Transcript Eval.Impact of Fiscal Multiplier on FS Economy _2013 PSEFx

Evaluating the Fiscal Multiplier Effects on the Free State
Economy: A Social Accounting Matrix Approach
by
O.S. Omoshoro-Jones, N.D. Mokalanyane & G.G. Mashibini.
Paper prepared for the 5th Annual Public Sector Economists’ Forum,
Cradle of Humankind Conference Centre, Maropeng, Magaliesburg, 27th – 29th Nov. 2013
© 2013. No part of this presentation should be distributed, reused or replicated in any form without a prior written consent of Free State
Economic Analysis Directorate.
1
Background
• Government spends money and levies taxes to finance its expenditure. Every
government must therefore regularly decide how much to spend, what to
spend it on and how to finance its expenditure.
• For effective and efficient fiscal policy, quantitative analysis should be done to
measure how much a direct effect is amplified or multiplied by indirect
linkage effects. The sum of all direct and indirect linkages associates with a
particular exogenous demand side shock measure the shock’s multiplier effect
• Although empirical studies on fiscal multipliers recently attracts lot of
attention globally, whilst none of existing empirical work have examine the
size of distributional effects on the provincial economy emanating from an
increase in government spending (a positive fiscal shock).
Research Objectives and Significance
The aim of this research is to:
• Examine the size of distributional effects on the provincial economy as a result of
government spending,
• Empirically estimate/review, the effect of discretionary fiscal policy on output and
household income with a specific focus on the Free State economy, and,
• Identify possible sector(s) and household groups with the highest/lowest gain from a
positive fiscal shock.
Why is this empirical research important?
•
Results form this study will provide empirical evidence to inform policy makers in FS on ‘how
to’ and ‘when to’ make decisions on provincial expenditure patterns in order to avoid wasteful
expenditure, implement fiscal prudence by effectively allocating funds given the availability of
funds .i.e. how much funds should be spent and/or transferred to low income households
relative to their high income counterparts to reduce poverty and income inequality?
• Empirical results will identify possible sectors (at microeconomic level) in the economy
can amplifier and maximize multipliers effects generated from a positive fiscal shock.
Review of Relevant Literature
• Extant literature investigating the impact of fiscal shock on the economy in SA remain scarce, whereas,
no empirical study specifically focusing on a provincial economy using SAM currently exist. To the best
of our knowledge, this study is the first pioneering paper to assess the influence of fiscal shock and the
corresponding multiplier effect on a provincial economy at macro and (micro) sectoral level.
• Theoretical and empirical literature suggests that fiscal multipliers differ widely across countries
• Jooste, Liu and Naraidoo (2012) shows that that multipliers effects depend an increases in
government expenditure. Authors finds that an increase in government expenditure have a positive
impact, albeit (at times) less than unity on the gross domestic product (GDP) in the short run but over
the long run, the impact of government expenditure on GDP is insignificant
• The empirical result of Spilimbergo et.al (2009) affirm that the size of the multiplier is larger if
“leakages” are few (that is, only a small part of the stimulus is saved or spent on imports), when the
monetary conditions are accommodating, where by interest rate does not increase as a consequence
of the fiscal expansion, and if the country’s fiscal position after the stimulus is sustainable.
• Given the empirical evidence in literature, an increase in government spending of 1 percent of GDP
ought to generate output and unemployment multipliers respectively of about 1.2 per cent (annually)
and 0.6 percentage points (at the peak).
• Monacelli, Perotti, and Trigari (2010) established that a percentage point increase in GDP will produces
an increase in employment of about 1.3 million jobs.
Data – set, treatment and sources
•
For the purpose of our analysis, we employ a highly disaggregated 2005 SAM
developed for Free State, which consists of:
• 34 commodities and 34 activities,
• 4 factors of production disaggregated as (labour, capital and enterprise),
• 48 households disaggregated as (whites, indians and blacks) and subdivided
into 12 percentiles each,
• 6 Government expenditures disaggregated as (national; provincial health,
provincial education, provincial social and provincial others; as well as local),
• 7 Government taxes disaggregated as (property income, transfers, direct tax,
indirect tax, subsidies, provincial government and local government),
• 2 Capital Accounts disaggregated as (government, corporate sector and
households), and
• 4 rest of South Africa disaggregated as (factor payments and transfers; goods
and services; transfers and balance on current account) and
• 4 rest of the world disaggregated as (factor payments and transfers; goods and
services; transfers and balance on current account).
