1. Lectures 10-12: Small Open Economies
Devaluation in small open economies
The Salter-Swan (NTGs) model
The Dutch Disease

2. LECTURE 10: DEVALUATION IN SMALL OPEN ECONOMIES
Key Question: If a country is too small to affect its terms of trade (i.e., it must take prices of its X & M as given on world markets), does that mean E has no effect on TB or BP?
Answer: No.
Real balance effect can reduce spending.
Output can shift from non-traded sector to traded.

3. After the big devaluations, Mexico in 1994
and Korea
& Southeast
Asia in 1997,
trade balances
“improved” sharply,
even though prices
of their exports
are mostly set
on world markets.

4. For now, assume all goods are traded and PPP holds.
The “real balance effect” of a devaluation.
For now, assume all goods are traded and PPP holds. {
The two formulations
are equivalent.
Three experiments:
1) Monetary expansion →
2) Devaluation →
3) Supply-side growth →

5. Effect of monetary expansion on the balance of payments.
Deficit.
Gradually unwound over time, as reserves flow out.
Figure 19.C.1

6. on the balance of payments:
Effect of devaluation
on the balance of payments:
Surplus.
Gradually unwound over time, as reserves flow in.
Figure 19.C.2

7. LECTURE 10 ½: INTRODUCTION TO SALTER-SWAN MODEL
Key Assumptions:
All Traded Goods (TGs) are aggregated together, => TB becomes: output of TGs minus consumption of TGs.
There is also a 2nd market, in NonTraded Goods (NTGs).
Key results:
Devaluation works also by changing relative price of NTGs.
To attain both internal and external balance (e.g., Y= & CA=0), you need both expenditure-switching and expenditure-reducing policies.

8. TWO ALTERNATIVE DEFINITIONS OF THE REAL EXCHANGE RATE
(I) Two-good model: Q
(II) Small open economy model, dependent-economy, Salter(1959)-Swan(1963), Australian, or NonTraded Goods model “Q”
Instead, “relative P of NTGs,”

9. Salter diagram. Start with a closed economy.
Production of TGs & NTGs coincide with their respective consumption quantities,
if: ● the price mechanism is used to allocate resources and ● total consumer spending = total income.
Salter (1959)

10. Experiment: Increase spending, A
Excess Demand for TG (trade deficit) at point F.
Would require fall in PN if Excess Demand for TG is to be eliminated at B:
output of TG↑ <= via => (XTG-CTG )↑ CTG ↓ (probably) { }

11. Rise in A
must be accompanied by a fall in PN if the Trade Balance is to be kept unchanged.
We have now derived the downward-sloping BB relationship.

12. Experiment: Increase spending, A
Excess Demand for NTG (overheating) at point F.
Would require rise in PN to eliminate Excess Demand for NTG at G.
output of NTG ↑
<= via => (XNTG-CNTG )↑
CNTG ↓ { }

13. A rise in A must be accompanied by a rise in PN if internal balance is to be kept unchanged.
We have now derived the upward-sloping NN relationship.

14. THE SWAN DIAGRAM 
The external balance line, BB, & internal balance line, NN,
divide the A-E space into 4 zones of macroeconomic “illness.”
Swan (1963)

15. SWAN DIAGRAM, continued   The Tinbergen-Meade principle of Targets and Instruments
To attain two goals -- internal and external balance -- you need two independent policy instruments:
expenditure-switching policies (exchange rate)
and expenditure-reducing policies (fiscal or monetary contraction).

16. Two policy experiments
 (1) Fall in Demand: A => recession at point H in fig
If PN is sticky & exchange rate fixed, downward adjustment to point E may be slow & painful.
 (2) Devaluation: E
Improves TB in two ways:
Real balance effect, reduces spending (at H).
Fall in PNTG /PTG , switches spending out of TG, & switches supply into TG (at E).

17. At point H, economy is in recession.
Eventually prices may fall
enough to clear markets.
But with sticky prices,
devaluation can speed up adjustment.

