## _wp08101

## Source details

**Canonical URL:** [_wp08101](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08101.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08101.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08101.pdf.json)

---

### Introduction: volatility and pro-cyclicality of aid
- Aid flows to poor, aid dependent countries are highly volatile and pro-cyclical.
- For a group of 72 aid-dependent countries (period 1975-1997):
  - The average measure of relative volatility of aid flows is close to 4, implying aid flows are four times more volatile than fiscal revenues (both normalized by GDP).
  - For countries with average ratio of aid to revenue greater than 10 percent, relative aid volatility is 4.96.
  - For countries where average ratio of aid to revenue is above 50 percent, relative aid volatility is 7.4.
- The average correlation coefficient between detrended aid and revenue series is positive for all groups, with the correlation increasing with the ratio of aid to government revenues.
- Contextual statistics:
  - Since the mid 1990s real resource transfers on debt (as a percentage of GDP) have been close to zero or negative, with outright grants gaining in relative importance.
  - Aid accounts for 12.5 percent of disposable income on average in Africa.

### Primary commodity exports and price volatility
- Many aid-dependent countries depend on one or two primary commodities; commodity prices are highly volatile.
- For 17 HIPC’s (average shares 1997–99):
  - 14 of the 17 countries have more than 20 percent of export revenues coming from a single commodity.
  - 15 of the 17 countries earn more than 30 percent of their export revenue from just three commodities.
  - For 16 out of the 17 countries, export revenues from the top 3 commodities constitute more than 5 percent of total GDP.
- Instability indices (1957–1999) — methodology:
  - Instability index = percentage standard deviation of the natural logarithm of real commodity price (deflated by The Economist’s Manufactured Unit Value Index) around the Hodrick-Prescott filtered trend.
- Empirical volatility findings:
  - Sugar: average annual percentage deviation of around 40 percent.
  - Average annual percentage deviation for all primary commodity prices: around 20 percent.
  - 10 out of 14 commodities have volatility above 15 percent.
- Implication:
  - Given average share of the leading primary commodity in total exports is about 40 percent for countries in the sample, commodity price volatility would imply an annual export-revenue volatility of 4 to 16 percent.

### Case study: Cotton (Burkina Faso)
- Burkina Faso characteristics:
  - Cotton exports account for 65 percent of total merchandise exports and 9 percent of total GDP.
  - Burkina Faso is a price taker in the world cotton market.
- Empirical relationships:
  - Aid flows are about five times more volatile than government revenues.
  - Rank correlation between aid and government revenue is positive.
  - Government revenue is positively correlated with world cotton prices.
  - Aid flows to Burkina Faso are positively correlated with world cotton price.
  - For five cotton exporters (Benin, Burkina Faso, Burundi, Chad, Mali) average correlation coefficient between aid and world cotton price is 0.3.
- Historical response of aid to GDP declines (five cotton exporters, 1981–2004):
  - Out of 49 cases of GDP decline by 5 percent or more, only 17 were accompanied by an increase in aid flows of 5 percent or more.
  - Out of 59 cases of GDP reduction by 1 percent or more, only 29 were accompanied by an increase in aid flows; the remaining 30 cases saw a reduction in aid flows.
  - Conclusion: a decline in GDP was almost as likely to be accompanied by an increase in aid as by a decline in aid.

### Model overview and key calibration
- Model summary:
  - Three-sector, dynamic general equilibrium SOE with no access to international capital markets.
  - Shocks: productivity, aid, and terms of trade.
  - Aid flows assumed more volatile than output and mildly pro-cyclical.
  - Capital is sector-specific; investment uses imported goods; quadratic capital adjustment costs included.
- Key numeric summaries (data and calibration):
  - Relative Volatility of AID-Revenue: 4.92
  - Correlation Coefficient for Revenue-AID: 0.12
  - Correlation Coefficient for Revenue-Price: 0.33
  - Correlation Coefficient for AID-Price: 0.18
  - Elasticity of substitution between exportable and non-tradable consumption (1/(1+μ)): 1.279
  - Share of exports in tradable consumption (ς): 0.15
  - Risk aversion coefficient (σ) used in simulations: 2.61 and 5
  - Capital shares: α = 0.3 (exportable), η = 0.4 (non-tradable)
  - Elasticity of substitution between sectoral capital (1/(1+υ)): -10
  - Rate of depreciation (δ): 0.09
  - Steady state investment rate targeted: 18 percent of GDP
  - Productivity shock: standard deviation 4 percent; first-order auto-correlation coefficient 0.4
  - Volatility of aid shocks: set to be five times the volatility of productivity shocks
  - Correlation coefficient between productivity and aid shocks: 0.4
  - Terms of trade shocks: standard deviation 11 percent; persistence 0.4
  - Steady state aid-to-GDP ratio for benchmark simulations: 12.5 percent
  - Comparative static aid-to-GDP ratios evaluated: 0, 1/10, 1/4, 1/2

