Transcript BASE - World Bank

Choosing the Balance between
Human Development and
Infrastructure Spending in Malawi
Carolina Diaz-Bonilla
DECPG, World Bank
Malawi, June 20, 2007
Joint work with:
Hans Lofgren, Pavel Lukyantsau, and Antonio Nucifora
Introduction
• The Malawi Growth and Development Strategy (MGDS)
marks a policy shift towards economic growth and
infrastructure development, and away from investments
in human development (HD):
– Belief that resource allocations in previous strategies were
tilted towards general administration and social services at the
expense of infrastructure services;
– MDGS thus emphasizes the need to adjust these allocations.
• Simulation analysis will:
– Create a base scenario in line with the GoM’s macro program
– Explore the consequences of changing the balance of
expenditures between infrastructure and HD sectors.
Introduction (cont)
• Use MAMS: Maquette for MDG Simulations.
– Economywide simulation model to analyze development
strategies.
– For Malawi: not targeting MDGs, but monitor them.
• Rationale for economywide approach:
– Economic effects at different levels: macro, sectors, labor
market.
– Sector-by-sector approach (partial equilibrium) analysis is
not sufficient on its own
• Results provide an indication of the broad direction of
changes in the economy and the likely trade-offs.
Outline
•
•
•
•
Model Structure
Database
Simulations: Assumptions
Simulations: Results
Model Structure
• Standard dynamic recursive computable general
equilibrium (CGE) model AND
• Additional module that links specific MDG or
poverty-related interventions to poverty and
other MDG achievements.
• Requires relatively detailed treatment of
government activities to make the link possible:
– Malawi: 9 government activities.
Model Structure (cont)
• Some features of open-economy, dynamicrecursive CGE models:
– Optimizing producers and consumers.
– Supply-demand balance in factor and commodity
markets (with flexible prices clearing most markets)
– Expenditures = receipts for the three macro balances:
government, savings-investment, rest of world
– Imperfect transformation/substitutability in trade.
– Updating of factor and population stocks and TFP;
endogenous/exogenous mix.
Table: Model Disaggregation
Activities/Commodities (12)
Non-government (3)
Agriculture
Industry
Private Services
Government (9)
Education primary
Education secondary
Education tertiary
Health
Government agriculture
Irrigation
Water & sanitation
Public infrastructure
Other government
Factors (14)
Labor with less than completed secondary education
Labor with completed secondary education
Labor with completed tertiary education
Capital (10) - one stock for each model gov activity
Land
Institutions (3)
Household
Government
Rest of the World
Model Structure (cont)
• MAMS treatment of government spending:
– Government purchases public services, disaggregated
by function.
– Government services produced using labor,
intermediate inputs, and capital.
– Provision of education, health, and water-sanitation
services contribute directly to MDGs and influence
factor productivity.
– HD (education) influences size and composition of
labor force.
– Sources of government income: taxes, domestic
borrowing, foreign borrowing, and foreign grants
Model Structure (cont)
• Education:
– Disaggregated by cycle.
– Endogenous student behavior:
• Shares of relevant totals that enter first grade;
• In a grade: shares that pass, continue, repeat, or drop
out within or between cycles.
– Within each cycle and between cycles, student
behavior determined by logistic-CE structure (for
arguments, see Table)
– Enrollment in each cycle = old students that
continue/repeat + graduates from earlier cycle +
new entrants to school system.
Model Structure (cont)
• Labor (by level of education) defined as the
sum of:
– Remaining stocks from last year
– Graduates and dropouts who enter the labor force
– Net entrants from outside the school system
Table: Determinants of MDG achievements
Other Determinants
MDG
Per-capita
real service
delivery
Poverty Rate
Net primary
completion
Under-5
mortality rate
Maternal
mortality rate
Access to safe
water
Access to basic
sanitation
Per-capita
Public
household
infrastructure
consumption
Wage
Incentives
Other
MDGs
X
4
X
X
X
X
X
X
X
7a,7b
X
X
X
7a,7b
X
X
X
X
X
X
Database
• Social Accounting Matrix (SAM): fiscal year 2004
• Macro SAM:
– National Accounts data averaged over calendar years 2003 and
2004 (SIMA, World Bank)
– Balance of Payment information for last two quarters of 2003
and first two quarters of 2004 (IMF data), and
– Government Budget for fiscal year 2004.