18. L12 Appendix: Rudiger Dornbusch, AER (1973)
“Devaluation, Money & Nontraded Goods”  
Combines NTG model , with MABP
Two automatic mechanisms of adjustment:
(i) PNTG flexible => always on NN;
PN rises instantly in response to ED.
(ii) reserve flows not sterilized;
Money adjusts in response to TD.
 E.g., two experiments
NDA => jump to point G. (Fig. 20.5)
E  => jump to point E. (Fig. 20.6) In each case, over time, reserve flows gradually bring the economy back to S (following the sequence of arrows).

19. THE DUTCH DISEASE
Question: What are the consequences of a natural resource boom in exports?
oil, minerals and agricultural commodities e.g., commodity booms of and

20. BP due to commodity boom:
THE DUTCH DISEASE
BP due to commodity boom:
P natural resource  => TB  or
oil discovery => capital inflow to develop oil; or, by analogy,
KA  due to stabilization or liberalization; or inflow of foreign aid.
Undesired side effect: real appreciation (& crowding-out of non-commodity X.) How?
Under fixed rate, Res inflows => MB  => inflation in PNTG (Also via G ) or
Under floating, appreciation E  => PTG  .
Either way, => (PNTG/PTG) 

21. The Dutch Disease in terms of the Salter diagram
A commodity boom stretches the Production Possibility Frontier rightward (H): can now afford to buy more TGs.
The new LR equilibrium point, E', (external balance & internal balance) now implies a higher relative price of NTGs, inducing land & labor to move out of non-commodity TGs, into the NTG sector.
TB>0 => real apprec. •
Jeffrey Sachs, 2007, “How to Handle the Macroeconomics of Oil Wealth,” in Escaping the Resource Curse, edited by Humphreys, Sachs & Stiglitz

22. Alternative strategy for dealing with inflows:
Movement to point EN′ may be rapid, especially if PNTG is flexible or exchange rate floats.
Alternative strategy for dealing with inflows:
Try to sterilize intervention, to avoid/postpone real appreciation, if BP shift thought temporary, e.g., transitory commodity boom (where resource shifts from EN to EN′ are costly), or if short-term capital inflows thought excessive.
Typically sterilization only works temporarily, especially if capital markets are open.

23. PN ≡ PNTG/PTG NN
BB shifts out. Again: the new equilibrium is a higher PNTG/PTG . But how do we get there? And is it wise, if the boom might reverse? E'
(3)
BB'
Response to Dutch disease. One plausible sequence: (1) Sterilize reserve inflow E
(2)
(1)
(2) Allow inflow to raise money supply BB
(3) Appreciate currency A

24. Another common aspect of the Dutch Disease: governments over-spend, in response to high revenue
For example, the government wage bill goes up – which is difficult to reverse when export revenues go back down.
This is one source of the pro-cyclicality of government spending that is so common among developing countries (esp. Latin America).
References for procyclical fiscal policy:
Gavin & Perotti, 1997
Kaminsky, Reinhart & Vegh, 2004
Talvi & Vegh, 2005
Alesina & Tabellini, 2005
Mendoza & Oviedo, 2006

25. Iran’s government wage bill has been heavily influenced by what oil prices were 3 years before.
Lagged oil prices

26. Indonesia’s government wage bill has been heavily influenced by what oil prices were 3 years before.
Lagged oil prices

27. Correlations between Gov.t Spending & GDP }
Adapted from Kaminsky, Reinhart & Vegh, 2004, “When It Rains It Pours”
procyclical
Pro-cyclical spending
countercyclicall
Counter- cyclical spending
G always used to be pro-cyclical for most developing countries.