Partial Equilibrium Framework …(1)
• Application of FS SAM to investigate the economic-wide impact of fiscal shocks
Computation Approach:
1. Values of Endogenous variables are determined by the economic model;
2. Values of Exogenous variables are determined outside the model;
3. The coefficients matrix of endogenous accounts will be estimated;
4. Links between the different variables in an economic system are captured
as systems of equations;
5. Estimated model is solved by determining the changes in the value of
endogenous variables (policy objectives) corresponding to changes in the
values of exogenous variables (policy instruments); and
6. Identify the distribution process of multiplier effects across sectors and
households as well as identify underlying structural bias using Relative
Distributive Measure (RDM) from these output and income multipliers.
Equation Notation in the Model…(1)
where;
A = Gross output of each commodity  i.e., A1 and A2 
Q = Gross demand for each commodity  i.e., Q1 and Q2 
V = Total factor income (denotes household income)
B = Total household income (represents total factor income)
E = Exogenous components of demand (government, investment, &
exports)
where ;
a = technical coefficients (i.e., input or intermediate shares in production)
b = share of domestic output in total demand
k = share of value-added or factor income in gross output
Also, Z = Aggregate demand in each sector { Sum of Dd for intermediate input;
household consumption Dd & other exogenous sources of demand E, e.g.
public consumption and investment}
Mathematical derivation of Multiplier Eqs. in the Model …(2)
Q  a A a A c BE
2
1
1
Agg, Dd for each commodity: Q1  a11 A1  a12 A

c
B

E
2
21 1
22
2
2
2
Gross output forms part of Total Dd:
A1  b1Q1 and A2  b2Q2
Eq. 3.1
Eq.3.2.
Total Dd = intermediate Dd + exogenous Dd
but...
Aggregate HH income depend on factor earnings per sector:
B  k1 A1  k2 B2
Substitute Eq. 3.2 into 3.3 to obtain total income (I matrix) B  k1b1Q1  k2b2Q2
Q1  a11b1Q1  a12b2Q2  c1 (k1b1Q1  k2b2Q2 )  E1
Replace A & B in Eq. 3.2.&3.4 to get: Q2  a21b1Q1  a22b2Q2  c1 (k1b1Q1  k2b2Q2 )  E2
If we isolate exogenous Dd, Identity, I
As such, Identity matrix, I:
1  a11b1  c1k1b1 a12b2  c2 k2b2  Q1  E1 
 a b  c k b 1  a b  c k b  Q   E 
 21 1 2 2 1
22 2
2 2 2
2
 2
1  a11b1  c1k1b1 a12b2  c2 k2b2 
 a b  c k b 1  a b  c k b   I  M
 21 1 2 2 1
22 2
2 2 2
Eq.3.3
Eq.3.4
Eq.3.5
Eq.3.8
Eq. 3.9
Mathematical derivation of Multiplier Eqs. in the Model …(2)
Subst. vectors Q and E, Eq.3.9 is rewritten as
 I  M Q  E
Eq.3.10
By taking the inverse of Eq. 3.10, we obtain the Multiplier’s formula
E  I  M  E
1
Eq.3.11
Total Dd = Multiplier Matrix X Exogenous Dd
To compute Unconstrained Multipliers (if production & consumption of HH is unchanged)
If fixed sectors, Z 2 import substitutes from domestic supply…eliminating sectoral growth linkages
Eq 3.5 becomes:
1  a11b1  c1k1b1  Q1   a12b2  c1k2b2  Q2  E1
 a21b1  c2 k1b1  Q1  1  a22b2  c2 k2b2  Q2  E2
Matrix format (isolating exogenous Dd)
a12b2  c1k2b2  E1 
1  a11b1  c1k1b1 0  Q1   1

   
 
 a21b1  c2 k2b1 1 E2   0 1  a22b2  c2 k2b2  E2 
1  a11b1  c1k1b1 0 

 I M *
 a21b1  c2 k2b1 1
a12b2  c1k2b2 
1
Rearranging Eq.3.15 to obtain, B

B
 0 1  a22b2  c2 k2b2 
Q
E
In a reduced from, Eq. 3.16 becomes:  I  M *  1    2 
 E2   Z 2 
 Q1 
 E1 
1
Therefore, the unconstrained Multipliers are modelled as:
    I  M * B  
 E2 
 Q2 
We have Identity matrix* denoted as:
Eq.3.11
Eq.3.12
Eq.3.15
Eq. 3.16
Eq. 3.17
Eq.3.18
Simulated Scenarios and Interpretation of Empirical Results
I.