• Micro SAM: created from macro SAM above and 1998
Malawi micro SAM from IFPRI
• Government sectors:
– Disaggregated using recurrent and development expenditure
information from several Malawi Ministries.
Figure. SAM Structure
Expenditures
Receipts
Activities
Market
sales
Domestic
Institutions
Home consumption
Intermediate
Commodities
inputs
Transactions
costs
Final
market
demands
Factors
Domestic
Institutions
Activities
Factors
Value
added
Taxes
Rest of
World
Totals
Commodities
Tariffs,
Taxes
Income,
Taxes
Rest of
World
Activity
income
Exports
Commodity
demand
Transfers
Factor
income
Transfers,
Transfers,
Taxes,
Savings
Savings
Commodity
supply
Institution
income
Foreign
exchange
outflow
Imports
Activity
spending
Totals
Factor
spending
Institution
spending
Foreign
exchange
inflow
Database (cont)
• Also need more detailed data related to different
MDGs in the labor market; ex:
–
–
–
–
Levels of service delivery to meet MDGs.
Stocks of students at different educational levels.
Stocks of labor by educational level.
Student behavioral patterns (ex: graduation rates)
• Elasticities in production, trade, consumption, and
in the different MDG functions.
GDP Growth in MAMS
• GDP growth is determined by:
– growth in factor employment
– growth in productivity or efficiency of factor
use (TFP)
GDP Growth in MAMS
• Factor employment
– Labor:
• Stock growth depends on functioning of education sector
• Un/underemployment responds to wage pressures.
– Capital:
• Government capital: grows in parallel with increases in
government services
• Private capital: investment ( stock growth) driven by
funding = [private savings] + [FDI] – [gov. borrowing]
– Land:
• Stock growth is exogenous.
GDP Growth in MAMS (cont.)
• TFP
– Endogenous part depends on economic
openness and growth in government
infrastructure stocks.
– Exogenous part captures what is not explained
in model (institutions, new technologies, ….)
Quality of Expenditures
• “Quality” is implicitly assumed constant: i.e.
efficiency in the use of expenditures does not
change
• The degree of efficiency is determined by the
assumptions in the model (elasticities)
• Simulations of the impact of improvements in
efficiency can be carried out by:
– Changing elasticities
– Decreasing the share of “Other Government Services”,
which is assumed as ‘non-productive’ expenditure in
MAMS
Simulations
• BASE: projection of economic program pursued by
current government, which is underpinned by the IMF
PRGF.
• Trade-off simulations:
– SIM-INFRA: reallocation of public expenditure toward
infrastructure, agricultural, and irrigation services (less HD)
– SIM-SOCIAL: reallocation of public expenditure toward
human development (education, health, and water-sanitation)
services (less infra).
• SIM-HISTORICAL
• Robustness simulation: Simulations under lower GDP
growth assumption.
BASE Simulation
• Based on the implementation of the economic
program pursued by the current government.
• Projects into the future the consequences of
continuing current policies and growth rates
GDP
Population
Growth Assumptions
6%
1.95% --> 2008
1.4-1.7% --> 2015
• Note: Assumption that adequate reforms adopted and Malawi is
not hit by exogenous shocks.
BASE Simulation (cont)
• All sources of financing for government
are exogenous and follow projections of
GoM economic program.
Growth Assumptions
Foreign Borrowing
3%
Foreign Grants
5%
Direct and indirect taxes
Domestic Borrowing
Fixed share of GDP
0%
Note:
Complete foreign debt relief occurs in 2007
BASE Simulation (cont)
• Endogenous government expenditure; clears
fiscal account.
