Transcript Lecture 3: Vulnerabilities Related to the Scaling Up of Aid and Other

Thorvaldur Gylfason
 Aid and other capital flows
 History, theory, evidence
 Foreign aid and economic growth
 Effectiveness: Does aid work?
 Macroeconomic challenges
Dutch disease
 Aid volatility

 Policy
options in managing aid flows and
lessons from recent experience
Preparing for scaling up aid
 Vulnerabilities
 Monetary and fiscal policy options
 Debt sustainability
 Governance issues

 Conclusions
and guidelines
Unrequited
transfers from donor
to country designed to promote
the economic and social
development of the recipient
 Excluding
commercial deals and
military aid
Concessional
loans and grants
included, by tradition
 Grant
element ≥ 25%
Development
aid can be
 Public
(ODA) or private
 Bilateral (from one country to another)
or multilateral (from international
organizations)
 Program, project, technical assistance
 Linked to purchase of goods and
services from donor country, or in kind
 Conditional in nature

IMF conditionality, good governance
 Moral duty
 Neocolonialism
 Humanitarian intervention
 Public good
 National (e.g., education and health care)
 International
Social justice to promote world unity
 UN aid commitment of 0.7% of GDP

 World-wide redistribution
 Increased inequality word-wide
 Marshall Plan after World War II
1.5% of US GDP for four years vs. 0.2% today
 But Think tank in Nairobi disagrees, see
www.irenkenya.com

Objectives
 Individuals
in donor countries vs.
governments in recipient countries
Who
should receive the aid?
 Today’s
Aid
poor vs. tomorrow’s poor
for consumption vs. investment
Conflicts
 Beneficiaries’ needs
 Donors’ interests
Aid
is a recent phenomenon
Four major periods since 1950
 1950s:
Fast growth (US, France, UK)
 1960s: Stabilization and new donors

Japan, Germany, Canada, Australia
 1970s:
Rapid growth in aid again due
to oil shocks, recession, cold war
 1980s: Stagnation, aid fatigue, new
methods, new thinking
Rapid
growth of development aid
US provided 50% of total ODA

To countries ranging from Greece to South
Korea along the frontier of the “SinoSoviet bloc”
France

To former colonies, mainly in West Africa
UK

provided 30%
provided 10%
To Commonwealth countries
Stabilization
of aid from traditional
donors and emergence of new donors

US contribution decreased considerably
after the Kennedy presidency (1961-63)
The
French contribution decreased
starting from the early 1960s
New
donors included Japan,
Germany, Canada, and Australia
Rapid
growth in aid from industrial
countries in response to the needs of
developing countries due to
 Oil
shocks
 Severe drought in the Sahel
The
donor governments promised to
deliver 0.7% of GNI in ODA at the UN
General Assembly in 1970
 The
deadline for reaching that target was
the mid-1970s
Stagnation
of development
assistance
 Donor
fatigue?
 Private investor fatigue?
12
United
States: largest donor in
volume, but low in relation to GDP
 US
aid amounts to 0.2% of GDP
Japan:
second-largest donor in volume
Nordic countries, Netherlands
 Major
donors to multilateral programs
 Sole countries whose assistance accounts
for 0.7% of GDP
EU:
leading multilateral donor
Even
though targets and agendas
have been set, year after year,
almost all rich nations have
constantly failed to reach their
agreed obligations of the 0.7% target
Instead of 0.7% of GNI, the amount
of aid has been around 0.4% (on
average), some $100 billion short
40
35
1985
1990
2000
30
25
20
15
10
5
0
sub-Saharan
Africa
Asia
Oceania
MEDA
Latin America
Europe
Sub-Saharan
Africa and Asia have
received the most aid, the former a
rising amount over time
Aid to Sub-Saharan Africa is high in
relation to GDP
 For
the 44 countries in the IMF’s Africa
Department, net official transfers are as
follows:
< 5% of GDP:
14 countries
6%-16% of GDP: 24 countries
> 20% of GDP:
6 countries
 The
Blair Report and the Sachs Report
called on world community to increase
development aid (particularly for
Africa) to enable developing countries
to attain the MDGs by 2015
 2005
G-8 Gleneagles communiqué called
for raising annual aid flows to Africa by
$25 billion per year by 2010
 2005 UN Millennium Project called for $33
billion per year in additional resources

