Growth Week 2011: Ideas for Growth Session 9 - Climate Change, Environment, Natural Resources

Harnessing Oil Revenues in Ghana

Rick van der Ploeg, Radek Stefanski and Samuel Wills*

Oxford Centre for the Analysis of Resource Rich Economies (OxCarre) Department of Economics, University of Oxford

www.oxcarre.ox.ac.uk

OxCarre Conference June 2011

1

http://www.oxcarre.ox.ac.uk/

Overview

• Ghana has discovered oil with estimated reserves of between 780 and 4000 million barrels, but this is being revised upwards frequently

• This is relatively modest on a global scale, though it will still comprise a significant component of Ghana’s GDP

• The oil windfall will also be temporary, so the issue is how to spread the new found wealth between present and future generations.

• Ghana will also have to cope with the notorious volatility of oil prices and the effects this will have on its budget and economy.

• To make the most of this windfall Ghana must consider all aspects of oil production, though our focus is on spending.

• Ghana should spend some of the income upfront to stimulate GDP growth, whilst considering inflation, absorption and Dutch disease. This differs from typical recommendation of establishing a Sovereign Wealth Fund

• Ghana should focus this spending on reducing foreign debt and investing in domestic capital to promote structural transformation of the economy

Harnessing Ghana’s Oil Windfall 2

Ghana has discovered oil

It is a small, temporary and volatile windfall

To harness the windfall they should repay debt and invest in capital, rather than a SWF

Ghana has discovered oil with estimated reserves of between 780 and 4000 million barrels, but this is being revised frequently

0

200

400

600

800

1000

1200

1400

1600

TotalJubilee

Total DWT TotalWCTP

Total Other

p10%p50%

Harnessing Ghana’s Oil Windfall 3 Source: Tullow Oil 2010 full yr results

Total Ghana oil reserves at different probs, m barrels

(non-Jubilee)

Ghana has discovered oil commercial oil reserves in two licences off the eastern coast

These reserves amount to between 780 and 4000 million barrels

This places Ghana at approximately 50th in the world by proven oil reserves, with significantly less oil than major producers


Ghana vs Top 20 countries by proven (90%) oil reserves, m barrels 2010

Source: CIA World Factbook, 2010

The reserves are small relative to Ghana’s population, however they may be significant as a proportion of GDP


0.00.20.40.60.81.01.2

05

1015202530354045

However, Ghana could be in the top 20 countries by reserves/GDP if most of its reserves are accessible

With potentially 160 barrels/head Ghana is small in terms of reserves per capita

Source: CIA World Factbook, 2010

Oil reserves/population, ‘000 barrels per person

Oil reserves/GDP, barrels per dollar

Current planned production from the Jubilee field is likely to be temporary and last for ~20 years, peaking from 2012-2015


Predicted average oil output by year, m barrels

Source: World Bank, 2009, “Economy-wide impact of oil discovery in Ghana”

Ghana collects the revenue from this production through four channels


Oil revenue has four components

Name Size

Royalty GNPC commercial profits Additional Oil Entitlement Income Tax

5% gross

13.75% net profit

10-25% if rate of return 18-33%

35% net profit

Jubilee 90% proven reserves These four components combine to give Ghana’s total oil income

Cumulative oil revenue when oil price=$75/barrel, $ m (2010)

The level of revenue depends closely on the oil price, and may amount to a potentially significant share of GDP and govt income per year


Ghanaian government oil revenue from Jubilee field, % 2010 GDP and % 2010 Govt Revenue

0%

2%

4%

6%

8%

2011 2013 2015 2017 2019 2021 2023 2025 2027 2029

100755030

$/barrel:

Source: World Bank 2009, Team Analysis

The “Additional Oil Entitlement” increases with the oil price

0%

10%

20%

30%

40%

50%

2011 2013 2015 2017 2019 2021 2023 2025 2027 2029

100755030

Oil revenue from Jubilee field as a share of 2010 GDP

Oil revenue from Jubilee field as a share of 2010 Government Income

Ghana will thus have a small and temporary windfall which has particular challenges compared to other resource rich countries


Taxonomy of different types of resource rich countries

Windfall size Windfall duration Challenges Example

Small Temporary • Speed up economic development • Provide for future generations

Ghana

Large Temporary • Speed up economic development • Provide for future generations • Manage absorption constraints • Prevent underutilisation of capital •Prevent inequality and corruption

Nigeria

Large (Large economy)

Long-lasting • Manage oil price volatility to safeguard recurrent spending (mostly government jobs) • Less focus on future generations

Iraq

Large (Small Economy)

Long-lasting • Avoid becoming a rentier state • No absorption constraints due to imports of skilled/unskilled labour and capital

Kuwait

Ghana also has its own specific challenges such as low GDP, low capital, unproductive agriculture and high inflation


Ghana vs the largest 48 oil producers, percentiles

Low GDP/capita

2,001 9,006 43,560

Ghana: 1,192 GDP/Capita 2005 USD

Low physical capital Capital/capita Proportion of US

0.02 0.19 0.92

Ghana: 0.01

Low human capital Human Cap/capita Proportion of US

0.54 0.67 0.86

Ghana: 0.63

Large and unproductive agricultural sector

Labor prod. Agr. 2005 USD

514 3,594 50,874

Ghana: 998

p10 p50 p90 p100

High inflation Inflation rate % pa, Mar 2011

9.20%

Source: World Development Indicatiors, 2011

To address these challenges Ghana must consider all aspects of oil production, though this work focuses on spending