Increase in government expenditure via an injection of R1 million into FS
economy (+ve fiscal shock)
Counterfactual experiment focused on the impact of a positive fiscal shock on:
• Commodities across the disaggregated 10 sectors ( i.e. agriculture, mining, utility,
construction, trade, transport, finance and services)
• 4 households categorised according to race ( i.e.White,African black, Indian/Asian)
Simulated Scenario
1. Increase in government expenditure via an injection of R1 million
into FS economy (+ve fiscal shock)
Counterfactual experiment focused on the impact of a positive fiscal
shock on:
• Commodities across the disaggregated 10 sectors ( i.e. agriculture,
mining, utility, construction, trade, transport, finance and services)
• 4 households categorised according to race ( i.e. White, African
black, Indian/Asian)
Analysis of Empirical Results…(i)
a. Government Expenditure Multiplier: Effect of R 1 million
Injections (+ve fiscal shock) on Commodities
Table 1: FS Govt. Expenditure Multiplier: Effect of Fiscal Shock on Sectoral Commodities
Commodities
Size of
multipliers
Commodities
AGR-C
MIN-C
MAN-C
UTI-C
CON-C
TRD-C
AGR-C
1.44
0.49
8.83
0.48
0.29
0.77
MIN-C
0.19
2.32
5.51
0.7
0.27
MAN-C
UTI-C
1.81
0.08
2.78
0.21
63.51
1.85
2.98
2.74
CNS-C
0.05
0.12
0.96
TRD-C
TRANC
FIN-C
0.1
0.18
0.41
0.49
SERV-C
Sum output
multiplier
BLKHouseholds
HH
CLDHH
ASNHH
WHTHH
Sum income multiplier
Income/output
multiplier
TRAN-C
FIN-C
SER-C
AVR
0.54
0.87
0.28
1.56
0.29
0.31
0.33
0.14
1.12
1.88
0.08
3.43
0.22
3.44
0.17
3.77
0.21
1.67
0.08
9.48
0.63
0.17
1.29
0.12
0.11
0.23
0.05
0.35
1.98
0.18
0.02
2.24
0.28
0.29
0.12
0.6
0.59
0.63
7
10.6
0.61
0.94
0.34
0.58
0.76
1.28
3.01
1.08
0.84
4.7
0.31
0.5
1.54
2.31
0.35
0.91
8.23
0.64
0.57
0.71
0.68
0.97
2.31
1.71
4.93
8.24
108.48
9.44
5.31
9.81
9.62
12.22
5.45
19.28
0.59
1.25
12.98
1.22
0.66
1.27
1.22
1.61
0.63
2.38
0.03
0.05
0.64
0.05
0.03
0.06
0.06
0.08
0.03
0.11
0.01
0.01
0.18
0.01
0.01
0.02
0.02
0.02
0.01
0.03
0.29
0.92
0.54
1.86
6.64
20.44
0.56
1.85
0.32
1.02
0.61
1.95
0.58
1.87
0.97
2.68
0.34
1.01
1.21
3.73
0.19
0.23
0.19
0.2
0.19
0.2
0.19
0.22
0.19
0.2
Table 2: Proportional Distribution of expenditure multipliers among commodities &
households
Commodities
Proportional distribution
Commodities
AGR-C
MIN-C
MAN-C
UTI-C
CON-C
TRD-C
TRA-C
FIN-C
SER-C
AGR-C
0.29
0.06
0.08
0.05
0.05
0.08
0.06
0.07
0.05
MIN-C
0.04
0.28
0.05
0.07
0.05
0.03
0.03
0.03
0.03
MAN-C
UTI-C
CONSC
TRADE
-C
TRANC
FIN-C
0.37
0.02
0.34
0.03
0.59
0.02
0.32
0.29
0.35
0.02
0.35
0.02
0.36
0.02
0.31
0.02
0.31
0.01
0.01
0.01
0.01
0.02
0.24
0.01
0.01
0.02
0.01
0.02
0.02
0.02
0.02
0
0.23
0.03
0.02
0.02
0.08
0.1
0.07
0.08
0.06
0.1
0.06
0.1
0.06
0.11
0.08
0.13
0.31
0.11
0.07
0.38
0.06
0.09
SERV-C
0.07
0.11
0.08
0.07
0.11
0.07
0.07
0.08
0.42
1
1
1
1
1
1
1
1
1
0.64
0.67
0.64
0.66
0.65
0.65
0.65
0.6
0.63