### Steady state and comparative-static implications
- Steady-state properties:
  - Steady state return on capital net of depreciation equals discount factor β.
  - Non-tradable consumption equals output (C_N = Y_N).
  - Investment steady state: i = δ k.
- Comparative-static findings for a permanent increase in aid:
  - Aid increases the relative price of the non-tradable good (real appreciation).
  - Aid lowers the share of labor and capital employed in the exportable sector.
  - Aid lowers the share of exportable output in total GDP.
  - Aid lowers investment as a share of total income (GDP + exogenous aid flows).
  - Aid increases the share of tradable goods in total consumption.
  - A permanent increase in aid raises steady-state consumption permanently but does not change steady-state saving or investment (steady-state capital stock determined by β and δ, independent of aid).
  - Additional aid tends to be consumed rather than saved.

### Indexed aid: designs, triggers, and operational rules
- Indexation objective: provide insurance against exogenous falls in income, reduce macroeconomic volatility, and improve welfare.
- Types of indexing described:
  - Terms of Trade (TOT) indexing — advantages/limitations:
    - Advantages: minimal moral hazard (donor observes world prices), timely (world prices known without lag).
    - Limitations: divergence between domestic and world prices due to transport/grade; misses quantity variation; applicability limited to countries with commodity concentration.
    - Permanent vs temporary shocks: indexation best applied to temporary fluctuations; not recommended for secular declines.
    - Cashin and Pattillo (2000) finding: out of 42 countries, 11 have permanent TOT shocks; among a 17-country sample, four (Côte d’Ivoire, Kenya, Uganda and Zambia) have non-dissipating/permanent TOT shocks and are not suitable candidates for TOT-based insurance.
  - Export-indexed aid:
    - Mechanics: aid_t = aid × [1 + λ_x × g_EXt], with g_EXt using max{0, EX_t/EX − 0.03}.
    - Cut-off: value of exports falling/rising by more than 3 percent triggers adjustments; one standard deviation TOT band roughly corresponds to a direct change of 0.5 percent in GDP and a 3 percent change in export revenues.
    - Parameter: xλ < 1 determines degree of indexation.
  - GDP-indexed aid:
    - Mechanics: aid_t = aid × [1 + λ_y × g_Yt], with g_Yt = max{0, Y_t/Y − 0.005}.
    - Cut-off: pre-specified level of 0.5 percent of GDP.
- Operational specification for TOT-indexing:
  - Use percentage deviation from a moving average trend (last four or five years) to classify periods.
  - If shock < 0.5 s.d.: period is normal (N); aid flows kept at mean level.
  - If shock is between 0.5 s.d. and 1 s.d.: adverse (A) if negative, propitious (P) if positive; aid is increased (decreased) by 5 percent.
  - If shock > 1 s.d.: A* (or P*); aid increased (decreased) by 10 percent.

### Welfare measurement and methodological notes
- Welfare metric: CES consumption following Lucas (1987).
- Welfare cost of consumption volatility: minimum percentage increase in consumption in every period making agent indifferent between stochastic consumption stream and deterministic one with same mean.
- Approximate closed form under CRRA and i.i.d. shocks:
  - Welfare cost ≈ (1/2) × σ × VC  (σ is coefficient of relative risk aversion; VC is variance of CES consumption aggregator around HP filtered trend).
- Caveat: welfare calculations focus solely on consumption volatility (ignore changes in consumption level due to precautionary savings and capital accumulation).