28. The procyclicality of fiscal policy, continued
Procyclicality has been especially strong in commodity-exporting countries.
An important development -- some developing countries, including commodity producers, were able to break the historic pattern in the most recent decade:
taking advantage of the boom of
to run budget surpluses & build reserves,
thereby earning the ability to expand fiscally in the crisis.
Chile is the outstanding model;
also Botswana, China, Indonesia, & Korea.
keep 28 28

29. Correlations between Government spending & GDP 2000-2009
procyclical
Frankel, Vegh & Vuletin (2011)
In the last decade, about 1/3 developing countries switched to countercyclical fiscal policy: Negative correlation of G & GDP.
countercyclical

30. The Dutch Disease is one component of the NRC
Natural Resource Curse: The size of the primary sector is not correlated positively with GDP growth
Source: Frankel (2012)

31. Appendix I: The Natural Resource Curse
Seven possible channels that some have suggested could lead to sub-standard economic performance:
* Long-term trends in world commodity prices (Prebisch-Singer hypothesis, But negative trend has not been borne out.) * Volatility (e.g., Hausmann & Rigobon, 2003) * Permanent crowding out of manufacturing (Matsuyama, 1992) * Unsustainability * Civil war (Collier, 2007…) * Poor institutions (Auty, Sachs-Warner, Engerman-Sokoloff…), and * Cyclical Dutch Disease.
JF, “The Natural Resource Curse: A Survey,” 2011

32. Natural resources need not necessarily be a “curse.”
The aim: to get the economic performance of a Chile, rather than a Bolivia;
of an Australia rather than an Argentina
of a Botswana, rather than a Congo;
of a Norway, rather than a Venezuela.
What institutions can best avoid the resource pitfalls?

33. The Dutch Disease & commodity price volatility are two components of the longer-run NRC.
Another important source of the NRC: natural resource abundance may be conducive to bad institutions, including rent-seeking & corruption.
The Engerman-Sokoloff hypothesis (e.g., North America vs. South America):
extraction by mine & plantation => monopoly/authoritarianism/inequality; => societies without private incentives, => ill-suited to develop manufacturing & services.
API Macroeconomic Policy Analysis I Professor Jeffrey Frankel, Kennedy School of Government, Harvard University
Copyright 2007 Jeffrey Frankel, unless otherwise noted

34. Institutional mechanisms to reduce cyclicality of fiscal policy
Independent central banks, to be able to resist political pressure to monetize budget deficits;
Budget rules, to be able to resist pressure to increase in spending overly when revenue is temporarily high;
Well-managed Sovereign Wealth Funds (SWFs) to insulate accumulated assets from pressure to spend (especially in the case of a depletable natural resource), or from temptation to allocate assets on political grounds.
API Macroeconomic Policy Analysis I Professor Jeffrey Frankel, Kennedy School of Government, Harvard University
Copyright 2007 Jeffrey Frankel, unless otherwise noted 34

35. Institutions to fix the procylicality of fiscal policy in commodity-producing countries: The case of Chile

36. Poll ratings of Chile’s
President over time
In 2009, the popularity of the Socialist President of Chile Michelle Bachelet rose sharply (both with respect to handling of the economy and overall), to the highest levels since the restoration of democracy 20 years earlier. More remarkable: the rise in the polls, from very low to very high, came just as the economy moved from rapid growth to slow growth -- not the usual pattern. Why?
Chart source: Eduardo Engel, Christopher Neilson & Rodrigo Valdés, “Fiscal Rules as Social Policy,” Commodities Workshop, World Bank, Sept. 17, 2009

37. Poll ratings of Chile’s Presidents and Finance Ministers
In August 2009, the popularity of the Finance Minister, Andres Velasco, ranked behind only President Bachelet, higher than any other minister since democracy. Why?
And the
Finance
Minister?:
August 2009
Chart source: Eduardo Engel, Christopher Neilson & Rodrigo Valdés, “Fiscal Rules as Social Policy,” Commodities Workshop, World Bank, Sept. 17, 2009

38. In 2008, a copper price spike had looked permanent to many
In 2008, a copper price spike had looked permanent to many. In 2009, the price reverted toward its long run trend.

