Introduction to

IMPACT ANALYSIS USING

SUPPLY-USE TABLES

IFPRI On-Line
2018
 What have we learned so far?
I: Usefulness of IOT and SUT
• Why IOT & SUT based analysis may be important
• Concerned with econ-wide, consistency & structure

II: Composition of IOT and SUT

• IOT: Industry x Industry
• SUT: Add commodities

III: Applications with IOT

• Small IOT multiplier models
• National
• Regional
 Next: SUT Multiplier Model
I: Extension of IOT Multiplier Model
• Introducing commodities in addition to activities
• Extending the income and expenditure loop
II: Setting up the model
• What are the questions
• Linear algebra of basic model
III: Extending the basic model (beyond imports)
• Income
• Employment
• Energy
Remember the Simplified Supply
Use Table
1 2 3 5 6
Domestic
Activities Commodities Exports Total (X)
Final Demand
Domestic Total Supply
1 Activities
Supply by Activities
Intermediate Domestic Total Use of
2 Commodities Exports
Use Final Demand Commodities

3 Wages Wages Wages

4 GOS GOS GOS

NIT on NIT on
5 Net Indirect Taxes Indirect Taxes
Prodn Products
6 Imports Imports Imports

Total Use Total Supply

7 Total (X’) by of Demand Exports
Activities Commodities
Circular Flow in SU Table
 Assume that exports of agricultural commodities rise.
 What is the flow of impacts?
 Endogenous variables: Output of Activities and Supply of Commodities
 Exogenous variables: Domestic Final Demand and Export

Activities Commodities Demand

AAgric AManuf CAgric CManuf Dom Export Total
AAgric 279 279
AManuf 394 394
CAgric 84 55 79 71 289
CManuf 50 99 257 152 558
Pay to Fact 145 240 86 471
Other Pay 10 164 135 3 312
Total Cost 279 394 289 558 471 312
 SETTING UP THE SUT MODEL
1. Follow the same modelling approach as before
2. Define appropriate exogenous variables (f)
 Now: need to express in terms of commodities!!
 No exogenous demand for goods produced by activities
3. Define endogenous variables (x), but now a mix of:
 Output of activities
 Supply of commodities

4. Create an appropriate econ structure B, a mix of:

 Activities supplying (or making) commodities (fixed %)
 Activities using commodities as intermed inputs (fixed %)
 …where B replaces A of model introduced earlier
5. How?
CREATING THE COEFFICIENT
New: supply of
MATRIX B
Production input commodities
coefficients are
formedAgrinAct
the Mnf Act Agr Com Mnf Com HH Other Total
Agr Act usual way 279 279
Mnf Act 394 394
Agr Com
Mnf Com
84
50
55
99
 79
257
71
152
289
558
Pay to Fact
Other Pay
145 240
10 164 135
86
3
471
312
Total Cost 279 394 289 558 471 312
Agr Act Mnf Act Agr Com Mnf Com HH Other Total
Agr Act 0.97
Mnf Act 0.71 Assume ratio of
Agr Com 0.30 0.14 0.17domestic
0.23supply
Mnf Com 0.18 0.25 0.55to total0.49
supply is
Pay to Fact 0.52 0.61 fixed 0.28
Other Pay 0.03 0.29 0.29 0.01
Total Cost
 Express supply per unit (of
commodity output)
The Algebra
Express use per unit (of
x commodity output

activity output)
x activity
AAgric output CAgric
AManuf CManuf FD Total
AAgric c11C1 X1
AManuf c22C2 X2
CAgric a11X1 a12X2 F1 C1
CManuf a21X1 a22X2 F2 C2
Pay to Fact V1 V2 Yh
Other Pay V*1 V*2 F3 Yo
Total Cost X1 X2 C1 C2

Material Balances now for activities as well as commodities

 

B xfx
0X1 + 0X2 +c11C1+ 0C2 + 0 =X1
0X1 + 0X2 + 0C1+ c22C2 + 0 =X2
a11X1+a12X2+ 0C1+ 0C2 +F1 =C1 ● + =
a21X1+a22X2+ 0C1+ 0C2 +F2 =C2

 
 Ignore this part of the
Interpretation ofmultiplier
SUT Leontief
matrix inverse: multipliers

What is impact of a R1 increase in Exogenous demand for Manufacturing Commodity?

(I-B)-1

0.17   X 1  Activity Output multiplier:

0.88 X  0.17+0.88=1.05.
    2
0.17   C1 
   
1 .25 
  2 C
Commodity output multiplier = 1.42
Why > than the activity output multiplier?
Round by Round Decomposition of the SUT Leontief inverse multipliers

Output in Agr
Act now also
increases
Why is the
increase in After 12
Agr Com
output of Mnf rounds we get
Starting with required
Act less than as
the same
a 1 unit 1??? intermediate
results as in
increase in inputs in Rnd
the multiplier
final demand 2
matrix
for Mnf Com
Next: from output to income

 National Accounting Rules:

 Avoid double counting
 Distinguish between gross & net output
 Output multipliers: gross output concept
 Net output preferred: GDP
 Reward for factors of production / factor incomes
 Omit intermediate demands like in national accounts
 How to go from gross to net output multipliers?
 Basic principal: fixed ratio of factor costs to output
 How does this work?
Remember: CREATING THE COEFFICIENTS
Now take Factor
Production input
Cost as a are
coefficients
AgrinAct
proportion
formed theof Mnf Act Agr Com Mnf Com HH Other Total
Agr Act output
usual way 279 So, if output in Agr 279
Mnf Act
rises394
by 1 Rm (say 394
Agr Com 84 55 as a result of a79 71 289
Mnf Com 50 99 257
multiplier process), 152 558
Pay to Fact