### Quantitative welfare results and comparisons
- Amplification by capital adjustment costs:
  - Specification 1 (no TOT shocks, no capital adjustment costs):
    - S.D. of consumption: 0.058
    - Welfare cost of aid volatility: 0.0059
  - Specification 2 (TOT shocks, no capital adjustment costs):
    - S.D. of consumption: 0.074
    - Welfare cost of aid volatility: 0.009
  - Specification 3 (benchmark: TOT shocks and capital adjustment costs):
    - S.D. of consumption: 0.13
    - Welfare cost of aid volatility: 0.04
- Welfare gains from indexation (benchmark specification comparisons):
  - Standard Deviation of Consumption:
    - Benchmark: 0.13
    - TOT indexed: 0.036
    - GDP indexed: 0.026
    - Export indexed: 0.014
  - Welfare Cost (compensating variation in CES consumption):
    - Benchmark: 0.045
    - TOT indexed: 0.003
    - GDP indexed: 0.0017
    - Export indexed: 0.0005
  - Correlation between Aid and Output:
    - Benchmark: 0.05
    - TOT indexed: -0.44
    - GDP indexed: -0.81
    - Export indexed: -0.799
  - Standard Deviation of Investment:
    - Benchmark: 0.376
    - TOT indexed: 0.24
    - GDP indexed: 0.144
    - Export indexed: 0.12
  - Standard Deviation of GDP:
    - Benchmark: 0.044
    - TOT indexed: 0.028
    - GDP indexed: 0.035
    - Export indexed: 0.022
  - Capital to GDP ratio:
    - Benchmark: 2.40
    - TOT indexed: 1.34
    - GDP indexed: 2.20
    - Export indexed: 1.24
  - Investment to GDP ratio:
    - Benchmark: 0.21
    - TOT indexed: 0.12
    - GDP indexed: 0.198
    - Export indexed: 0.112
  - Consumption to GDP ratio:
    - Benchmark: 0.37
    - TOT indexed: 0.412
    - GDP indexed: 0.414
    - Export indexed: 0.436
  - Exportable Output to GDP ratio:
    - Benchmark: 0.54
    - TOT indexed: 0.39
    - GDP indexed: 0.44
    - Export indexed: 0.34
  - P_n (relative price of non-tradables):
    - Benchmark: 0.84
    - TOT indexed: 1.08
    - GDP indexed: 0.95
    - Export indexed: 1.16
- Interpretation:
  - Indexing aid to terms of trade reduces consumption volatility materially and lowers welfare cost by about 4 percent of permanent consumption relative to stochastic aid.
  - GDP and export indexing provide slightly greater reductions in consumption volatility and welfare cost than TOT indexing because they more directly cushion exogenous falls in income or export revenues.
  - Across specifications, indexation reduces volatility of investment and GDP, lowers precautionary capital accumulation, and raises long-run consumption to GDP ratios.

### Sensitivity analysis highlights
- Lower risk aversion (σ = 2.61) — qualitative conclusions hold; absolute welfare losses and absolute gains from indexation are smaller:
  - S.D. of consumption (Benchmark): 0.12; Welfare Cost (Benchmark): 0.0197.
  - S.D. of consumption (TOT indexed): 0.032; Welfare Cost (TOT indexed): 0.0013.
- Higher average aid (doubling aid) — indexing still yields substantial welfare gains and can increase absolute gains because permanent consumption levels rise:
  - S.D. of consumption (Benchmark): 0.123; Welfare Cost (Benchmark): 0.045.
  - S.D. of consumption (TOT indexed): 0.04; Welfare Cost (TOT indexed): 0.0035.

### Policy-relevant findings and recommendations
- Volatility and pro-cyclicality of aid, together with volatile terms of trade and lack of export diversification, are crucial sources of macroeconomic instability in many developing countries.
- Capital adjustment costs and terms of trade shocks substantially amplify welfare losses from aid volatility; ignoring capital adjustment costs understates losses.
- Rearranging aid flows to insure recipient countries against exogenous terms of trade shocks materially improves welfare by lowering consumption volatility.
  - For reasonable risk aversion values, gains in welfare are between 2 to 4 percent of permanent consumption.
  - Alternatively, donors can reduce average aid flows by 1.5 to 3 percent without lowering recipient welfare by indexing aid to exogenous shocks.
- Design considerations for indexation:
  - Index temporary fluctuations (not secular declines).
  - Prefer indices with low moral hazard and timely availability (world prices for TOT indexing).
  - Tailor index choice to countries with high commodity concentration and volatile export revenues.
- Additional considerations and caveats:
  - Model omits conditionality, fungibility, absorption constraints, and productivity spillovers; computed welfare gains are likely a lower bound.
  - As donors consider increasing aid to meet development goals, structuring aid to promote growth and reduce volatility is crucial.

*Source: _wp08101 (IMF Working Paper content provided).*

### 1. Dynamic behavior of Aid..............................................................................................