11 Harnessing Ghana’s Oil Windfall

Precepts Overarching Issues

Decision to Extract

Fiscal Regime

Contracts & Operations

Tax & Royalty Collection

Revenue Management

Sustainable Development

International Actors

1. Maximising benefits to citizens

3. Realising full benefit subject to attracting investment, with stable and robust policies

4. Using competition to award contracts and development rights 5. Protecting or compensating the environment and local society

6. Operating nationally owned resource companies transparently and competitively

7. Promoting growth through high levels of investment 8. Smoothing spending through stabilization funds or limited foreign borrowing 9. Effectively spending to increase efficiency and equity

10. Building private investment to stimulate and diversify growth

11. Requiring and enforcing best practice amongst the international community 12. Following best practice amongst resource companies

2. Ensuring openness and accountability

Stage The twelve precepts of the Natural Resource Charter

Source: http://www.naturalresourcecharter.org

Focus of this work

http://www.naturalresourcecharter.org/

The spending decision can be divided into two questions: whether to consume or invest the windfall, and what to consume or invest in

12

Consume or invest the windfall?

What to consume or invest in?

A

B

Harnessing Ghana’s Oil Windfall


13



A

B


The Ghana Petroleum Revenue Management Act (PRMA) has recently been passed, outlining the planned spending/savings mix


Petroleum Account (BoG)

Annual Budget

Petroleum Funds

Long-term investment

Consumption

Stabilisation Fund

Heritage Fund

70%*

30%

30%

min 70%

70%

min30%

Spending

Saving

•Unless otherwise directed by the national development plan, allocated to 11 priorities:

•Unallocated

• Built up quickly to capped level, which is reviewed regularly • Used to cushion the impact of adverse price/production changes •After production ends, combined with Heritage fund for permanent income

• Built up slowly initially, then receives all contributions once stabilization fund established • Used to support welfare of future generations once resources exhausted

•agriculture •transport •rural •security •environment

•human resources •education /health •water/sanitation •institutions/governance •alternative energy

•welfare

*: Mix based on “Benchmark Revenue”, moving avg of past and predicted oil prices and output. Can vary from 50-70% Source: Petroleum Revenue Management Act 2011

A

To assess various spending rules we construct a simple model of an intertemporally optimizing agent who can either consume or save abroad


As a benchmark we take the permanent income, spend-all and bird-in-hand rules

Household chooses consumption and foreign assets to maximise intertemporal utility

Consumption is perfectly smoothed, and the permanent income from the windfall is consumed, when r=ρ

The spend-all rule dictates that all oil income is consumed as it is received

The bird-in-hand rule consumes a fixed proportion of foreign assets: 4% in the case of Norway

The permanent income rule is then adjusted for a range of assumptions

Substitutability and impatience

Finite lives (Blanchard Yaari constant hazard rate)

Productivity growth

Population growth

Precautionary savings

Dynamic programming following Skinner (1998), see Backup

A

On the spectrum of spend/save options, The PRMA is closer to the spend-all than the permanent-income or bird-in-hand rules


Spend all

*: Assumes oil price is constant at $75/barrel. r=2.5%. Ignores current debt position Source: Team analysis

0%

1%

2%

3%

4%

5%

2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 2031 2033

-10%

0%

10%

20%

30%

40%

50%

201120132015201720192021202320252027202920312033

Consumption*, % 2010 GDP

Assets*, % 2010 GDP

PRMA

PRMA

Permanent Income

Spending options Benefits Spending and Asset Profile

• Better returns earned through domestic investment than foreign Sovereign Wealth Fund

• Discounting future generations welfare (impatience); e.g., finite lives

•Substitutability of generations welfare

• Spending stability

• Less inflation and currency appreciation

• Less Dutch disease

• Precautionary savings

• Population growth

• Absorption constraints

Elaborated in next slides

A

Spend All Permanent Income

PRMA Bird in Hand (4%)

Spending the windfall upfront can make sense if policymakers have a short decision horizon or are very utilitarian

Harnessing Ghana’s Oil Windfall 17 *: Assumes oil price is constant at $75/barrel. Ignores current debt position Source: Team analysis

-30%

-20%

-10%

0%

10%

20%

30%

40%

201120132015201720192021202320252027202920312033

Consumption*, % 2010 GDP

Assets*, % 2010 GDP

Adjustment Description Spending and Asset Profile

• Impatience describes the rate at which future periods are discounted • This analysis assumes the real rate of interest r=2.5%, ρ=20% and EIS=0.5 • It could be thought of as a 17.5% chance of the government being removed from office each year

•Finite lives may be another reason why policymakers are impatient • We use a stylised adjustment setting the average lifetime to 61 years (Blanchard constant death rate=1.64%)

Impatience

Finite Lives

Substitutability Impatience


Substitutability

• Utilitarian (perfect substitutability between utility of different generations, EIS = ∞) versus Rawlsian (no substitutability, EIS = 0). • More substitutability brings consumption forward, so that it is not affected by discounting •This analysis assumes r=2.5%, ρ=20% and EIS=1 (spending peaks at 7.5% of GDP)

Finite Lives

A

0%

1%

2%

3%

4%

5%

2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 2031 2033

Comparison of spending rules to permanent income baseline

Alternatively, saving beyond the PI rule may make sense if there is population growth, though precautionary savings is only a minor concern

Harnessing Ghana’s Oil Windfall 18 *: fit to annual Brent Crude data, 1970-2010 Source: Team analysis