145 240 factor income is up 86 471
Other Pay  10 by 0.52Rm
164 135
and 3 312
Total Cost 279 394 289 0.61Rm558 in Mnf471 312
Agr Act Mnf Act Agr Com Mnf Com HH Other Total
Agr Act 0.97
Mnf Act 0.71Defined here as
Agr Com 0.30 0.14 fincxr’:
0.17factor 0.23
Mnf Com 0.18 0.25 income / output0.49
0.55
Pay to Fact 0.52 0.61 ratios 0.28
Other Pay 0.03 0.29 0.29 0.01
Total Cost
 Remember the SUT Leontief inverse: now apply to
factor income to create income (GDP) multipliers
What is impact of a R1 increase inThus, following
Exogenous a Rm1
demand forincrease in
Manufacturing Commodity?
demand for Mnf Com, output in Agr
  Act goes up by 0.17Rm and VA by
0.09. Total VA is up by
Note: element by element multiplication
(I-B) -1 0.09+0.55=0.64 which constitutes the
GDP or (Fct) Income multiplier for
Mnf.

Activity Output multipliers:

0.17   X 1 
If output in Agr is up by 0.17
0.88 X  and in Mnf by 0.88 then by
 
=
    2 (element) multiplying with
0.17   C1  fixr’ we get the impact on
   
1 .25 
  2 C factor income as a proxy for
GDP
Employment & other multipliers

 Policy makers concern: employment

 What is the impact of an intervention on employment
 Is the impact more or less on low skilled employment?

 Suppose we know employment in activities

 Ideally, full time equivalents (FTEs)
 Perhaps even a disaggregation by skill (or otherwise)

 Employment multipliers
 Basic principal: assume fixed ratio of FTEs to output
 How does this work? Suppose employment in
Agr = 56
Mnf = 39
Employment ratios and multiplier preparations

 Lets
?
  calculate employment / output ratios
 So, if: employment in Act1: 56 & Act2: 39

 This means that

 for every R10 of output, Agr employs 2 FTEs, and
 for every R10 of output, Mnf employs 1 FTE
 Thus, if Agr X up by 10 units, employment up by 2
 Demand for labour to increase by 2 FTEs.
 But is that the end of the story?
 Remember the SUT Leontief inverse: now apply to
employment to create employment multipliers
What is the impact of a R1 increase in Exogenous demand for Manf Commodity?

 
Thus, following a R1for
If demand increase in
Mnf Comms increases by
demand for Mnf total
Similar
R100, Com,
set-upempl in Agr
can be
employment Act be for
constructed
could up
goes up by 0.034
energy
by units
12.2 or and in Mnf
emissions.
FTEs. by in
Explored
0.088 units. Total
detail inemployment
exercise sheet goes upmore
with
by 0.122. than one energy type
Activity Output multipliers:
0.17   X 1  If output in Agr is up by 0.17
0.88 X  and in Mnf by 0.88 then by
 
=
    2 multiplying with empxr we
0.17   C1 
    get the impact on
1 .25 
  2 C employment as a proxy for
demand for labour
Exercises
 Open file: SUT Basic + Extended Multiplier 6 Comm Exercise.xlsx
 SU Table highly aggregated
Additional
 5 Activities: ANAT / ALMF / AHMF / AUTIL / ASERV
commodity
 Producing
 6 commodities: CNAT / CLMF / CCHM / CHMF / CUTIL / CSERV
Exercises (cont 1)
 Additional accounts and external variables
 LAB = Labour
 CAP = Capital
 GOVT = Government
 ROW = Rest of the World

 emp = Employment (FTEs)

 eelc = Electricity Use (Tj)

 egas = Gas Use (Tj)
 eptr = Petroleum Use (Tj)
 ecol = Coal Use (Tj)
Exercises (cont 2)

 Recommended Range Names (already given!!)

 Z = Endogenous Variables
 x = Output / Supply
 B = Input / Supply Coefficients
 I = Indentity Matrix [can also use =munit(4) directly]
 f = Original final demand (as in the data)
 Df = Exogenous change in final demand
 finc = Factor income (value added as in data sum of factors)
 fincxr = Factor income / output ratio
 emp = Original employment (as in data)
 empxr = Employment / output ratio
 ENG = Energy use data (matrix)
 ENGXR = Energy use per unit of output (matrix)
Exercises (cont 3)
4 Sheets in file: SUT Basic + Extended Multiplier 6 Comm Exercise.xls :
1. Index
2. Basic Preparations
 Calculate Coefficients
 Calculate Leontief Inverse
 Perform Consistency Check
3. Basic Multipliers
 Simulation: 10% increase in CHMF (can be changed manually)
 Impact on output
 Impact on factor income (GDP @ factor costs)
 Multipliers
4. Extended multipliers
 Same simulation
 Impact on employment
 Impact on energy use
Exercises (cont 3)
Energy use in Tj:
1. 2005 data from www.dme.gov.za uprated to 2015
 Using constant price GDP growth rates
 Implies fixed energy ratios
2. 4 types of energy
 Electricity
 Gas
 Petroleum
 Coal
Exercises (cont 4)
Tasks
 Complete calculations in empty ranges
 Answer questions: discuss with group partner
 Group Discussion
More detailed exercises
Same format
4 Sheets in file: SUT Basic + Extended Multiplier Exercise 21 Comm.xls:
1. Index
2. Basic Preparations
3. Basic Multipliers
 Simulation: 10% increase in CIAS (can be changed manually)
4. Extended multipliers
End