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.32
0.29
0.32
0.3
0.31
0.31
0.31
0.36
0.34
1
1
1
1
1
1
1
1
1
Sum output
multiplier
Households
BLKHH
CLDHH
ASIANHH
WHTHH
Sum income multiplier
Table 3: Total Expenditure Multiplier Effects by the type of Multiplier
1
Agric
2
Mining
3
4
Manufacturing
Utility
5
6
Construction
Trade
7
Transport
8
9
Finance
Community
Services
Output
5.26
8.43
108.48
9.44
5.4
9.81
9.62
12.22
5.45
GDP
2.28
4.52
48.14
4.52
2.31
4.57
4.52
6.57
2.28
Income
0.92
1.86
20.44
1.85
1.02
1.95
1.87
2.68
1.01
TOTAL
8.47
14.8
177.06
15.82
8.73
16.34
16.01
21.47
8.74
Analysis of Empirical Results…(ii)
b. Government Expenditure Multiplier: Effect of Fiscal Shock on
disaggregated Households & race group.
Table 4: Multipliers Effects of Transfers to Households: R 1 million transfer to HH
Households
Size of multipliers
Commodities
AGR-C
MIN-C
MAN-C
UTI-C
CONS-C
TRADE-C
TRAN-C
FIN-C
SERV-C
Sum output
multiplier
Households
BLK-HH
CLD-HH
ASIAN-HH
WHT-HH
Sum income multiplier
Income/output multiplier
Blacks
5.94
2.02
Coloured
6.58
2.05
Asia/Indian
7.45
2.15
Whites
7.01
2.13
Average
6.75
2.09
25.48
1.36
0.78
1.71
4.75
8.60
1.68
26.87
1.32
0.62
1.83
4.63
7.00
5.25
28.55
1.47
0.68
1.86
5.44
7.33
5.63
27.87
1.47
0.64
2.08
5.05
7.05
5.32
27.19
1.41
0.68
1.87
4.97
7.5
4.47
52.32
16.24
0.33
0.09
3.53
20.19
0.39
56.15
6.62
12.41
0.09
3.4
22.52
0.40
60.56
7.07
0.34
12.11
3.61
23.13
0.38
58.62
6.87
0.33
0.09
15.55
22.85
0.39
56.91
9.2
3.35
3.10
6.53
22.17
0.39
Table 5: Proportional Distribution of the Expenditure Multiplier across Commodities
and Households – transfers of R 1million to Households
Households
Size of multipliers
Commodities
Blacks
Coloured
Asia/Indian
Whites
AGR-C
0.11
0.12
0.12
0.12
MIN-C
0.04
0.04
0.04
0.04
MAN-C
0.49
0.48
0.47
0.48
UTI-C
0.03
0.02
0.02
0.03
CONS-C
0.01
0.01
0.01
0.01
TRADE-C
0.03
0.03
0.03
0.04
TRAN-C
0.09
0.08
0.09
0.09
FIN-C
0.16
0.12
0.12
0.12
SERV-C
0.03
0.09
0.09
0.09
1
1
1
1
Blacks
0.8
0.29
0.31
0.30
Coloured
0.02
0.55
0.01
0.01
Asia/Indian
Whites
0
0.18
0
0.15
0.52
0.16
0
0.68
1
1
1
1
Sum output multiplier
Households
Sum income multiplier
Analysis of Empirical Results…(ii)
c. The Government Expenditure Multiplier: Effect of Fiscal
Shock on Households Income Distribution
Table 6: Comparison of Average Annual Household Consumption Income By Population Group of
Household Head
(Source: Stats SA, IES: 2011, own calculation)
IES 2010/11
(R)
South Africa
IES 2005/06
(R)
119542
Real Growth
102401
Increase In
Rand Terms
16.7%
17141
Population Group of Household Head
Blacks
69632
51773
34.5%
17859
Coloured
139190
109038
27.7%
30152
Indian/Asian
White
252724
387011
184711
385599
36.8%
0.40%
68013
1412
Table 7: Total Multiplier Effects by the Type of Multiplier: R 1 million injection
Blacks
HOUSEHOLDS
Coloured
Asian/Indian
Whites
Output
56.77
56.15
60.56
58.62
GDP