• No changes in the current composition of
expenditures; shares across sectors are
maintained constant (as in 2003/04 fiscal year).
• Special functioning of education sector:
– government expenditure set to grow at a rate such that
spending per enrolled student (“educational quality”)
remains constant between 2004 and 2015.
BASE Simulation (cont)
• Government infrastructure capital stock (roads) has an effect on
the productivity of other sectors in the economy:
– Elasticity of TFP generated such that, ceteris paribus, the sum of the GDP
changes across all activities linked to the public infrastructure capital stock
per additional Kwacha spent on investments in this capital stock is equal
to 0.2 (an implicit rate of return on public capital)
• Agriculture activity also affected in all simulations by the
productivity in both the government agriculture and irrigation
sectors.
– Assumption linked to GoM strategy assumption that Malawi would
increase its efficiency to help with macroeconomic stability.
• Elasticity of factor productivity for labor with respect to health:
– Linked to under-5 and maternal mortality rates relative to base year.
BASE: Some Results
(annual growth rates)
2004
Household consumption
Government consumption (total)
Government investment
Private investment
Exports
Imports
Real exchange rate
- Kwacha per Foreign Curr. Unit
GDP at factor cost
Base
Values
Units
growth
178.7
30.8
21.0
6.3
49.9
98.2
100.0
bn Kwacha
4.9
5.9
5.1
8.2
7.6
5.1
0.02
167.3
bn Kwacha
bn Kwacha
bn Kwacha
bn Kwacha
bn Kwacha
indexed
to 100
bn Kwacha
6.0
BASE: Some Results
(annual growth rates)
2004
Base
Values
Units
growth
Wage of labor with less than secondary education**
Wage of labor with secondary education
Wage of labor with tertiary education
14584.8
69027.4
167075.8
Kwacha
Kwacha
Kwacha
MDG 1: headcount poverty rate
MDG 2: primary (1st cycle) net completion rate***
MDG 4: under-5 mortality rate (share of live births)
MDG 5: maternal mortality rate (share of live births)
MDG 7a: acess to safe drinking water
MDG 7b:acess to safe sanitation
52.4
8.0
133.0
960.0
66.1
63.7
%
%
per 1000
per 100000
%
%
4.0
6.3
7.6
2015 value
31.4
15.4
108.1
499.0
67.9
64.9
Trade-Off Simulations
• Public expenditures in infrastructure vs. HD sectors.
• SIM-INFRA simulation:
– Growth in spending on infrastructure sectors is exogenously
set higher by 1.5 times the growth rate in BASE.
– Implies that expenditure growth rates for the HD sectors are
endogenously scaled down to stay within fiscal space limits.
– Fiscal space defined by foreign inflows and domestic revenue
rules that are unchanged across the simulations.
• SIM-SOCIAL simulation:
– Growth in spending on infrastructure sectors is exogenously
set lower by 1.5 times the growth rate in BASE.
– Implies that expenditure growth rates for the HD sectors are
endogenously scaled up to stay within fiscal space limits.
• In both, public expenditure on the remaining sector (“other gov
services”) grows at the same rate as in BASE.