### 1. Dynamic behavior of Aid

### Introduction: volatility and pro-cyclicality of aid
- Aid flows to poor, aid dependent countries are highly volatile and pro-cyclical.
- For a group of 72 aid-dependent countries (period 1975-1997):
  - The average measure of relative volatility of aid flows is close to 4, implying aid flows are four times more volatile than fiscal revenues (both normalized by GDP).
  - For countries with average ratio of aid to revenue greater than 10 percent, relative aid volatility is 4.96.
  - For countries where average ratio of aid to revenue is above 50 percent, relative aid volatility is 7.4.
- The average correlation coefficient between detrended aid and revenue series is positive for all groups, with the correlation increasing with the ratio of aid to government revenues.
- Many aid-dependent countries have little or no access to international capital markets and are dependent on one or two primary commodities with highly volatile prices; thus volatility and pro-cyclicality of aid have important implications for growth and welfare.
- Model overview (high level):
  - A three sector, dynamic general equilibrium model of a small open economy (SOE) with no access to international capital markets.
  - The SOE faces productivity and terms of trade shocks and receives an exogenous stochastic transfer of tradable goods (aid) every period.
  - Aid flows are assumed to be more volatile than output and mildly pro-cyclical.
  - The SOE cannot borrow abroad; it can invest in domestic capital stock.
  - Representative household engages in precautionary saving via capital accumulation.
  - Capital is imperfect for consumption smoothing due to diminishing marginal product; building capital to buffer shocks yields declining marginal returns.
  - Presence of capital adjustment costs increases the welfare cost of aid volatility substantially (about four times).
- Key quantitative welfare results from the model:
  - Indexing aid flows to temporary terms of trade shocks, while keeping average aid constant, significantly reduces volatility of disposable income and consumption, lowering cost of business cycle fluctuations.
  - Reduction in the cost of business cycle fluctuations due to indexed aid is about 4 percent of permanent consumption relative to the stochastic aid benchmark.
  - Alternatively, the same level of welfare can be provided to the recipient country with 3 percent less aid by indexing aid to terms of trade shocks.
- Contextual points:
  - Since the mid 1990s real resource transfers on debt (as a percentage of GDP) have been close to zero or negative, with outright grants gaining in relative importance.
  - Aid accounts for 12.5 percent of disposable income on average in Africa.
  - Pro-cyclical aid can force countries to raise taxes or curb expenditures during recessions, dampening growth.
  - Empirical evidence suggests aid flows are often driven more by political and strategic considerations than by recipient economic prudence or policy performance.

### Primary commodity exports and price volatility
- Many aid-dependent countries are heavily dependent on one or two primary commodities for export revenues, and those commodity prices are highly volatile.
- For 17 Highly Indebted Poor Countries (HIPC’s) (average shares 1997–99):
  - 14 of the 17 countries have more than 20 percent of export revenues coming from a single commodity.
  - 15 of the 17 countries earn more than 30 percent of their export revenue from just three commodities.
  - For 16 out of the 17 countries, export revenues from the top 3 commodities constitute more than 5 percent of total GDP.
- Instability indices (1957–1999) — methodology:
  - Instability index = percentage standard deviation of the natural logarithm of real commodity price (deflated by The Economist’s Manufactured Unit Value Index) around the Hodrick-Prescott filtered trend.
  - These indices measure volatility of primary commodity export prices relative to volatility of manufactured goods imported.
- Empirical volatility findings:
  - Sugar has the highest volatility (average annual percentage deviation of around 40 percent), followed by cashew nuts.
  - The average annual percentage deviation for all primary commodity prices is around 20 percent.
  - 10 out of 14 commodities have volatility above 15 percent.
- Implication:
  - Given the average share of the leading primary commodity in total exports is about 40 percent for countries in the sample, commodity price volatility would imply an annual export-revenue volatility of 4 to 16 percent (depending on the commodity).

### Case study: Cotton (Burkina Faso)
- Burkina Faso characteristics:
  - Small open economy dependent on primary commodity exports.
  - Cotton exports account for 65 percent of total merchandise exports and 9 percent of total GDP.
  - Burkina Faso is a price taker in the world cotton market.
- Empirical relationships (Burkina Faso and cotton exporters):
  - Aid flows are about five times more volatile than government revenues (consistent with broader sample).
  - Rank correlation between aid and government revenue is positive.
  - Government revenue is positively correlated with world cotton prices.
  - Aid flows to Burkina Faso are positively correlated with world cotton price (contrary to an insurance role); for five cotton exporters with data (Benin, Burkina Faso, Burundi, Chad, Mali) average correlation coefficient between aid and world cotton price is 0.3.
- Historical response of aid to GDP declines (five cotton exporters, 1981–2004):
  - Out of 49 cases of GDP decline by 5 percent or more, only 17 were accompanied by an increase in aid flows of 5 percent or more.
  - Out of 59 cases of GDP reduction by 1 percent or more, only 29 were accompanied by an increase in aid flows; the remaining 30 cases saw a reduction in aid flows.
  - Conclusion: a decline in GDP was almost as likely to be accompanied by an increase in aid as by a decline in aid.