25.46
24.83
26.49
25.74
Income
22.79
22.52
23.13
22.85
105.02
103.51
110.18
107.21
Total
The Government Expenditure Multiplier: Effect of Fiscal Shock on Employment
Source: QLFS (2013), StatsSA
•
An increased in unemployment rate across all race group between Q1:2013 and Q2:2013
•
Highest unemployment rate is observed in Black household followed by Coloured households.
•
Apparent unemployment rate among the black and coloured household can be explained by level of
education and skill orientation – low education level / non-skilled
•
Notably, White and Indian/Asian households are considered as highly skilled and semi-skilled respectively –
suitable for lucrative employment.
Table 8a: Output Multiplier of a R1Million Investment
SECTOR
Agriculture
Wholesale and Retail
MULTIPLIER
(Numbers of Employment Created)
10.5
3.3
Manufacturing
3
Construction
2.5
Finance
1
Mining
0.5
Transport & Communication
0.1
Electricity
0.1
Table 8b: Output Multiplier of a R1Million Investment (used PAIRS BAU)
Growth Rate over 10 years
Output
Employment
Investment
Household consumption
Real wages
Exports
Imports
Fiscal revenue
Simulated as growth path of
3.4%
R184 bn
158,000
R116 bn
R136 bn
R61 bn
R52 bn
R102 bn
R62 bn
Simulated as growth path of
10%
R537 bn
454,000
R339 bn
R398 bn
R177 bn
R151 bn
R297 bn
R182 bn
Interpreting the results in Table 8b:
• A direct boost in the manufacturing output in SA will have a direct impact on the macro variables in the
system. i.e. faster adjustment process exists as the shock affects output directly. Given a 3.4% growth per
annum, overall economic output of the manufacturing sector will increase by a cumulative R184 billion, and
by R537 billion given a sustainable 10% growth per annum.
• Theoretically, modelling the baseline scenario growth path (3.4%), the cumulative additional jobs that will be
generated will come to an estimated 158, 000 leading to a real wage increase of about R61 billion over the
same period. Investment and household consumption spending will receive an estimated R116 billion and
R136 billion boost, respectively. An estimated R52 billion additional exports will be generated under the
assumption that global demand will pick up moderately in favour of domestic manufactured goods.
Summarized Empirical Findings (1)
Our results suggests that…
• The need for FS government to focus its strategic policy and priority on revitalising the
fading mining and the agriculture sectors to reduce the existing high unemployment rate.
• An increase in provincial government spending would only improve general welfare across
households, specifically the previously disadvantage households (the African Blacks), IF FS
government implement a subsidized rate scheme targeted at manufacturing sectors to
stimulate output, sectoral performance and provincial contribution to the national GDP.