For comparison, US gave $20 billion in 2004,
not $70 billion as suggested by UN goal
The
recent increase in aid flows
toward developing countries
(particularly Africa) poses crucial
questions for both recipient
countries and donors
 What
is the role of aid?
 What is the macroeconomic impact of aid?
 Is the impact of aid necessarily positive,
or could aid have adverse consequences?
Aid
fills gap between investment
needs and saving and, if well
managed, can increase growth
 Poor
countries often have low savings and
low export receipts and limited
investment capacity and slow growth
Aid
is intended to free developing
nations from poverty traps
 E.g.,
capital stock declines if saving
does not keep up with depreciation
To understand the link between aid and
investment, consider Resource
Constraint Identity by rearranging the
National Income Identity:
Y=C+I+G+X–Z
I = (Y – T – C) + (T – G) + (Z – X)
Sp
Sg
Sf
In words, investment is financed by the
sum of private saving, public saving,
and foreign saving
This
is where aid enters the picture
Rearrange again:
Y+Z=E+X
where E is expenditure
E=C+I+G
Total supply from domestic and foreign
sources Y + Z equals total demand E + X
Aid increases recipient’s ability to
import: Z rises with increased X, incl. TR
Poor
countries are trapped by poverty
 Driving
forces of growth (saving,
technological innovation, accumulation of
human capital) are weakened by poverty
 Countries become stuck in poverty traps
Aid
enables poor countries to free
themselves of poverty by enabling
them to cross the necessary
thresholds to launch growth
 Saving
 Technology
 Human capital

Is it feasible to lift all above a dollar a day?
 How much would it cost to eradicate extreme
poverty? Let’s do the arithmetic (Sachs)
Number of people with less than a dollar a
day is 1.1 billion
 Their average income is 77 cents a day, they
need 1.08 dollars

 Difference amounts to 31 cents a day, or 113
dollars per year

Total cost is 124 billion dollars per year, or
0.6% of GNP in industrial countries
 Less than they promised! – and didn’t deliver
Several
empirical studies have
assessed the impact of aid on growth,
saving, and investment
The results are somewhat inconclusive
 Most
studies have shown that aid has no
significant statistical impact on growth,
saving, or investment
However,
aid has positive impact on
growth when countries pursue “sound
policies”
 Burnside
and Dollar (2000)
Regression
analysis to measure
the impact of aid on
 Saving
 Investment
 Public
finance
 Economic growth
 Saving

Negative effect on saving
Substitution effect? I.e., crowding out?
 Boone 1996; Reiche 1995


Positive effect for good performers

E.g., South-East Asia, Botswana
 Investment


No impact on private investment
Positive impact for good performers
 Public


finance
Uncertain effect on public investment
Positive effect on public consumption
Growth:
Mixed results
 Most
early studies showed no
statistically significant impact
 Some more recent studies show
negative impact

Selection bias and endogeneity issues
Need
to distinguish between
different types of aid
 Leakages,
cash vs. aid in kind
aid has
sometimes been
compared to natural
resource discoveries
 Aid and growth are
inversely related
across countries
 Cause and effect
 156 countries,
1960-2000
Per capita growth adjusted for initial income (%)
 Foreign
r = rank correlation
r = -0.36
6
4
2
0
-2
-4
-6
-8
-20
0
20
40
60
Foreign aid (% of GDP)
80
 No
robust relationship between aid and
growth
 Aid works in “countries with good
policies”
 Aid works if measured correctly
 Distinction between fast impact aid
(infrastructure projects) and slow
impact aid (education)
 Infrastructure:
High financial returns
 Education and health: High social returns
So,
empirical evidence is mixed
Need to distinguish between
different types of aid
Need to acknowledge diminishing
returns to aid as well as limits to
domestic absorptive capacity
Need to clarify interaction with
governance and good policies
Special case: Post-conflict situations
Aid
may lead to corruption
Aid may be misused, by donors as well
as recipients
 Donors:
Excessive administrative costs
 Recipients: Mismanagement, expropriation
Aid
may be badly distributed,
sometimes for strategic reasons
 Supporting
opposition
government against political
Aid
increases public consumption,
not public investment
Aid is procyclical
 When
Aid
it rains, it pours
leads to “Dutch disease”
 Labor-intensive
and export industries
contract relative to other industries in
countries receiving high aid inflows
 Dutch disease may undermine external
sustainability
 Aid
volatility and unpredictability
may undermine economic stability in
recipient countries
 Economic
vs. social impact
 Growth
is perhaps not the best
yardstick for the usefulness of aid
 Long

