Supply Constrained Multipliers With SUT

You might also like

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 15

SUPPLY CONSTRAINTS

IN IO AND SUT MODELS


ANALYSIS

IFPRI On-Line
2018
The Consequences of Supply
Constraints
 Have examined standard IO and SUT multiplier models
 They assume:
 Final demand is exogenously determined
 Prices are fixed
 Output is demand driven
 Fixed proportions production technology
 Supply can expand to satisfy any demand increase
 What if a sector is capacity constrained & cannot expand output?
 Other sectors using its output as inputs may not be able to expand
 But what happens to the multiplier  0?
Examples of Supply Constraints
 Wine exports & the local supply of grapes by agriculture
 After 1994 global demand for South Africa wine increased
 No increase in output of agriculture
 Export of premium wine  local shortage of cheap wine  imports
 New power station & the supply of coal
 No increase in output of coal
 Export of coal diverted to new power station
 Supply constrained multipliers not zero
 … because inputs can be sourced from other uses

 But likely to be lower


Setting up a Supply-constrained IO
Multiplier Model (SUT to follow later)
 Standard IO Model: demand driven model
 All Final demand is exogenously determined
 Remember how we assumed fixed final demand
 All Outputs are endogenously determined
 Remember how we solved for output

 Supply Constrained IO model:


 Make output of the constrained sector exogenous
 Make final demand for the constrained sector endogenous

 How can this be done???


Linear Algebra of the Basic Multiplier Model
Exogenous

x  I  A
1
f
For three industries this can be written as:

bar
X1 F1
bar
X2 = (I-A)-1 • F2
bar
X3 F3
Find equivalent to the Leontief
Instead, challenge is to find a new matrix such that output in industry 1 is fixed
that satisfies the new relation
bar
X1 F1
bar
F2 = ?? • X2
bar
X3 F3
Algebra for sector 2 to be constrained
Set up basic (unconstrained) solution
x  A xf xA x  f x I  A  f
(1  a11 ) X 1  a12 X 2  a13 X 3  F1
 a21 X 1  (1  a22 ) X 2  a23 X 3  F2
 a31
31 X 1  a32 X 2  (1  a13 ) X 3  F3

Swap…for the 2nd sector to be output constrained


Rewrite: exogenous to rhs & endogenous to lhs
(1  a11 ) X 1  0  a13 X 3  F1  a12 X 2
a21 X 1  F2  a23 X 3  (1  a22 ) X 2
a31 X 1  0  (1  a13 ) X 3  F3  a32 X 2
Solution for constrained model
Now, rewrite in Matrix Algebra
 (1  a11 ) 0 a13   X 1   1 a12 0   F1 
     
  a21 1  a23 F
 2    0  (1  a22 ) 0  X2 
 a 0 (1  a33 )   X 3   0 1   F3 
 31 a32
M mendo  N mexo
Solution
mendo  M 1 N mexo
1
 X 1   (1  a11 ) 0 a13  1 a12 0   F1 
      
F
 2    a21 1  a23   0  (1  a22 ) 0  X2 
 X   a 0 (1  a33 )  0 1   F3 
 3  31  a32
Comparison with unconstrained
1
Comparing constrained solution mendo  M N mexo
1
 X 1   (1  a11 ) 0 a13  1 a12 0   F1 
      
F
 2    a21 1  a23   0  (1  a22 ) 0  X2 
 X   a 0 (1  a33 )  0 1   F3 
 3  31  a32

…..with base / unconstraint solution (multiplied by I)


1
 X 1   (1  a11 ) a12 a13   1 0 0   F1 
      
X
 2    a21 (1  a22 )  a23   0 1 0   F2 
 X   a    0 0 1 F 
 3  31 a32 (1 a33 
)   3 
We swap:
1. columns in Leontief (A) and I and changes their signs
2. entries in x and f
Now for small changes
Apply to small D in mixed exogenous variables

1
m endo  M N m exo
1
 X 1   (1  a11 ) 0 a13  1 a12 0   F1 
      
F
 2    a21 1  a23   0 (1  a22 ) 0   0 
 X   a   0  0 
 3  31 0 (1 a33 
)  a 32 1  

1
 X 1   (1  a11 ) 0 a13  1 a12 0  0 
      
F
 2    a21 1  a23   0  (1  a 22 ) 0 X
 2 
 X   a   0  0 
 3  31 0 (1 a33 
)  a 32 1  
SUPPLY CONSTRAINTS: SUT
MODEL
The Basic Set-up
 In SUT can now choose to constrain
 activities
and/or
 commodities
 Activity constraints = capacity bottlenecks
 Commodity constraints = production and import constraints
 Initially assume only activities can be supply constrained
 …since commodities can be imported
Next Step: Supply Constrained in SUT

mendo   M N  m exo compare with x   I  B  f


1 1

In which
• 𝐌−1 ∙ 𝐍 replaces 𝐈 − 𝐁 −1
• 𝐦𝐞𝐱𝐨 vector of exogenous variables
• With at least one element of activity output
• Allows for exogenous change in supply (activity output)
• 𝐦𝐞𝐧𝐝𝐨 vector of endogenous variables
• With at least one element of activity final demand
• Contradiction: as activities do not have final demand
• Convert to ΔFD of commodities with supply coefficients
Exercises using same SUT as multiplier model
SUT Supply Constrained Multiplier Exercise 6 Comm.xls sheets:
1. Methodology
2. Supply Constr Multiplier Model
 I, B, M and N are given
 Check toggle, conditional format & if-statement to set-up M and N
 Calculate constrained inverse with constrained act: AUTL
 Run scenario: 10% increase in exports of CHMF commodities
 Also calc unconstrained (basic Leontief) inverse (I-B) for comparison

SUT Supply Constrained Multiplier Exercise 21 Comm.xlsx:


 Similar structure as 6 comm sheet
 Calculate constrained inverse with constrained act: AELW
 Run scenario: 10% increase in exports of CIAS commodities
Bonus in 6 comm solution (only):
Hypothetical Extraction of Activity
Related question is sometimes raise:
 What is the impact of an activity closing down
 Iron & Steel Industry in SA
1
 Full shut down  1  a11  0 a13 
 
 Check M&B L HX  0 1 0 
 Second activity  a 0 1  a33  
 31

 Can compare multipliers (column totals) as % of base multipliers


 Which activities are impacted most
 Shut down to a degree, say 80%
1
 1  a11  1  0.8 * a12 a13 
 
LHX   1  0.8  * a21 1  1  0.8 * a22  1  0.8  * a23 
 
 a31 1  0.8  32
* a  33  
1  a
END

You might also like