0%

1%

2%

2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 2031 2033

-10%0%

10%20%30%40%50%60%70%

201120132015201720192021202320252027202920312033

Consumption, % 2010 GDP

Assets, % 2010 GDP

Adjustment Description Spending and Asset Profile

• Precautionary savings is when people save more when income is volatile as a buffer against future income shocks. This delays consumption • This assumes oil is the only source of income • CRP=3 and CRP=11 (very conservative) • P_O=$75, StdDev_O=24*

• Population growth delays consumption to allow more when there are more people in the future •This analysis assumes population growth at 1.85% (Ghana 2011 rate)

Precautionary Savings

Population Growth

Prec Saving CRP=3 Prec Saving CRP=11


A

Zoomed In

Comparison of spending rules to permanent income baseline Pop’n Growth

On balance, Ghana’s windfall spending can be brought forward relative to the PI rule. What it should be spent on is discussed in the next section


Spending rule

Precautionary Savings

Impatience

Finite Lives

Substitutability

Population Growth

Permanent Income

PRMA

Spend All

GDP growth PI

Importance

Spend now

Save now

High

High

N/A

High

Medium

Low

High

Medium

Medium

Comment

•If properly considered then should borrow heavily to smooth consumption across generations

• Too aggressive due to inflation and absorption constraints

•Less spending than spend-all and is just as volatile which is not so good. However, does accumulate some assets for future •A more utilitarian social welfare function leads to much more

consumption by present generations at expense of future generations made possible by large-scale borrowing.

•Consumption is much more upfront if politicians are myopic due to the fear of being removed from office.

•Allowing for finite lives (no bequest motive) implies more consumption upfront and less in future, so initially borrowing and less asset accumulation in the long run than the PIH.

•Good benchmark for developed countries, but development needs mean more should be spent upfront in countries like Ghana

•Prudence leads to less consumption upfront and more precautionary buffers

•Realistic population growth also leads to less consumption upfront and to a gently rising stock of assets.

A


20



A

B


Once the decision of whether to consume or invest the windfall is made, Ghana must decide what to consume or invest in


Investment

Consumption

i Repay foreign borrowing

ii Accumulate foreign capital (SWF)

iii Accumulate domestic capital

iv Citizen dividends

v Lower taxes or higher public consumption

vi Subsidies to specific industries or consumers

We focus on investment

B

i) Repaying foreign borrowing may reduce spreads and the risk of lower creditworthiness due to resource-driven conflict


Lower foreign borrowing will reduce interest rate spreads, boosting capital accumulation and development

Lower foreign borrowing will also reduce the risk of reduced creditworthiness due to resource-driven conflict

Ln bond spread residual vs debt/GNI residual, Ln bond spreads vs resource exports/GDP

•Expect oil wealth to improve credit worthiness and lower spreads •But, in more fractionalized, corrupt societies oil wealth may cause conflict and civil war (Collier, Hoeffler) •Creditworthiness falls and bond spreads rise

B

Slope = 1.89

Source: van der Ploeg and Venables (2011)

i) If Ghana faces an increased cost of borrowing due to foreign debt, it is optimal to postpone windfall consumption and quickly repay borrowings

Harnessing Ghana’s Oil Windfall 23 Source: van der Ploeg and Venables (2011), World Bank Datacentre, Bloomberg (2011)

If Ghana’s debt is increasing the cost of borrowing it can be represented by a kink in the interest rate

* for and * ( ) * for r r F F r r F r F F= ≤ = +Π > >•Kink allows interest premium and endogenous choice of F in steady state

Ghana may be facing high interest spreads due to its stock of foreign debt

•External stock of public debt is ~37% of GNI (2009) • S&P rates 2007 10yr $750m Eurobond as “B”, 3 steps below Egypt’s “BB”. Yield range from 6.7-7.1% past 6 mth

The interest rate premium makes it optimal to postpone consumption in the short term to repay debt

•Solving a standard CRRA maximisation, with a risk premium on debt:

•The inclusion of the debt premium alters the Euler equation, depending on the level of debt

•Both converge to the steady state

B

i) Following a windfall, consumption should rise slightly and debt should be repaid quickly. A SWF is only suitable if the windfall is large


The dynamics from a small and large windfall can be expressed in a phase diagram

Small windfall

Large windfall

Initial jump in total consumption:

λu is eigenvalue with positive real part > r*, so smaller in a smaller capital-scarce economy as Π′ pushes up λu.

It shows that a SWF is only suitable if the windfall is sufficient to completely reduce the interest premium

Time Consumption, C Debt, F

After announcement

Consumption path jumps up

Borrowing, increasing level of debt

During extraction

Steep increase in consumption

Rapid pay down debt

Small windfall case:

After depletion Resume growth path but ‘further along’ the development path

Large windfall case:

During extraction

Run debt down to during extraction. Start building SWF

After depletion Support permanent increase in consumption from interest on SWF

F


B

i) Now, if there is endogenous capital formation, then the windfall should be spread between debt repayment and public infrastructure


We assume the government can choose public capital, transfers and taxes to maximise welfare

•Government chooses time paths for lump sum transfers T, distortionary taxes τ, and public capital stock S, and hence paths of K, Y, W, D, C, to maximise

•Production with private & public capital:

•Foreign capital supply:

(Marginal product of capital = world interest rate)

•Consumption: C =W + T

•α=0.4, γ=0.25, ρ=r*=0.05, σ=0.75, ψ=0, δK=δS=0.05

)==>=− −− τατ γαα ,(*)1( 11 SKKrSLK

[ * ( )]D r D D G T S S N Yδ τ• •

= +Π + + + + − −

1 1/ 1 1/

0exp( )