run vs. short run
E.g., increased saving reduces level of
GDP in short run, but increases growth of
GDP in long run
See my “Dutch Disease” in the New Palgrave Dictionary of Economics Online
 Appreciation
of currency in real terms,
either through inflation or nominal
appreciation, leads to a loss of export
competitiveness
 In 1960s, Netherlands discovered natural
resources (gas deposits)


Currency appreciated
Exports of manufactures and services suffered,
but not for long
 Not
unlike natural resource discoveries, aid
inflows could trigger the Dutch Disease in
receiving countries
Foreign
exchange is converted into
local currency and used to buy
domestic goods
Fixed exchange rate regime
 Expansion
of money supply leads to
inflation and an appreciation of real
exchange rate
Flexible
 Increase
exchange rate regime
in the supply of foreign exchange
leads to an appreciation of the nominal
exchange rate, so the real exchange rate
also appreciates
Review
theory of Dutch disease
in two rounds
 Demand
and supply model
 Two-sector model
Demand effects
 Supply effects
 Exchange rate volatility

Real exchange rate
Payments for imports
of goods, services, and
capital
Imports
Earnings from exports
of goods, services, and
capital
Exports
Foreign exchange
eP
Q
P*
Devaluation or
depreciation of e
makes Q also
depreciate unless P
rises so as to leave Q
unchanged
Q = real exchange rate
e = nominal exchange rate
P = price level at home
P* = price level abroad
1. Suppose e falls
Then more dinars per dollar,
so X rises, Z falls
2. Suppose P falls
Then X rises, Z falls
3. Suppose P* rises
Then X rises, Z falls
Summarize all three by supposing
that Q falls
Then X rises, Z falls
eP
Q
P*
Real exchange rate
Aid leads to appreciation,
and thus reduces exports
C
B
A
Imports
Exports
plus aid
Exports
Foreign exchange
Real exchange rate
Oil discovery leads to appreciation,
and reduces nonoil exports
C
B
A
Imports
Exports
plus oil
Exports
Foreign exchange
Real exchange rate
Composition of exports
matters
C
B
A
Imports
Exports
plus oil
Exports
Foreign exchange
 Dutch
disease is a real phenomenon, not
monetary
 Real exchange rate always floats
 Recall: Q = eP/P*
 Flexible exchange rate regime

Nominal appreciation
 Fixed

exchange rate regime
Inflation
 Look
at this more closely in two-sector
model of traded vs. nontraded goods: Skip
A
large inflow of foreign aid -- like a
natural resource discovery -- can
trigger a bout of Dutch disease in
countries receiving aid
A real appreciation reduces the
competitiveness of exports and might
thus undermine economic growth
 Exports
have played a pivotal role in the
economic development of many countries
 An accumulation of “know-how” often
takes place in the export sector, which may
confer positive externalities on the rest of
the economy
 Aid
 It
is likely to lead to Dutch disease if
leads to high demand for nontradables
Trade restrictions may produce this outcome
 Recipient country uses aid to buy nontradables
(including social services) rather than imports

 Production

is at full capacity
Production of nontradables cannot be increased
without raising wages in that sector
 Aid
is not used to build up infrastructure
and relax supply constraints
 Price and wage increases in nontradables
sector lead to strong wage pressure in
tradables sector
The
risk that aid flows might have an
adverse impact on the economy as a
result of aid-induced Dutch Disease
crucially depends on how aid is used
in the recipient countries
We can identify four different cases
on the basis of how the aid is spent,
and in which the macroeconomic
implications of aid flows are different
 Aid
spending can take several forms,
with different macroeconomic
implications:
 Case
1: Aid received is saved by recipient
country government
 Case 2: Aid is used to purchase imported
goods that would not have been
purchased otherwise (grants in kind)
 Case 3: Aid is used to buy nontradables
with infinitely elastic supply
 Case 4: Aid is used to buy nontradables
for which there are supply constraints
Aid
received is saved by recipient
country government
 Aid
receipts leads to accumulation of
foreign exchange reserves in Central Bank