• Only a vibrant manufacturing sector in the FS would have a significant direct impact
on macro-economic variables such as household consumption, GDP, income and investment.
• At the sectoral level, manufacturing, wholesale & retail, trade, finance, construction and
mining sectors are capable of creating massive jobs and improve the Provincial growth over a
forecasted ten (10) year period. However, manufacturing sector has the strongest
multiplier effect that filters into other sectors.
•
The manufacturing sector remains the most important sector in FS capable of producing
more income, boost GDP growth and raise aggregate output level. For SA, cf. Ngadu et.al
(2010)
Summarized Empirical Findings (2)
White and Asian/Indian Household groups benefited more than the African Black households. This
findings aligns with results of Thurlow (2006), Thurlow and van Seventer (2008), Hérault et.al
(2005), Hérault et.al (2009), Mabugu (2007).
…This finding is plausible because the manufacturing sector: (i) require mostly the highly skilled and semiskilled workers to operates its machineries and few unskilled workers (larger made up of the previously
disadvantage racial groups), (ii) is more capital-intensive and less labor intensive.
•
•
The previously disadvantage (African Blacks) will only benefit from an increase in
government expenditure only if the Free State government initiate the revitalization of
labour intensive sectors such as agriculture and mining. Policy designs targeted towards
beneficiation in the mining sectors, agro-processing and construction of infrastructures will assist the
Provincial government to tackle the existing high rate of unemployment, poverty incidence and
inequality while improving the Provincial economic performance and contribution to the national
gross GDP.
•
Overall, empirical evidence shows that the largest multiplier lies in the manufacturing
sector, hence, South Africa and Free State government should align its investments into strategic
sectors as agricultural and mining value chains, manufacturing and services advocated in the New
Growth Path frame work (2011:24)
POLICY RECOMMENDATIONS (i)
• Four policy related recommendations are suggested to policy
makers in Free State based on our empirical findings.
1.
To enhance output productivity and reduce the burden of high production
cost of energy intensive industries (e.g. manufacturing), FS government should
standardise electricity price at a subsidized rate. In effect, this would allow
competitiveness across the sectors/ industries.
Long run effect : Lower production cost
High productivity
Increase in labour
demanded by energy intensive industries such as manufacturing sectors
Lower
Provincial unemployment rate & raise labour absorption capacity of energy
intensive industries due to rise in aggregate demand emanating from higher
income received by households for supplying labour.
POLICY RECOMMENDATIONS (ii)
2.
Urgent need for Provincial government develop more technical trainings
centres to encourage skills development of workers focusing on meeting the
pre-requisite skill demand in the capital intensive sectors. Specifically, FS
government should spearhead new skill development programmes and
intensify skills upgrade via its EPWP & other Public Works programmes.
The objective of a typical skill development programmes must
focus on skill upgrades for artisans, technicians, engineers etc.
3.
FS government should implement strategic policy via Department of
Economic, Tourism and Environmental Affairs (DETEA) to accelerate the
beneficiation in the mining sector to spur both down-and upstreammanufacturing industries.
POLICY RECOMMENDATIONS (iii)
4.
FS government should revitalize the fading agriculture sector to counteract
the current high unemployment rate in the Province.
•
An effective policy strategy focusing on the Free State agriculture sector is important
not only to absorb the large chunk of obsolete skills that made up of the high
unemployment pool, but a vibrant agricultural sector would provide tangible and
value-adding benefits for the creation of other important sectors such as agroprocessing, commercial and industrialised farming and animal-husbandry.
•
Ultimately, these sectors will generate large pool of employment given their highly
labour intensive nature and absorbs substantial number of workers with little/no
technical or formal skill.
THANK YOU
for being a delightful Audience!
QUESTIONS & DISCUSSIONS
O.S. Omoshoro-Jones
Snr. Econometrician (Manager: Modeling & Forecasting sub-directorate)
Free State Provincial Treasury
Tel: +27 51 405 4065
Fax:+27 405 4999
Email: [email protected]
30
© 2013. No part of this presentation should be distributed, reused or replicated in any form without a prior written consent of FS EAD.