1 1/C G t dt

σ σψ ρσ

− −∞ +− −

∫

γαα SLKY −= 1

This lets us find the optimal mix of policy after a resource windfall

When transfers are available, taxes are set to zero, and the windfall is spent on public capital and repaying debt

•Optimal income tax rate is zero

•Intratemporal smoothing:

•Optimal infrastructure:

•Optimal time profile of private consumption:

G Cσψ=

( ,0) * ( ) '( )S SW S r D D D δ= +Π + Π +

[ ]( ) '( ) for , else 0C C D D D D D Cσ= Π + Π > =

When transfers aren’t available, consumption must rise through lower taxes and higher public infrastructure

•Private consumption is now C = W(S,τ) •Marginal Cost of Public Funds increases with tax rate: - Depresses demand for public vs private C - resources relax this •Public infrastructure:

1 11

1 1

φα τα τ

= > − − −

W( , )G Sσ

ψ τφ

=


B

i) If lump sum transfers are possible, income taxes are stopped, debt is repaid and public capital is accumulated


No windfall Anticipated windfall

Response of economy to anticipated temporary windfall


When a windfall is announced (division of 1st yr resource revenues): -Transfers rise (68%) - Debt is quickly repaid (11%) - Public capital is accumulated (21%)

This results in: - lower r - more private K -higher output - higher wage - high consumption brought forward

B

i) If lump sum transfers are not possible, income tax is reduced, debt is repaid and public capital is accumulated


No windfall Anticipated windfall

Response of economy to anticipated temporary windfall


When a windfall is announced: -Tax falls - Debt is quickly repaid - Public capital is accumulated

This results in: - lower r - more private K - higher output - higher wage - high consumption brought forward

B

ii) If the windfall is large enough, accumulating foreign capital in a Sovereign Wealth Fund may help with volatility and absorption


Arguments for and against setting up a Sovereign Wealth Fund

For

Against

Argument Discussion

1. Providing for future generations

2. Smoothing against oil price volatility

3. Holding funds temporarily until absorption constraints are alleviated

Part Next slide >> Part iii)

A

1. Greater marginal benefit from current consumption or investment in domestic capital

B

Part iii)

B

B

ii) Oil volatility is a major part of the resource curse and should be managed by hedging, stabilisation funds and a flexible economy


Hedging

• Use derivatives to hedge against adverse price movements

• Used by Mexico (spent $1.5bn on option, earned $8bn), Ecuador, Colombia, Algeria, Texas, Louisiana

• Unlikely to become widespread: • Political risks when lose • Market impact of hedging (information and market power)

Mexico oil export price, $ barrel

B

ii) Oil volatility is a major part of the resource curse and should be managed by hedging, stabilisation funds and a flexible economy


Stabilisation Fund

Flexible Economy

Stabilisation funds should be considered separately to “future generations” funds:

The size of a stabilisation fund should be determined according to four criteria:

• Cost of volatility to the domestic economy?

• Opportunities for borrowing in downturn?

• Stochastic process governing resource?

• Political risk – fund is lootable?

StabilisationStab/SavingsSavingsNone

It is impossible to fully insulate an economy from oil price volatility

Therefore, the domestic economy should be designed to handle residual volatility

SWF of 31 oil producers, 2005

Source: IMF

•2008-early 2009, MENAP FOREX reserves fell $40 bn and non-oil growth fell 5% points. •There were transmission channels other than revenue: - Resource sector investment - Other private sector responses

- Capital mobility – Zambia

• Encourage flexible labour and capital markets • Avoid hard to reverse commitments • Diversify…..

B

ii) A sovereign wealth fund can be used to smooth “Dutch disease”: a contraction of the traded sector and a real appreciation during an oil boom

Managing Resource Revenue in LIC's 31

•Oil output increases

•Spending rises on traded (T) and non-traded (NT) goods

•T goods can be imported, but NT goods must be produced domestically

• Labour (and capital) switch from T to NT

• The relative price of NT also rises – a real appreciation

Wealth effect

Substitution effect • If total labour (L) fixed, workers will move from T to NT as NT goods can’t be imported

• If total labour (L) flexible, workers still leave T as these goods are imported, but they choose to retire instead

Dutch disease overview Dutch disease simulations

These effects will be mitigated if capital

and labour are imported

B

ii) Sovereign wealth funds can also be used to park funds temporarily abroad to avoid absorption constraints binding

• If there are absorption constraints, i.e., it takes nurses to train nurses, it takes roads to build new roads, etc., there may be real absorption constraints so that windfall can in the short run not be properly spent.

• In that case, the real exchange rate will appreciate and reverse back as absorption constraints are relaxed.

• This happens via gradually running down capital in the traded sector via wear and tear if traded sector is capital intensive or via gradual build up of home-grown capital if non-traded sector is capital intensive.

• Message is that there may a justification to temporarily park revenue from windfall abroad until domestic capacity is big enough.

• Must avoid investing in white elephants. • See van der Ploeg and Venables (2010)

Managing Resource Revenue in LIC's 32

B

iii) To complement debt reduction, accumulating domestic capital will boost GDP and begin structural transformation


B

Low GDP

Capital Investment

Structural Transformation

Smooth Transition

• Ghana has both low GDP and low GDP growth •This can be attributed to all sectors

•To boost GDP growth Ghana must invest in domestic capital •Traded capital can be imported •Non-traded capital must be “home-grown”: teachers teaching teachers

•Investment should be in: •physical capital (infrastructure ) •human capital (education and health) •stimulating risk taking, entrepreneurship and R&D (via generic tax subsidies).