… and, unlike increased aid that is spent, are
not allowed to enter the spending stream
 No
effect on money supply
 No inflation
 No appreciation of nominal exchange rate
 No risk of Dutch disease
Aid
is used to purchase imported
goods that would not have been
purchased otherwise (grants in kind)
 Import
purchases lead to transfer of real
resources from abroad, but not to
increased spending at home
 No effect on money supply
 No inflation
 No appreciation of nominal exchange rate
 No risk of Dutch disease
Aid
is used to buy domestic
nontradables with infinitely elastic
supply due to underutilized resources
(labor and capital) in economy
 Increased demand for nontradables
 Because some resources are unemployed,
greater demand leads to increased supply
 This has a positive impact on production
without increasing the price of
nontradables
 No risk of Dutch disease
Aid
is used to buy nontradables for
which there are supply constraints,
since all available resources are
already in use (e.g., social services)
 Increased
demand for nontradables
 Increased prices for nontradables
 Shift of resources away from the tradables
(exports) and into nontradables
 Real appreciation of the currency
 Dutch disease!
 Monetary
policy response determines if real
appreciation of currency will be caused by
inflation or by nominal appreciation
If foreign currency is used to increase the
reserves of the Central Bank, aid spending on
nontradables leads to an increase in money
supply and to inflation
 If Central Bank sterilizes the impact of aid
spending in nontradables on money supply by
selling foreign exchange, currency appreciates
in nominal terms

To
recapitulate, the risk of Dutch
disease varies, and depends on
 How
aid is used (saved or spent) – CASE 1
 The presence of an aid absorption
constraint – CASE 2
 The impact of aid on productivity in the
nontradable goods sector – CASE 3
 The existence of externalities in the
nontradable goods sector affecting the
rest of the economy – CASE 4
 Aid
can give rise to Dutch disease
when the recipient country’s
government uses the aid to purchase
nontradables rather than imported
goods and when there are constraints
on increasing production in the
nontradables sectors
 The risk of Dutch disease is greater
when aid is used in social sectors that
face constraints on increasing their
production due to resource scarcity
(aid absorption constraint)
How
can recipient countries avoid
translating aid into Dutch disease?
 Save
aid received and increase central
bank reserves (gross, not net) by not
allowing the increased aid to enter the
spending stream
 Use aid to purchase imported goods
 Boost aid absorption capacity in the
nontradables sector
Policymakers
in recipient countries
need to pay attention to potential
early warning signals of aid-induced
Dutch disease such as
A tendency for wages and prices in the
nontradables sector to increase
 A decline in the profitability and sales
of the export and import-competing
industries

 Once
more, the macroeconomic
impact of aid depends critically on
the policy response to aid
 Interaction
between fiscal policy and
monetary policy is crucial
 To
highlight this interaction, apply
two related but distinct concepts
 Absorption:
Monetary policy
 Spending: Fiscal policy
Absorption
 Extent
to which the non-aid current
account deficit widens with increased aid

Captures the amount of net imports financed
by an increase in aid
 Given
fiscal policy, absorption is
controlled by Central Bank’s decision
about how much of the aid-induced
foreign exchange to sell in the markets

If Central Bank uses the full increment of aidinduced foreign exchange to bolster reserves,
aid will not be absorbed
Spending
 Extent
to which the non-aid fiscal deficit
widens with increased aid

Captures the extent to which the government
uses aid to finance an increase in expenditures
 Given
monetary policy, spending is
controlled by the government’s decision
about how much of the aid to spend, on
either imports or non-traded goods

