

Precipitation

River water Groundwater

Available water in irrigable land:

Available water in rainfed land:

28

Level of

canal

irrigation

Level of

underground

irrigation

For example,

Depreciation Investment

Cost :

29

Thank you very much!