### Implications for policy and aid design
- The combination of aid volatility, volatile terms of trade, and lack of export diversification is a crucial source of macroeconomic instability in many developing countries.
- Because most developing countries cannot control prices of their main exports and imports, it is useful to study aid volatility jointly with terms of trade volatility.
- Potential policy response highlighted by the analysis:
  - Indexing aid flows to temporary terms of trade shocks can provide insurance, reduce disposable income and consumption volatility, and yield material welfare gains while keeping average aid constant.
- Model-based policy quantification:
  - Indexation to terms of trade shocks can reduce the welfare cost of business cycle fluctuations by about 4 percent of permanent consumption compared to stochastic aid.
  - The same welfare level could be achieved with 3 percent less aid if aid were indexed to terms of trade shocks.

*Source: _wp08101 - 1. Dynamic behavior of Aid (IMF Working Paper content provided).*

### Section 3 describes the solution procedure. Section 4 presents some comparative static

### _wp08101 - Section 3 describes the solution procedure. Section 4 presents some comparative static

### Model overview
- Benchmark: SOE real business cycle model based on Mendoza (1995).
- Households consume exportable, importable, and non-tradable goods; labor normalized to one and freely substitutable across sectors; capital is sector-specific and the only investable asset.
- Three sources of uncertainty: productivity shocks, aid shocks, and terms of trade shocks.
- Imports used as numeraire good.

### Key numeric summaries (data and calibration)
- Volatility and correlations:
  - Relative Volatility of AID-Revenue: 4.92
  - Correlation Coefficient for Revenue-AID: 0.12
  - Correlation Coefficient for Revenue-Price: 0.33
  - Correlation Coefficient for AID-Price: 0.18
- Calibration parameters:
  - Elasticity of substitution between exportable and non-tradable consumption (1/(1+μ)): 1.279
  - Share of exports in tradable consumption (ς): 0.15
  - Risk aversion coefficient (σ) used in simulations: 2.61 and 5
  - Capital shares: α = 0.3 (exportable), η = 0.4 (non-tradable)
  - Elasticity of substitution between sectoral capital (1/(1+υ)): -10
  - Rate of depreciation (δ): 0.09
  - Steady state investment rate targeted: 18 percent of GDP
  - Productivity shock: standard deviation 4 percent; first-order auto-correlation coefficient 0.4 (implying "60 percent of any deviation of productivity from its mean level dissipates in each period")
  - Volatility of aid shocks: set to be five times the volatility of productivity shocks
  - Correlation coefficient between productivity and aid shocks: 0.4
  - Terms of trade shocks: standard deviation 11 percent; persistence 0.4
  - Steady state aid-to-GDP ratio for benchmark simulations: 12.5 percent
- Comparative static aid-to-GDP ratios evaluated: 0, 1/10, 1/4, 1/2

### Household (structure and constraints)
- Lifetime expected utility: Σ_{t=0}^∞ β^t C_t^{1-σ}/(1-σ).
- Aggregate consumption Ct is CES of non-tradable and tradable composites; tradable composite combines exportable and importable goods with share ς.
- Household income components: wages (w_t), rental income (r_t), stochastic transfer of tradable goods (AID_t), and returns from capital.
- Resource constraint (expressed in tradables and non-tradables prices): EX p_t C_EX + N p_t C_N + IM C_IM + w r k + AID_t = ... (full expression in source).

### Firms and production
- Cobb-Douglas production in each sector:
  - Y_Nt = A_t K_Nt^{η} L_Nt^{1-η}
  - Y_EXt = A_t K_EXt^{α} L_EXt^{1-α}
- Labor perfectly mobile across sectors: L_Nt + L_EXt = 1
- Sector-specific capital with transformation curve κ(k_EX, k_N) specified (see source); perfect homogenous capital special case 1/υ = -1.
- Quadratic capital adjustment costs; capital accumulation:
  - k_{t+1} = (1-δ)k_t + i_t - φ/2 * (k_{t+1}/k_t - 1)^2 (full functional form and parameters in source)
- Investment uses imported goods; investment identity links export revenue, consumption, aid and investment (see source).

### First-order conditions and equilibrium conditions
- Factor prices equal marginal products across sectors (wage and rental equalities given in equations III-12 and III-13).
- Intertemporal Euler equation linking consumption, capital returns, terms of trade, and CES price index (equations III-14 and III-15 in source).
- Goods and input market clearing conditions provided (equations III-10 through III-25 in source).