•Domestic investment will begin the structural transformation away from agriculture, which should be promoted

•Now that the PRMA is passed this is the major question facing Ghana

•Although there will be pressure to support agriculture, this should be done only to smooth the transition to more productive industries

Elaborated in following slides

iii) Ghana’s GDP per worker is low and growing slowly, driven largely by small and slow growing capital stock

Harnessing Ghana’s Oil Windfall 34 Source: Penn World Table, UN, IFPRI, Own Calculations

Ghana has low GDP per capita and low GDP per capita growth

This is driven in large part by a small and slow-growing capital stock

B

Contribution to GDP growth, 1993-2007

Growth accounting following Caselli (2005):

Data flaws mean employment in manufacturing is overpredicted in Nigeria

iii) This suggests Ghana is far from its steady state. To analyse the effect of the oil windfall we therefore must capture its transition path

We use a simple three sector model to generate structural transformation and capture Ghana’s transition path

This is driven by exogenous growth rates and different factor intensities in each sector that drive overall growth

•Non-homothetic preferences for agriculture • Exogenous growth. Highest in Manufacturing, Services then Agriculture • Services are L intensive, manufacturing is K intensive • As capital accumulates, draws labour into M, then S (eg Rybczynski effect for 1 country over time)

Source: Acemoglu and Guerrieri (2006); Gollin et al (2002)

Parameterisation of simplified Acemoglu and Guerrieri (2006) model

B

iii) By including growth we find that the importance of oil declines with time. The level of Dutch disease depends on the stage of development

As the economy grows the relative size of the oil shock declines

The shock causes a small reallocation of factors from T to NT, the extent will depend on the stage of transformation.

“Dutch disease”

Source: van der Ploeg, Stefanski and Wills (2011)

•The model is fitted to data from 1993-2007. It also assumes constant growth rates, based on these years. This explains why there isn’t a large hump in manufacturing factor shares.

B

• Without capital markets (relative to a world with no oil) capital stock should be driven down in anticipation of the shock

•Once shock hits, capital should be accumulated

•With international bond markets, the effect is weakened but still dominant

iii) As well as sector effects, the optimal response of total capital is to fall before the shock, and accumulate during it, to smooth consumption

Optimal response of total capital to oil shock, expressed as ratio of K in oil vs non-oil economy

Source: van der Ploeg, Stefanski and Wills (2011)

B

•Relative prices follow similar path to capital – first depreciation of RER (as capital is driven down) then appreciation as boom hits

•This reflects the anticipation effect and the higher labor intensity of the non-traded sector :

• Since capital declines initially, labor more abundant relative to capital

•Price of sector that uses labor more intensively (NT) goes down

•As capital increases relative to labor, opposite effect

•Notice the relatively small magnitudes! Reflects small oil find and (assumed) flexibility of labor and capital.

Optimal response of P_S/P_M, expressed as ratio of oil vs non-oil economy

iii) The fluctuations in K cause a small real depreciation then appreciation, as capital becomes relatively scarce then abundant B

iii) To boost growth Ghana should invest in domestic capital, especially as the largest sector (agriculture) is the least productive

Harnessing Ghana’s Oil Windfall 39 Source: Kuralbayeva and Stefanski (2011)

Sectoral employment estimates, labour productivity and TFP

1%

1%

12%

31%

55%

Mining and utilities

Construction

Manufacturing

Services

Agriculture

21354

18280

1011

7462

849

14

341

35

180

2

Sector

Employment Share %

Labour Productivity 2005 USD

TFP Levels

•Large size and low productivity are linked (Lagakos and Waugh). •Cocoa has been crucial for combating poverty.

B

iii) By investing in domestic capital Ghana will raise its genuine savings rates


-10

0

10

20

30

1970 1975 1980 1985 1990 1995 2000 2005

Ghana Malaysia Venezuela Kuwait

Adjusted net savings (Genuine savings)* 2008 excluding pollution damage, % Gross National Income

*: Gross savings – depreciation of fixed capital + education expenditure – depletion of natural resources Source: World Bank

-10

0

10

20

30

1970 1975 1980 1985 1990 1995 2000 2005

East Asia & Pacific Sub-Saharan Africa Latin America & Caribbean

Ghana’s genuine savings rate is currently negative

As is Sub-Saharan Africa’s as a whole

B

iii) This will involve investing in education to boost intangible capital, which is the main creator of wealth

0%10%20%30%40%50%60%70%80%90%

100%

Low Middle High (OECD) World

Natural

Produced

Intangible


Composition of wealth, $ per capita and % share, 2000

Note: All dollars at nominal exchange rates. Oil states excluded. National wealth is PV sustainable consumption 2000-25 using discount rate of 4%. Produced capital from PIM. Source: World Bank (2006, Table 2.1).

Total Wealth $ per capita

7,532 27,616 439,063 95,860

Type of capital

Country income group

B

Expanded next slide

iii) And moving away from the reliance on natural capital which characterises low income countries

0%

5%

10%

15%

20%

25%

30%

Low Middle High (OECD) World

Pastureland

Cropland

Protected

NTFR*

Timber

Subsoil


Composition of land resources, percent of total wealth, 2000

*: NTFR = Non-Timber Forest Resources Source: World Bank (2006, Table 1.2).

Total land resources

$ per capita

1,925 3,496 9,531 4,011

Type of land resource

Country income group

B

Summary


• Ghana has discovered oil with estimated reserves of between 780 and 4000 million barrels, but this is being revised upwards frequently

• This is relatively modest on a global scale, though it will still comprise a significant component of Ghana’s GDP

• The oil windfall will also be temporary, so the issue is how to spread the new found wealth between present and future generations.

