Transcript Analysing environmental issues in an open economy

875-6.ppt
6 Analytical extensions and
policy issues
1. Model of development in a resourceintensive economy (from class 5)
1. Model structure
2. Comparative statics: trade and policy shocks
2. Analytical extension: H-O manufacturing
3. Some related topics and papers
1
875-6.ppt
Forest
& forestry
Upland economy
H-O upland
economy
Land
Labor
Upland food
Food
market
Tree crops
Lowland economy
Land
Capital
Lowland food
Ricardo-Viner
lowland economy
Labour
market
Manufacturing
Labor
2
875-6.ppt
Cerebral comparative statics
1. Effects of price changes, e.g.:
•
•
Tree crops in uplands (H-O prediction?)
Manufacturing price increase, e.g. tariff. (Cf.
Deacon, JEEM 1995).
2. Effects of tech. progress and factor
endowment growth
•
•
‘Green revolution’ in lowland food sector
Capital investment in manufacturing
3
875-6.ppt
A mathematical statement
•
Fundamental assumptions about
technology, preferences
General equilibrium
Intersectoral and interregional marketclearing conditions
•
•
•
•
•
Spatial: labor market & migration
Food market clearing when non-traded
Open access to forest resource for land
4
875-6.ppt
• Two regions, upland and lowland, with p rice-taking r evenue-maximi sing
producers.
• In uplands, two goods are produced: food (n) and non-food, or tree crops;
vectors of upland prices and outputs are PU and YU respectiv ely. Defin e p
as relativ e price of food to tree crops in uplands.
• Upland goods produced with VU input s, containing labor (LU) and upland
land (T).
• Land must be cleared for production, using labor according to T = L.
• Upland producer's problem is captured by revenue function:
U
U
U
R(T, LU – T, p) = max
{P

Y
|
V
}.
T ,Y
• 'Default ' assumption upland food is labor-int ensive, or RnL > 0; RnT < 0
5
875-6.ppt
• Lowland region produces food (n) and manufactures (m) in output vector
YL. Price vector PL has elements p and q, where the latter is the price of
manufactures.
• Lowland food is different to that used in upl ands.
• Each lo wland indust ry uses a sector-specifi c factor (irrig ated land and
manufacturing capital), which we summarize as a vector K = (Kn, Km), and
labor.
• Lowland producer's problem
L
L
L
S(K, L – LU, p, q) = max
{P

Y
|
V
}.
Y
6
Consumption. Assumption s:
875-6.ppt
• Utility deriv ed from consumption of goods and from exis tence of forest.
• Forest-clearing decision takes no account of consumer preferences so
quantity of land cleared is exogenous to the consumer.
• Consumer’s problem is to maximi ze utility subject to income and the
quantity of standing forest (utility is assumed separable between these).
• Forest is quantity-rationed to consumer, so we have conditional
expenditure function:
E(P, F, u) = min {PC | u}
where u =utility, F is quantity of forest land, and P and C contain
prices and quantiti es of food, tree crops and the manufactured good.
* T = 1 – F. ET,= virtu al price of land cleared, or –1* margin al amount the
consumer is WTP to preserve standing forest. Therefore, we have ET Š 0.7
875-6.ppt
Equilibrium
• Assume manufactures (upland non-food, food) to be import-competing
(exportable, non-traded) with price q (1, p). q is exogenous, p endogenous.
Aggregate budget constraint:
E(p, q, T, u) = R(T, LU–T, p) + S(K, L–LU, p, q).
8
875-6.ppt
FONC:
Assume that food is not internationally traded, so at optimum:
Rn + Sn = En.
(4.2)
Forest clearing for upland agricultu re, at optimum:
RT  RL = 0.
(4.3)
Note: assumption that forest clearing de cisions do no t cons ider social costs. Hence in
equili brium t here is m ore tree-clearing than is socia lly optim al, which con fers a neg.
externalit y on consu me rs.
Labor mig rates between regions in response to changes in productivity , so::
RL  SL = 0.
(4.4)
• The solution to (4.1)-(4.4) provid es values for endogenous variables T and
LU, p, and u, giv en exog. variables L, K, q.
9
875-6.ppt
Actual comparative statics
• Terms of trade shocks
• Tariff policy reforms
10
875-6.ppt
ToT shock: a rise in q
(a) When food is a traded good and labor is regionally mobile
Eq. (4.2) does not hold. Taking tot al diff of (4.3 ) and (4.4) and solving:
( RnT  RLn )dp


Rvv
(RLT  RLL ) dT  

 u   S dq  ( R  S )dp  S dL  S dK 
(R


R
)
(R

S
)
dL 
 LT
 Lq
Ln
Ln
LL
LK

LL
LL
LL 


 

The determin ant of the coefficient matrix , DL, is positive by the strict
concavity of the revenue function. An in crease in q giv es:
dLU dq  RvvSLq / D L  0
dT dq  (RLT
where 
(4.6a)
dLU 
 RLL )SLq / D    
0

dq


L
(4.6b)
 RLT  RLL 
 
 0
R


vv
High er labor productivity in lowlands causes down-slope migr ation; high er
labor costs and diminish ed upland labor supply both cause the quantity of
11
upland cleared for agriculture to dimini sh.
875-6.ppt
ToT shock (food is traded good and labor is regionally mobile, cont’d)
Welfare change
Denote the excess demand for manufactures by
Zq = Eq(p, q, T, u) – Sq(K, L–LU, p, q),
and noting that Zq > 0 for a net import and Zq < 0 for a net export.
Taking an in crease in q we obtain:
Eu
u
T
 ET
  Zq < 0.
q
q
(4.10)
12
875-6.ppt
ToT shock
(b) when labor is immob ile but food is non-tr aded
Eq (4.4) does not hold. By total differentiation of (4.1)–(4.3):
 Eu
du   Zq dq  RL dLU  SK dK  SLdL 
0
ET