If the government decides to save the full
increment in aid, aid will not enter the
spending stream
 Different
combinations of absorption and
spending define the policy response to a
surge in aid inflows
Absorption and spending are equivalent if aid
is in kind or if it is spent on imports
 Absorption and spending differ when the
government provides the aid-related foreign
exchange to Central Bank and chooses how
much to spend on domestic goods while the
Central Bank decides how much of the aidrelated foreign exchange to sell in markets

 Studies
assessing empirical relevance
of Dutch disease as caused by aid
flows have produced mixed results
 Aid
was associated with real
appreciation in Malawi and Sri Lanka
 Aid was associated with with real
depreciation in Ghana, Nigeria, and
Tanzania
 Ethiopia,
Ghana, Tanzania, Mozambique,
and Uganda experienced a surge in aid
1998-2003 (Berg et al. 2007)
The net aid increment ranged from 2% of GDP
in Tanzania to 8% of GDP in Ethiopia
 High everywhere, from 7% to 20 % of GDP

 In
Ghana, sharp increase in 2001 followed
by a slump in 2002 and another surge in
2003

In all other countries, the surge in aid was
persistent, i.e., after the initial jump, aid
inflows remained higher than before
In
the five countries, no evidence of
aid-induced Dutch-Disease
 Real
exchange rates did not appreciate
during the aid surges
 Only Ghana had a small real appreciation
while the others experienced a real
depreciation

From 1.5% in Mozambique (2000) to 6.5% in
Uganda (2001)
Why?
 The macroeconomic
policy response
was meant to avoid a real appreciation
 Countries
were reluctant to absorb the
surge in aid


Only Mozambique absorbed two-thirds
Aid surge led to reserve accumulation

So, currency did not appreciate in real terms
 Mozambique,
Tanzania, and Uganda spent
most of new aid

They had attained stability, so reducing domestic
financing of the budget deficit was not a major goal
 Ghana

and Ethiopia spent little of the aid
They had a weak record of stability and low
reserves, so reducing the domestic financing of the
budget deficit was a consideration not to spend aid
Two types of policy response
1. In Ethiopia and Ghana, aid impact was
limited because only a small part of it
was either absorbed or spent

New aid was saved and reserves built up
2. In Mozambique, Tanzania, and Uganda,
spending exceeded absorption, creating
a pressure on prices
Money supply expansion was sterilized
through treasury bill sales
 Foreign exchange sales were kept
consistent with a depreciation of
currency to maintain competitiveness

Was
aid-induced Dutch disease a
problem?
No evidence of significant real
appreciation following surge in aid
 Macroeconomic
policy response (fiscal
and monetary policy mix) avoided real
appreciation
 “Not absorb and not spend” vs. “spend
more than absorb”
The
choice in some countries to “not
absorb and not spend” new aid
preserved competitiveness while
allowing the replenishing of
international reserves
The choice in some other countries to
“spend more than absorb” went
along with sterilization of public
spending that contained inflationary
pressures
 Aid
can play a key role in the development
of recipient countries, but it can also
generate macroeconomic vulnerabilities
 Recipients need to implement appropriate
policies to manage aid flows to avoid
macroeconomic hazards

The appropriate policy response needs to take
into account
Potential impact of aid on competitiveness
 Existence of constraints to aid absorption
 Risks linked to aid volatility and to external debt
sustainability

 Aid
is increasingly volatile and unpredictable
 Aid flows are 6-40 times more volatile than
fiscal revenue
 Volatility is largest for aid dependent
countries (Bulir and Hamann 2003, 2007)
 Volatility increased in the 1990s
 Aid delivery falls short of pledges by over 40%
 Reasons for aid volatility
 Donors: Changes in priorities; administrative
and budgetary delays
 Recipients: Failure to satisfy conditions

IMF conditionality often guides donors, helping them
decide if the country’s policies are on track
Impact of large sudden inflows
Supply constraints in absorbing aid
 Real exchange rate overshooting and
volatility
 Negative impact on budget management
 Negative impact on export industries
 Ratcheting up spending commitments
without adequate consideration of exit
strategy
 Infrastructure investment without adequate
planning for recurrent expenditure


