Transcript The Macroeconomic Challenge of Aid Volatility

The Macroeconomic Challenge
of Aid Volatility
Andrew Berg
Calvin McDonald
African Department
International Monetary Fund
Two dimensions
Horizon: short (intra-year), medium (1-3
years), and long
Agent: donors and recipients.
Short-run predictability: problems
Lots of volatility, even here.
Mix of reasons: sometimes country,
sometimes “fickle donors”, usually in
between
“innocent of what”
Program and project very different.
Short-run predictability: solutions
Donors: can and are doing better
But move to budget support can make
problem more acute.
Recipients should be able to smooth
IMF programs have come a ways on this
Even if IMF programs allow through
adjusters, headline domestic borrowing
and deficit targets can go awry.
Medium-run predictability:
problems
Again, lots of volatility
First principle component in a panel of aid flows
explains 45 percent of the variation (Kang, et al.
2007)
Macro risks: can cause inflation, exchange rate and
interest rate volatility.
Mainly because of problems of fiscal/monetary
coordination.
May cause increase in size of domestically-financed
public sector
Makes it hard to plan.
Medium-run predictability: solutions
Donors
multi-year commitments etc.
Break link between donation and aid flow: global funds,
Rethink budget support?
Recipients
Create good MTEFs, scaling up scenarios
Smooth aid
How many reserves do you need, and how costly?
Many countries have enough
Surprisingly hard to do.
Complications of exchange rate management
Separate central bank/fiscal decision making and even objectives
Behavior of Aid Over Medium Term.
Aid Flows -Three Year Non-Overlapping Average
30
Aid Flows as Percentage of GDP
25
Ethiopia
Ghana
Malawi
Mozambique
Tanzania
Uganda
20
15
10
5
0
1990-92
1993-95
1996-98
Period
1999-01
2002-04
Behavior of Aid Over Medium Term
- Longer Series.
Aid Flows - Thre Year Non-Overlapping Averages.
25
Three Year Average
20
Ethiopia
Ghana
Malawi
Mozambique
Tanzania
Uganda
15
10
5
0
1975-77
1978-80
1981-83
1984-86
1987-89
1990-92
Period
1993-95
1996-98
1999-01
2002-04
How Many Reserves Do We Need
To Smooth Aid Shocks?
Reserve Adequacy
24
Average Reserve to GDP Ratio.
20
Broken line shows the
reserves needed to cover
the bottom 10 percentile of
negative Aid shocks.
Actual Reserves to GDP Ratio (2003-06)
17.7
17.4
18
16.8
16
12
11.3
8
4.4
4
0
Ethiopia
Ghana
Malawi
Mozambique
Country
Tanzania
Uganda
Country Specific Reserves.
Largest Negative Aid Shock Since 1990.
Negative Aid Shock as Percentage of GDP.
25
Size of the Largest Negative Aid Shock
Actual Reserves to GDP Ratio (2003-06)
20
15
10
5
0
Ethiopia
Ghana
Malawi
Mozambique
Country
Tanzania
Uganda
Long-run predictability
Irreducible and large uncertainty
Donors?
Recipient reactions
Smoothing infeasible and expensive
Aid-led strategy is fundamentally risky
Malawi fertilizer subsidy
Final provocations/thoughts
Countries should figure out how to live with aid
volatility
Vertical funds have merits
Notion of aid/policy interaction may be harmful as
well as a bit flimsy
“We” should be prepared to support alternative/nonaid-led strategies too
We don’t pay enough attention to private capital
flows: aid shocks seem to result in comparable
private outflows on impact.