### Steady state results and implications
- Steady state return on capital net of depreciation equals discount factor β (equation III-18).
- Key steady-state relationships:
  - Marginal rates of substitution equated to relative prices (e.g., C_IM/C_EX and terms of trade; equation III-19).
  - Non-tradable consumption equals output (C_N = Y_N).
  - Investment steady state: i = δ k (equation III-23).
- Comparative-static findings (sensitivity to permanent increases in aid):
  - Aid increases the relative price of the non-tradable good (real appreciation).
  - Aid lowers the share of labor and capital employed in the exportable sector.
  - Aid lowers the share of exportable output in total GDP.
  - Aid lowers investment as a share of total income (defined as GDP + exogenous aid flows).
  - Aid increases the share of tradable goods in total consumption.
  - A permanent increase in aid raises steady-state consumption permanently but does not change steady-state saving or investment (steady-state capital stock determined by β and δ, independent of aid).
  - Additional aid tends to be consumed rather than saved.

### Sensitivity analysis (comparative statics)
- Comparative static checks across variations in capital intensity and capital substitution elasticity show:
  - Decline in exportable output with increased aid holds across all specifications.
  - Increase in the relative price of non-tradable goods holds as long as capital is sector-specific.
  - If capital were perfectly substitutable, no real appreciation would be necessary because capital could shift costlessly across sectors.
  - The magnitude of decline in capital share in exportable sector depends on elasticity of substitution of capital: the more sector-specific the capital (lower substitutability), the smaller the decline in the exportable sector capital share.
  - Labor share in exportable sector declines in all models; magnitude depends on capital substitution elasticity.
  - Ratio of investment to total income declines in all models (additional aid is consumed rather than invested).

### Dynamics and solution procedure
- The stochastic dynamic problem rewritten as Bellman equation (see V-1 in source).
- Shocks: productivity, aid, and terms of trade; productivity shocks affect both sectors equally.
- Numerical solution: value function iteration and transition probability iterations on discrete grids used because indexing of aid introduces kinks in the budget constraint making linear approximation inappropriate.
- Exact-solution procedure based on value function and transition probability iterations used to obtain policy and value functions.

### Indexed aid: objectives, methods, and trade-offs
- Objective of indexation: provide a cushion against exogenous falls in income (reduce macroeconomic volatility and improve welfare).
- Three methods of indexing aid are described (detailed methods follow in source); pros and cons discussed at length.
- Terms of trade (TOT) indexing:
  - Advantages:
    - Few developing countries have market power over world prices → less moral hazard from recipient manipulation.
    - World prices are known without lag → more timely moderation of aid flows versus schemes based on GDP or exports which have lags and revisions.
  - Limitations:
    - Transport costs and grade differentials can cause divergence between domestic export/import prices and international prices.
    - Quantity variation in exports can be as important as price variation; TOT-only schemes miss this.
    - Applicability limited to countries with high commodity concentration in exports and imports.
  - Trade-off in choosing index variable: extent of insurance vs potential moral hazard.
  - Permanent vs temporary TOT shocks:
    - Indexation best applied to temporary fluctuations; not advisable to shield against secular declines in export prices.
    - Cashin and Pattillo (2000) findings cited: out of 42 countries, 11 have permanent TOT shocks; among the 17-country sample referenced in the paper, four—Côte d’Ivoire, Kenya, Uganda and Zambia—have non-dissipating or permanent TOT shocks and thus are not suitable candidates for TOT-based insurance schemes; the remaining 13 countries could potentially benefit.
- Operational specification of the TOT-indexing rule (implementation thresholds):
  - Period classification based on magnitude of terms of trade shock measured as percentage deviation from a moving average trend (last four or five years).
  - If shock is less than 0.5 standard deviation (s.d.), period is normal (N) and aid flows kept at mean level.
  - If shock is between 0.5 s.d. and 1 s.d., period is adverse (A) if negative (propitious P if positive) and aid flows are increased (decreased) by 5 percent.
  - If shock exceeds 1 s.d., period is A* (or P* if positive) and aid flows are increased (decreased) by 10 percent.