Maintenance
Impact of aid promised, but not disbursed
Mismatch between revenues and scheduled
expenditures
 Spending commitments cannot be financed
 Necessitates difficult expenditure choices
 Aid volatility translates into public expenditure
volatility
 Can be costly if it compels government to cut
down on, delay, or abandon productive
investments
 To avoid this, government may resort to printing
money or borrowing
 Hence, negative impact on stabilization


Volatility in money supply, inflation, exchange rates
 Donors
need to disburse aid according to the
agreed schedule and increase transparency
toward recipient country governments
 Recipient countries need to respect the
conditionality of development aid
disbursements
 Recipients need to be granted more flexibility
in their choices to spend or save aid flows,
specifically in light of the time span for the
aid they receive

E.g., during 2000-03, Ghana chose to save
unexpected aid increases and to supplement its
Central Bank reserves
A
substantial acceleration in aid flows
could adversely affect the external
debt sustainability of recipients
 Development aid may take the form
of grants or concessional loans
Grants are unrequited transfers
 Concessional loans increase outstanding
debt and the amount of resources needed
to service that debt

Studies
have shown that debt
sustainability may deteriorate even if
loans are concessional
 Daseking
It
and Joshi (2005)
is crucial for donors to choose an
appropriate mix of grants and loans in
order for recipients to achieve the
MDGs without undermining their
external debt sustainability
Advantages
of grants
 Do
not increase debt burden
 Useful for social projects with uncertain or
delayed returns (health care, education)
Advantage
 Mobilize
of concessional loans
more resources
 Increase debt management capacity
 Useful for projects yielding quick returns
(infrastructure)
Choice
between grants and loans
must balance the benefits of larger
available resources against the risk of
a heavier debt burden
Since loans force recipients to repay
in future, they have an incentive to
 Choose

more profitable projects
This leads better allocation of aid
 Improve
external debt management
Efforts
to find an appropriate balance
between loans and grants can be
based on
 Project-based
approach
 Country-based approach
 Grants
 To
finance investments with a
significant social impact but whose
return is uncertain or difficult to
appropriate or which need a longer
period to be profitable

E.g., education and health care
 Loans
 To
finance projects that yield profits
more quickly

E.g., infrastructure
 An
appropriate balance between grants
and loans is determined case by case

Based on the sustainability of recipient’s debt as
well as its exposure to revenue/growth
volatility
 Poorest
countries receive a larger
proportion of aid through grants
 Countries with higher growth rates and
sound economic policies receive a larger
proportion of loans
Corruption
and economic performance
 Impact
on growth
 Likelihood of disbursement
Anticorruption
strategies
 Reduce
state role
 Improve regulatory environment
 Punish offenders
 Liberalize and reform institutions
Improving
public expenditure
management systems
From
aid fatigue to new initiatives
Aid effectiveness is ambiguous
 Positive
results likely with better policies
and governance
Five Primary Guidelines
 Minimize risks of Dutch disease
 Enhance growth – Always a good idea!
 Assess the policy mix
 Promote good governance and reduce
corruption
 Prepare an exit strategy







Bulir and Hamann 2003, “Aid volatility: An empirical
Assessment,” IMF Staff Papers.
______, 2007, “Volatility of Development Aid: An Update,”
IMF Staff Papers.
Daseking and Joshi, 2005, Debt and New Financing in LowIncome Countries, IMF.
Isard, Lipschitz, Mourmouras, and Yontcheva, 2006,
Macroeconomic Management of Foreign Aid: Opportunities
and Pitfalls, IMF.
Gupta, Powell, and Yang, 2006, Macroeconomic Challenges of
Scaling up Aid to Africa: A Checklist for Practitioners, IMF.
Rajan and Subramanian, 2005, “Aid and Growth: What Does
the Cross-Country Evidence Really Show?,” IMF Working Paper.
______, 2005, “What Undermines Aid’s Impact on Growth?,”
IMF Working Paper.
Aid
can play an important role in
the growth and development of
recipient countries …
…
but it can also create macroeconomic
vulnerabilities
Recipient
countries need to manage
aid flows so as to avoid hazards

Need to consider potential impact of aid on
 Competitiveness
 Constraints to aid absorption
 Risks linked to aid volatility and to external
debt sustainability These slides will be posted on my website:
www.hi.is/~gylfason