• Ghana will also have to cope with the notorious volatility of oil prices and the effects this will have on its budget and economy.

• To make the most of this windfall Ghana must consider all aspects of oil production, though our focus is on spending.

• Ghana should spend some of the income upfront to stimulate GDP growth, whilst considering inflation, absorption and Dutch disease. This differs from typical recommendation of establishing a Sovereign Wealth Fund

• Ghana should focus this spending on reducing foreign debt and investing in domestic capital to promote structural transformation of the economy

Ghana has discovered oil

It is a small, temporary and volatile windfall

To harness the windfall they should repay debt and invest in capital, rather than a SWF

Background material

44

It is also likely to affect the exchange rate as it will amount to a large component of exports per year


-

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Food Manuf Agriculture Ores andMinerals

Fuel Oil at peakproduction*

2008 Ghanaian merchandise exports by sector, $ million

*: Based on 120,000 bopd at USD 75/barrel Source: World Bank WDI, 2011

Wealthy future generations are often used to justify upfront spending, though if taken seriously spending should rise even further


-

2,000

4,000

6,000

8,000

10,000

12,000

2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 2031 2033

No oil, no growth

Oil, no growth

Oil and growth (0.5% pa)

A

PI rules for government spending under different growth and oil assumptions, $ m (2010)

*: Based on 20% of GDP accounted for by government expenses Source: World Bank data, team analysis

If the wealth of future generations is properly considered, then government should borrow heavily now to smooth consumption across generations

The effects of precautionary savings were solved using dynamic programming


Dynamic programming methodology, following Skinner (1998)

The stochastic Euler equation is given by:

The second order Taylor expansion is:

Simlifying and solving recursively gives:

This gives the full system of equations:

Ultimately Ghana should focus on spending upfront to stimulate GDP growth, whilst considering inflation, absorption and Dutch disease


A

Mature economy

Developing economy

Aim Policy

•Protect wealth for future generations

• Invest windfall abroad in SWF - Diversify amongst bonds, equity and real estate, ensuring it is orthogonal to the stochastic path of the oil price

•Develop wealth for future generations

• Invest windfall domestically: - Focus on generic ways of promoting entrepreneurial spirit and research & development, eg. education, health, infrastructure

With these aims in mind, Ghana’s Petroleum Revenue Management Act looks to be appropriate. The challenge now is to avoid the pitfalls

Pitfalls

• Fund governance - Ensuring it is preserved for future generations

•Political bias - investing in illiquid, partisan projects to avoid political rivals raiding liquid, non-partisan funds •“White elephants” - high visibility, low use investments •Absorption constraints - Some capital must be “home-grown”: eg. existing teachers must train new teachers - This may justify temporarily investing abroad. •Inflationary pressures •Dutch disease

•For explanation see backup

iii) Ghana’s low TFP growth can be attributed to all sectors, and is lagging behind other oil exporters like Nigeria and Malaysia


All sectors have low TFP growth in Ghana. But, lowest growth has been in industry and agriculture. Nigeria and Malaysia have had more TFP growth in agriculture and services.

Source: UN, IFPRI, Own Calculations

TFP decomposition by sector: Total to Agriculture, Industry , Services; 1970=1

B

iii) Genuine savings rates are a particular risk for Ghana as a resource rich country, many of which have negative intangible capital


Intangible capital, $ per capita and percentage share of total wealth

Source: World Bank (2006, pg 29).

Intangible capital $ per capita

6,029 2,176 1,173 4,360

Type of capital -400

-300

-200

-100

0

100

200

300

Natural Produced Intangible

-3,215 -1,598 -3,418 -1,959 -12,158

Intangible capital % total wealth

B

iii) As Ghana develops, labour will shift out of agriculture as part of the structural transformation process, which is happening in Malaysia


In contrast to Ghana and Nigeria, Malaysia has been making a steady transformation of moving people out of agriculture into industry and services.

Sectoral employment share, %

Source: UN, IFPRI, Own Calculations

B

iii) Cocoa is a large part of the unproductive ag. sector and may be hurt by the transformation, though supporting it sustains low growth


•A large part of the unproductive agricultural sector is cocoa.

•Dutch disease effects (appreciation of the currency) may especially hurt the export of cocoa and thus the livelihood of many Ghanaians.

•One should be cautious of supporting this sector which could exacerbate low aggregate TFP growth.

•There may be more efficient ways to alleviate poverty.

Share of sectoral value added, % 2005

Source: IFPRI

39%

8% 9%

11%

33%

Agr. Mining Mfg.Cstrn. Serv.

Share of sector: •Cocoa: 16% •Yams: 12%

B

Finally, consuming the windfall has been done in a number of ways with varying levels of success

• (i) Citizen dividends: Alaska hands out the windfall to its citizens. The idea being is that the natural resources belong to them and that they know best what to do with it.

• (ii) Lower taxes or higher public consumption: Another way is to let the oil revenue flow into a fund and withdraw say 4% from it each year for the general budget as Norway does. This can then be used for cutting taxes (higher private consumption) or raising public consumption.

• (iii) Subsidies: The Netherlands has used its gas windfall to raise welfare benefits in the 1970s and 1980s (but later used it for a fund for investing in the domestic infrastructure). Iran, Kazakhstan, Netherlands and many other countries use the windfalls for fuel subsidies to consumers or ‘pet’ industries, but that is very inefficient indeed. Better is to use the windfall in that case for conditional transfers (e.g., to stimulate education or risk taking).