*Italic: Source — _wp08101 - Section 3 describes the solution procedure. Section 4 presents some comparative static*

### 1.1 if 1 .  .

### _wp08101 - 1.1 if 1 .  .

### Extracted content
- 1.1 if 1 .  .
- 1.05 if  -0.5..-1..
-  if  0.5  .  .0.5  .  .
- 0.95 if 1 .0.5  . 

### Preserved numeric expressions
- "1.1"
- "1.05"
- "-0.5"
- "-1"
- "0.5"
- "0.5"
- "0.95"
- "1"
- "0.5"

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08101.pdf*

### 0.9 if 1 .  .

### _wp08101 - 0.9 if 1 .  .

### Export-Indexed Aid
- Concept: Export-indexed aid behaves like a combination of a bond and an insurance contract with payments linked to the value of exportable good output in terms of importable goods.
- Trigger/cut-offs:
  - A one standard deviation shock in the terms of trade band roughly corresponds to a direct change of 0.5 percent in the GDP.
  - Translated into export revenues, a one standard deviation decline in terms of trade is roughly equal to a direct loss of 3 percent.
  - Cut-off points used: 0.5 percent (GDP) and 3 percent (export indexed aid).
- Mechanics:
  - Aid flows to the recipient country are adjusted upwards (downwards) proportionately when the value of exports falls (rises) by more than 3 percent below (above) the steady state level.
  - Aid flows specification (as given): aid_t = aid × [1 + λ_x × g_EXt] (V-3)
  - Exports gap definition (g_EXt) uses max{0, EX_t/EX − 0.03} structure and steady state EX_Y and EX_t variables (V-4).
- Parameter:
  - xλ < 1 is the adjustment coefficient determining the degree of aid indexation.

### GDP-Indexed Aid
- Concept: Aid flows adjusted when GDP falls below (rises above) its steady state level by more than a pre-specified level.
- Trigger/cut-off:
  - Pre-specified level used: 0.5 percent of GDP.
- Mechanics:
  - Aid flows specification (as given): aid_t = aid × [1 + λ_y × g_Yt] (V-5)
  - g_Yt = max{0, Y_t/Y − 0.005} where Y is steady state GDP and Y_t is current GDP (V-6).

### Welfare Costs (Methodology)
- Welfare metric: CES consumption following Lucas (1987).
- Definition: Welfare cost of consumption volatility is the minimum percentage increase in consumption in every period needed to render the agent indifferent between the stochastic consumption stream and a deterministic one with the same mean.
- Business cycle focus: Uses cyclical component of Hodrick-Prescott (HP) filtered series.
- Approximate closed form under CRRA and i.i.d. shocks:
  - Welfare cost ≈ (1/2) × σ × VC  (presented compactly as (1/2) WVC σ = where σ is coefficient of relative risk aversion and VC is the variance of the CES consumption aggregator around the HP filtered trend).
- Caveat: Welfare calculations focus solely on consumption volatility (ignore changes in consumption level due to precautionary savings and capital accumulation).

### Results — Model Specifications and Benchmarking
- Three specifications compared:
  - Specification 1: No terms of trade shocks and no capital adjustment costs.
  - Specification 2: Terms of trade shocks but no capital adjustment cost.
  - Specification 3 (benchmark): Both terms of trade shocks and capital adjustment costs.
- Key finding: Capital adjustment costs materially amplify welfare losses from aid volatility by making domestic capital a less effective consumption-smoothing device.
- Table 6 (Welfare Cost Under Alternative Model Specifications):
  - Specification 1:
    - S.D. of consumption: 0.058
    - Welfare cost of aid volatility: 0.0059
  - Specification 2:
    - S.D. of consumption: 0.074
    - Welfare cost of aid volatility: 0.009
  - Specification 3:
    - S.D. of consumption: 0.13
    - Welfare cost of aid volatility: 0.04