Infra vs HD: Some Results
(annual growth rates)
2004
Government consumption (total)
Total Education
Primary Education services
Health services
Irrigation services
Gov agriculture services
Water-sanitation services
Public infrastructure services
Other services
Government investment
Values
Units
30.8
7.5
5.1
3.6
0.2
1.3
0.2
2.2
15.9
21.0
bn K
bn K
bn K
bn K
bn K
bn K
bn K
bn K
bn K
bn K
SIMULATIONS
Base
Sim-Infra Sim-HD
Annual growth 2005-2015 (%)
5.9
8.6
2.2
4.8
4.8
4.8
4.8
4.8
4.8
5.1
5.8
7.2
0.7
3.3
9.0
9.0
3.3
9.0
4.8
5.6
5.8
9.5
3.0
5.6
-0.1
-0.1
5.6
-0.1
4.8
4.6
Infra vs HD: Some Results
(annual growth rates)
2004
Household consumption
Government consumption (total)
Government investment
Private investment
Exports
Imports
Real exchange rate
- Kwacha per Foreign Curr. Unit
GDP at factor cost
Values
Units
178.7
30.8
21.0
6.3
49.9
98.2
100.0
bn K
167.3
bn K
bn K
bn K
bn K
bn K
indexed
to 100
bn K
SIMULATIONS
Base Sim-Infra Sim-HD
Annual grw 2005-2015 (%)
4.9
5.9
5.1
8.2
7.6
5.1
0.0
5.2
5.8
5.6
8.4
8.3
5.5
-0.2
4.6
5.8
4.6
7.9
6.7
4.6
0.2
6.0
6.3
5.6
Infra vs HD: Some Results
(annual growth rates)
2004
Values
MDG 1: headcount poverty rate
MDG 2: primary (1st cycle) net completion rate***
MDG 4: under-5 mortality rate (shr of live births)
MDG 5: maternal mortality rate (shr of live births)
MDG 7a: acess to safe drinking water
MDG 7b:acess to safe sanitation
Units
Simulations
Base Sim-Infra Sim-HD
2015 Values
52.4
%
31.4
8.0
%
15.4
133.0 per 1000 108.1
960.0 per 100000 499.0
66.1
%
67.9
63.7
%
64.9
29.6
10.8
113.4
587.0
67.4
64.7
33.9
18.5
106.4
472.6
68.1
64.8
MDG 1: Headcount Poverty Rate
55
50
45
%
40
base-prgf
35
sim-infra
sim-social
30
25
2004 2005 2006 2007 2008 2009
2010 2011 2012 2013 2014 2015
MDG 2: Net Primary School
Completion Rate (%)
20
18
base-prgf
16
sim-infra
14
sim-social
12
10
8
6
4
2
0
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
Real GDP at Factor Cost
(2003/04 bn Kwacha)
350
base-prgf
sim-infra
300
sim-social
250
200
150
100
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
Trade-Off Simulations (cont)
• Focus on infrastructure results in an increase in
the growth rate of real GDP at factor cost from
6% to 6.3% per year.
• Faster growth rate, exports, investments and
(monetary) poverty reduction, but slower
progress in other human development indicators.
• Focus on social sectors leads to a slower
economic growth rate and slower poverty
reduction, but more rapid progress on human
development indicators.
• GDP decreases to 5.6% per year
SIM-HISTORICAL Simulation
• Analyzes the expected economywide outcomes of
continuing with pre-2004 trends and unchanged
policies.
• Results: scenario is unsustainable
– High amount of government expenditure (holding
constant the trends in net foreign borrowing, foreign
grants) results in high levels of domestic debt and
interest payments that are unsustainable.
– Household per capita consumption plummets, and
therefore poverty rises rapidly.
Robustness Simulation
Lower Growth
• Repeat BASE, SIM-INFRA, and SIM-SOCIAL
simulations under lower growth assumptions.
– BASE 4% rather than 6%
• Lower growth in all macro aggregates.
• Smaller budget for government => lower
government expenditure per sector.
• Improvements in poverty and all other HD
indices are more modest.
Summary
• Sound macroeconomic policies are critical for
both growth and human development indicators.
• Higher infrastructure spending leads to faster
GDP growth (and reductions in monetary
poverty), but at the expense of slower
improvements in HD indicators
• Higher HD sector spending leads to faster growth
in health, education, and other HD indicators, but
at the cost of a slower growth rate and poverty
reduction.
• Note: we assume that “quality” of expenditures
and investments does not change.
Thank You
All analysis has limitations. If we limited presentations to
analysis without limitations, then we would live in a world
without presentations.