B

Ghana was previously a net-importer of oil, but net exports are soon to comprise more than half of production

54

0

20

40

60

80

100

120

140

Production Net Import Consumption

Ghana oil production and consumption, ‘000 bbl/day 2009 vs 2012

*: Assuming constant oil consumption for comparison Source: CIA World Factbook, Tullow Oil

0

20

40

60

80

100

120

140

Production Net Export Consumption*

2009 Production 2012 Production

Ghana experienced a period of hyperinflation during the 1970s-80s, which has since come under control but remains high

55

CPI Inflation, % pa

Source: World Bank Datafinder

Military Coups 1966-81

Ghana’s economy is largely focused around agriculture and non-traded services

56 Source: UN, IFPRI, Own Calculations

0%

5%

10%

15%

20%

25%

30%

35%

40%

Agriculture Services Mining andUtilities

Construction Manufacturing

Value Added

Employment

Value added and employment share of total, % 2007

Ghana’s agricultural sector is dominated by cocoa, and this accounts for 1/5 of global production

57

Cocoa is the largest part of Ghana’s agricultural sector Ghana accounts for approximately one fifth of world production

5%

1%

1%

1%

1%

1%

2%

2%

4%

4%

5%

5%

6%

Other

Rice

Cocoyams

Other meats

Sorghum and millet

Plantains

Maize

Fishing

Cassava

Vegetables (domestic)

Yams

Forestry

Cocoa beans 13%

2% 2%

2% 4% 4%

5% 6%

9% 11% 12%

14% 16%

Product Share of GDP % of GDP

Share of Agr % of Agr

Total Agriculture 39% 100%

Source: GSS, IFPRI, Own Calculations, UNCTAD

1

3

4

5

5

10

13

21

38

Malaysia

Ecuador

Brazil

Cameroon

Nigeria

Other

Indonesia

Ghana

Ivory Coast

Country Cocoa production % world, 2005

Total World 100%

Gujarat Pollution Control Board : Improving Industrial Pollution Control

Hardik Shah Member Secretary

Gujarat Pollution Control Board Gujarat, INDIA

IGC 21 September 2011

1

Gujarat

Area 196,024 sq.km. (5.96 % of India)

Capital Gandhinagar

Climate Tropical

Population 50.60 million as per 2001 census (4.93% of India)

Urbanization 38 % (Compared to the national average of 28%)

Population Density 258 persons per sq.km. vis-à-vis 324 of national average

Official Language Gujarati

Net State Domestic Product Rs 1,050,230 million (=US$ 22,036 million) in 2001-02

Share of secondary sector in SDP 38.5% in 2001-02 at current prices

Per capita income (in 2009-2010) Rs 21,276 (=US$ 446) 2

GUJARAT LAND OF MAHATAMA GANDHI

The Mother Earth Provides for Needs of Everyone

But Not for the Greed

of Everyone

Economic Snapshot

5

Gujarat- Strong Industrial Base

6

GPCB’s monitoring of industrial emissions includes three strategies:

• Regulatory inspections of industrial plants • However, in the face of high industrial growth, staff time constraints limited GPCB’s

in-house capacity to expand inspection operations

• Court-mandated third-party environmental audit programme • However, concerns about auditor objectivity exist, since industry selects and pays

auditors

• Third-party environmental monitoring involving Technical Institutes • However, it can only be complimentary and not substitutive to GPCB’s monitoring

• GPCB tested two innovative solutions to these challenges.

• GPCB partnered with external evaluators to measure the impact of changing these two programmes.

Challenges to Regulating Industrial Pollution

Making environmental audits independent

• Auditors paid from central pool, and not by individual firms

• Auditors randomly assigned to firms, not chosen by them

• Audits back-checked by independent team from a local technical university; auditors’ payments based on their accuracy

Question: Would changing auditors’ incentives make reporting more accurate? Do reliable audits induce plant compliance?

The Innovations

• Under the status quo (control group), auditors often reported readings just below the norms. Their reports were much lower than the readings from random back-checks conducted by the evaluators.

• Under the modified programme (treatment group), auditors reported significantly higher pollution readings consistently.

• Auditors who used to report readings just below PCB norms now reported higher readings that matched back-checks

Evidence from the Evaluation

9

• The preliminary results from this evaluation were shared with GPCB officials and third-party auditors

• Auditors suggested that adopting parts of the modified audit programme permanently would improve the quality of work they are able to provide

Using evidence for policy change

10

GPCB may consider changes to the audit policy in response to this evidence and feedback from the auditor conference. • GPCB centrally administers a random

assignment of auditors to firms, instead of allowing firms to select an auditor.

• GPCB sets a fee structure for audits and verifies that auditors are paid accordingly, instead of allowing firms and auditors to negotiate a price.

Using evidence for policy change

11

• GPCB is partnering with researchers to test another pilot programme for air pollution regulation with two components:

• Continuous emissions monitoring (CEMs) • Gives GPCB more detailed information on the total

load of particulates emitted by industry

• Emissions trading system • Tests the use of market-based regulatory

instruments to reduce airborne particulate matter

A Continuing Collaboration

12

India-Environment Protection-Religion

• ALL IN THIS MANIFESTED WORLD, CONSISTING OF MOVING AND

NON-MOVING ARE COVERED BY THE GOD. USE ITS RESOURCES

WITH UTMOST RESTRAINT. DO NOT COVET THE WEALTH OF

OTHERS.