  
U
Znn
(RnT  RLn )dp  Znq dq  RnLdL  SnK dK  SnLdL .
 Eu

 0
dT  
(
R


R
)
R
0
vv
nT
Ln



  

The determin ant of the coefficient matrix , Dp < 0. For an in crease in q:


dp dq   Znq  mZq Rvv / D p
> 0 if m = 0
(4.13a)
dT dq  Z nq  mZq RnT  RLn / D p < 0 if m = 0
(4.13b)


13
875-6.ppt
Tariff reform in manufacturing
Equilibrium:
E(p, q, T, u) = R(T, LU–T, p) + S(K, L–LU, p, q) + tZq,
(4.30)
(a) When food market clears through trade, results are as for terms of trade
shock (4.6).
14
875-6.ppt
Tariff change
(b) When labour is immobil e and food market clears through domestic price
adjustment,


Znq  mtZqq Rvv
dp

dt
Dp, t
> 0 if m = 0
(4.33a)
 R  RLn dp 
dT
   nT


dt
Rvv

 dt 
< 0 if m = 0.
(4.33b)
Overall welfare when the tariff rate is altered depends on int eractions
between the trade policy and the environmental externality . From (4.30),
du
dT
dp
.
Eu (1 tcM )  ET
 tZ qq  tZnq
dt
dt
dt
or, using (4.33b ) to elimin ate Žp/Žt,

Znq RnT  RLn dT
du
< 0.
Eu (1 tcM )  tZqq  ET  t



dt
Rvv

 dt
(4.34)
15
875-6.ppt
Other comparative static
experiments
• ‘Green revolution’ in lowland agriculture
• Other technical change shocks
• Exogenous growth of endowments:
– Labor
– Mfg. capital stocks
– Lowland land stocks
16
875-6.ppt
Concluding remarks
• U-L model combines two ‘small’ models to
obtain richer specification and results
• Predictions of comparative static effects
depend on key parameter values
– Can define different economic ‘types’ based on
alternative parameter sets (see OEE Chapter 3)
• Empirical and micro research should guide
structural and parameter assumptions.
<--17
875-6.ppt
Analytical extension: H-O
manufacturing sectors
• Heckscher-Ohlin manufacturing sector, RV-J agriculture (Coxhead/Jayasuriya 2003)
• Simplification: only one upland sector
• Extension: clean and dirty mfg industries
• All traded goods (for simplicity!)
• Other aspects same as before:
• Open access to forests for ag. land
• Migration is costless and immediate
18
875-6.ppt
Two regions, four sectors
A= agricultural ‘region’, producing upland (U) and food (F) crops.
M = manufacturing ‘region’, producing labor-intensive (X) and
capital-intensive (H) goods.
cj = unit cost (i.e. zero profit) frontier for industry j
wA
cF cU
rU
rF
LF
0A
wM
LA
LM
L
LH
cX c H
0M
rM
19
875-6.ppt
Price increase for upland
agriculture
wA
cU cF
rU rF
LF
0A
wM
LA
LM
L
LH
cX cH
0M
rM
Increase in upland agr. price increases upland land demand &
returns to upland land, reduces rF, rM and industrial output
20
875-6.ppt
Tariff reduction for importcompeting industry
wA
cU cF
rN
rT
wM
LF
0A
LA
LM
L
LH
cX cH
0M
rM
NB.: Assumes large employment share of EOI industry
21
875-6.ppt
Some conclusions
• Extended model provides different
prediction on NR effect of tariff change
• More generally, issue of multiple env./NR
problems
– And added complexity of establishing welfare
results in economy with multiple distortions,
insufficient instruments
22
875-6.ppt
Some related topics and issues in
analytical models
1. Individual and firm motivations for
changes in NR or env. use patterns
•
•
Subsistence farmers and ag. land expansion
(Indonesia): Angelsen, WD 1995
‘Household EKC’ and cooking smoke
(Pakistan): Pfaff, manuscript, 2003.
23
875-6.ppt
Other topics and issues
2. Sectoral policies other than trade measures
•
Commercial farmers and deforestation
(Brazil): Binswanger, WD 1991
•
•
Ag. subsidies, land titling measures
Capital subsidies, poverty interactions (Latin
America): Lopez, WD, Feb. 2003
24
875-6.ppt
Other topics and issues
3. Macro measures and shocks
•
•
Structural adjustment and stabilization
measures: Munasinghe and Cruz, WB Env.
Paper #10, 1995
Exchange rate instability and depreciations:
Sunderlin et al., WD 2001; Coxhead et al.,
Land Econ. 2001
25
875-6.ppt
Other topics and issues
4. Voluntary and informal controls
•
•
Informal regulation of industrial pollution
(Indonesia): Pargal and Wheeler, JPE 1996
Legal and community institutions in lake
management (Philippines): Malayang, 1993
26