39. Chile’s structural budget rule
Government must set a fiscal target:
In booms, can only spend structural revenue,
must save the cyclical component.
Structural ≡ economy at full employment & price of copper at its long-run level
Under Bachelet, structural deficit target was 0.
Estimates of structural vs. cyclical are made by commissions of experts, not politicians, which avoids wishful thinking.
In other countries, official fiscal forecasts have optimism bias.
JF, “A Solution to Fiscal Procyclicality: The Structural Budget Institutions Pioneered by Chile,”
API Macroeconomic Policy Analysis I Professor Jeffrey Frankel, Kennedy School of Government, Harvard University
Copyright 2007 Jeffrey Frankel, unless otherwise noted 39

40. Copper prices spot, forward, & forecast 2001-2010
Forecasts internalize the tendency for copper prices to revert toward long-run equilibrium
Copper prices spot, forward, & forecast
spot price
official forecast

41. 2. Hedge commodity revenues in options markets
Innovations to deal with the resource curse: Ways to reduce exposure of economy to volatility in world price of export commodity
1. Index contracts with foreign companies to the world commodity price.
2. Hedge commodity revenues in options markets
3. Denominate debt in terms of commodity price
4. Try Product Price Targeting as a monetary regime
instead of targeting the CPI or exchange rate.

42. Commonly suggested model:
Institutional mechanisms to deal with the resource curse: Sovereign Wealth Funds
Commonly suggested model:
Norway’s National Petroleum Fund (now “Pension Fund”)
When oil prices are high, save it in a fund to offset depletion of reserves.
Internationally diversified.
Even better model:
Botswana’s Pula Fund
Professionally managed; no political interference.
API Macroeconomic Policy Analysis I Professor Jeffrey Frankel, Kennedy School of Government, Harvard University
Copyright 2007 Jeffrey Frankel, unless otherwise noted 42

43. Institutions to deal with the natural resource curse, continued
Extractive Industries Transparency Initiative (UK, 2000)
International oil companies “publish what you pay.”
Nigeria
attempt to save its oil revenues in “excess crude” account.
Proposal to distribute directly to the people - Sala-i-Martin & Subramanian, 2003.
World Bank plan to safeguard Chad oil revenue
revenue would have gone to Citibank escrow account in London;
law dedicated 70-90% for spending on health, ed., & roads,
10% for “future generations fund”
Chad backed out
Collier (2007): International charter: members pledge formal revenue audits.
The World Bank or IMF holds the kitty.
API Macroeconomic Policy Analysis I Professor Jeffrey Frankel, Kennedy School of Government, Harvard University
Copyright 2007 Jeffrey Frankel, unless otherwise noted 43

44. Appendix II: CONTRACTIONARY EFFECTS OF DEVALUATION
Why were the real effects of the East Asia currency crises so severe?
High interest rates raise default probability. The IMF may not have sufficiently realized this – according to J.Furman & J.Stiglitz; and S.Radelet & J.Sachs; both in Brookings Panel on Ec. Activity, 1998.
Devaluation is contractionary: 10 possible channels, including real balance effect & balance-sheet effect.

45. 10 POSSIBLE CONTRACTIONARY EFFECTS OF DEVALUATION
7 Negative effects on AD:
High import bill and low elasticities
Real balance effect (MABP)
Distribution effect: Diaz-Alejandro (1963) MPC urban workers > MPC rich landowners
Balance sheet effect: difficulty servicing $-denominated debts
End of speculative buying: Cooper (1971)
Other aspects of simultaneous stabilization/reform programs -- Removal of import quotas (not causal) -- Monetary/fiscal contraction (not causal)
Tariff revenue rise: Krugman-Taylor (1978)

46. 10 POSSIBLE CONTRACTIONARY EFFECTS OF DEVALUATION (continued)
3 negative effects on AS:
Rise in Pimported inputs , e.g., oil
Rise in W, e.g., where indexed to CPI
Rise in cost of firms’ working capital
TO BE CONTINUED IN LECTURE 22
ON CRISES IN EMERGING MARKETS