– UPANISHAD (RELIGIOUS GRANTHA WRITTEN CENTURIES AGO)

Thank You

14

Marketing Improved Cook-Stoves

Mushfiq Mobarak

Yale School of Management

[Primary Collaborator: Grant Miller (Stanford Medical School)]

Drawing on collaborative projects with BRAC (Bangladesh), Rob Bailis and P. Dwivedi (Yale FES), Sandro Gomez (Yale Mech. Engr.), S.

Barnhardt (IFMR, India), Biolite Stove (USA)]

Understanding the Low Demand

• Inexpensive welfare-improving technologies are often not adopted by poor households – Insecticide treated bed-nets, new varieties of seeds

and fertilizer, improved cook-stoves, migration

• Puzzle: Why do so many rural households refuse to adopt stoves even when the benefits are not external?

– ARI: leading killer of children under 5 worldwide. Accounts

for 22% of all non-communicable child deaths (WHO 2005) – Biomass combustion is the leading environmental “risk

factor” for female mortality worldwide

Southern Bangladesh, Weds 8/9

0

1000

2000

3000

4000

5000

6000

7000

11:00 11:15 11:30 11:45 12:00 12:15 12:30 12:45 13:00

Time (local Bangladesh)

10-sec averagesavg over entire cooking periodavg over 30 mins

U.S. 24-hr PM2.5 Standard = 65 ug/m3

Cook not in kitchen ~70ug/m3 1300 ug/m3

640 ug/m3

Why Don’t People Adopt? • Some hypotheses:

– Lack of liquidity, Information failure or learning externalities (inefficiently low experimentation), Intra-household externality, “Taste” and tradition

• Disentangling different reasons for adoption has important policy implications – do we need to address costs, risk aversion, a stove attribute (food taste), or an information failure?

• De we push existing technologies or do we need to develop new ones that people like better?

Forming hypotheses

Traditional stove

Improved stove 1: Portable

Improved stove 2: Chimney

2900 Households in 58 Villages, 2 Districts

2100 Households in 42 Villages 800 Households in 16 Villages

Stove at Full Price

Stove at Half Price

No Opinion-Leader Information

A

B

Publicizing Opinion-Leaders’ Adoption Decisions

C

D

Husband Makes Choice

Wife Makes Choice

Choice of Free Chimney or Free Portable Stove

E

F

Choice of Tk. 250 Chimney or Tk. 50 Portable Stove

G

H

Experiments

Pricing Results • Highly price-elastic and non-linear demand. • Very low adoption at education plus “financially-

sustainable” pricing • Inelastic chimney demand implies households less

elastic with respect to health costs than time costs • The “refusal rate” (drop from stove orders to stove

purchase) highly positively correlated with price – Suggests that liquidity / savings constraints are key

• Adoption far from universal even when free – Non-price factor (e.g. stove characteristics) matter

Opinion Leaders

• Asymmetric Effects – unanimous ‘no’s generally have a stronger effect on behavior – If the leader adopts, it’s not necessarily right for me,

but if he doesn’t, then it cannot possibly be right for me.

• OL influence larger for portable stove than the chimney stove

• After households gain more experience with stoves, the OL influence smaller, and the difference between chimney and portable stoves almost disappears

• In the free treatment, women prefer stoves, and they prefer the healthier chimney stoves – Women have larger valuation for own and child health

• Once we start charging for stoves (and relative price of chimney stove is increased), women less likely to purchase altogether, and shift towards the cheaper stove (relative to men) – Women more liquidity constrained, and cannot act on their

preferences

Stated Adoption

Total No Stove Yes Stove Portable ChimneyE - Men Choice, Free 197 12 185 94% 36 149 81%

F - Women Choice Free 202 0 202 100% 26 173 86%G - Men Choice Subsidy 197 55 142 72% 27 115 81%

H - Women / Subsidy 203 63 140 69% 29 111 79%Overall 799 130 669 118 548

If Yes, Type Chosen

+5%

-2%

Stove Orders

Any Stove

Order

Ordered a chimney stove out of those

who ordered a stove

Ordered a chimney stove out of those offered a

stove

Free Stove Condition -0.0597*** -0.0724* -0.121*** (standard error) (0.0175) (0.0375) (0.0391) sample size 397 382 394

Subsidized Stove Condition 0.0259 0.0140 0.0502 (standard error) (0.0457) (0.0450) (0.0482) sample size 398 282 398

p-value for equality of coefficients on 'male' between free and subsidized cases

0.0400** 0.0764* 0.00384***

Stove Purchase

Any Stove

Purchase

Purchased a chimney stove out of those

initially accepting a stove

Purchased a chimney stove out of those

offered a stove

Refused to Purchase of those who

Ordered one

Free Stove Condition -0.0101 -0.108** -0.0803 -0.0496 (standard error) (0.0477) (0.0492) (0.0516) (0.0462) sample size 397 275 394 397

Subsidized Stove -0.0276 -0.00802 -0.00350 0.0535 (standard error) (0.0455) (0.0746) (0.0399) (0.0505) sample size 398 111 398 398

p-value for equality of coefficients on 'male' between free and subsidy

0.652 0.231 0.0809* 0.0041***

Policy Implications • Can’t really market to either men or

women. • Solution: bundle an attribute that men

want with the stove. • Need financing solutions for a liquidity

or credit constraint, or risk aversion about a new/unknown product

• Clever marketing and persuasion techniques have limited effects when households can evaluate the technology for themselves

Growth Week 2011: Ideas for Growth Session 9 - Climate Change, Environment, Natural Resources

Business

oil commercial oil reserves

oil reservesgdp

oil producers

tullow oil

harnessing oil revenues

windfall ghana

fourchannels oil revenue

aspects of oil production