### Results — Welfare Gains from Aid Indexation (Benchmark Specification)
- Comparing benchmark volatile aid to three indexed schemes (TOT, GDP, Export):
  - Standard Deviation of Consumption:
    - Benchmark: 0.13
    - TOT indexed: 0.036
    - GDP indexed: 0.026
    - Export indexed: 0.014
  - Welfare Cost (compensating variation in CES consumption):
    - Benchmark: 0.045
    - TOT indexed: 0.003
    - GDP indexed: 0.0017
    - Export indexed: 0.0005
  - Correlation between Aid and Output:
    - Benchmark: 0.05
    - TOT indexed: -0.44
    - GDP indexed: -0.81
    - Export indexed: -0.799
  - Standard Deviation of Investment:
    - Benchmark: 0.376
    - TOT indexed: 0.24
    - GDP indexed: 0.144
    - Export indexed: 0.12
  - Standard Deviation of GDP:
    - Benchmark: 0.044
    - TOT indexed: 0.028
    - GDP indexed: 0.035
    - Export indexed: 0.022
  - Capital to GDP ratio:
    - Benchmark: 2.40
    - TOT indexed: 1.34
    - GDP indexed: 2.20
    - Export indexed: 1.24
  - Investment to GDP ratio:
    - Benchmark: 0.21
    - TOT indexed: 0.12
    - GDP indexed: 0.198
    - Export indexed: 0.112
  - Consumption to GDP ratio:
    - Benchmark: 0.37
    - TOT indexed: 0.412
    - GDP indexed: 0.414
    - Export indexed: 0.436
  - Exportable Output to GDP ratio:
    - Benchmark: 0.54
    - TOT indexed: 0.39
    - GDP indexed: 0.44
    - Export indexed: 0.34
  - P_n (relative price of non-tradables):
    - Benchmark: 0.84
    - TOT indexed: 1.08
    - GDP indexed: 0.95
    - Export indexed: 1.16
- Interpretation:
  - Indexing aid to terms of trade reduces consumption volatility by about ten percent and welfare cost by about four percent relative to benchmark.
  - GDP and export indexing provide slightly greater reductions in consumption volatility and welfare cost than terms of trade indexing because they more directly cushion exogenous falls in income or export revenues.
  - Overall gains in welfare are of the same order of magnitude across TOT, GDP, and export indexation.

### Results — Mechanisms and General Equilibrium Effects
- Aid indexation stabilizes disposable tradable income of the small open economy (SOE), reducing volatility of investment and hence output.
- Insurance via indexation reduces precautionary capital accumulation -> smaller long-run capital to GDP ratio and lower long-run investment to GDP ratio.
- Long-run consumption to GDP ratio increases with aid indexation (more output devoted to consumption than saving).
- Share of exportable goods in GDP falls with aid indexation due to:
  - Reduced demand for tradable goods (investment uses tradable goods),
  - Higher relative price of non-tradable goods,
  - Reallocation of labor and capital toward non-tradable sector.

### Sensitivity Analysis
- Lower risk aversion (σ = 2.61) — Table 8:
  - Standard Deviation of Consumption:
    - Benchmark: 0.12
    - TOT indexed: 0.032
    - GDP indexed: 0.027
    - Export indexed: 0.016
  - Welfare Cost:
    - Benchmark: 0.0197
    - TOT indexed: 0.0013
    - GDP indexed: 0.001
    - Export indexed: 0.0006
  - Correlation between Aid and Output:
    - Benchmark: 0.041
    - TOT indexed: -0.393
    - GDP indexed: -0.801
    - Export indexed: -0.743
  - S.D. of Investment:
    - Benchmark: 0.373
    - TOT indexed: 0.232
    - GDP indexed: 0.128
    - Export indexed: 0.116
  - Other ratios (Capital to GDP, Investment to GDP, Consumption to GDP, Exportable Output to GDP, P_n) remain the same as in the benchmark table.
  - Interpretation: Lower σ reduces measured welfare losses and reduces absolute gains from indexation, but qualitative conclusions hold.
- Higher average aid (doubling aid) — Table 9:
  - Standard Deviation of Consumption:
    - Benchmark: 0.123
    - TOT indexed: 0.04
    - GDP indexed: 0.031
    - Export indexed: 0.02
  - Welfare Cost:
    - Benchmark: 0.045
    - TOT indexed: 0.0035
    - GDP indexed: 0.003
    - Export indexed: 0.001
  - Correlation between Aid and Output:
    - Benchmark: 0.23
    - TOT indexed: -0.296
    - GDP indexed: -0.650
    - Export indexed: -0.621
  - Standard Deviation of Investment:
    - Benchmark: 0.57
    - TOT indexed: 0.21
    - GDP indexed: 0.137
    - Export indexed: 0.16
  - Standard Deviation of GDP:
    - Benchmark: 0.03
    - TOT indexed: 0.0193
    - GDP indexed: 0.028
    - Export indexed: 0.0281
  - Capital to GDP ratio:
    - Benchmark: 2.50
    - TOT indexed: 1.38
    - GDP indexed: 2.35
    - Export indexed: 1.09
  - Investment to GDP ratio:
    - Benchmark: 0.22
    - TOT indexed: 0.125
    - GDP indexed: 0.212
    - Export indexed: 0.0986
  - Consumption to GDP ratio:
    - Benchmark: 0.45
    - TOT indexed: 0.46
    - GDP indexed: 0.458
    - Export indexed: 0.469
  - Exportable Output to GDP ratio:
    - Benchmark: 0.36
    - TOT indexed: 0.266
    - GDP indexed: 0.34
    - Export indexed: 0.16
  - P_n:
    - Benchmark: 1.03
    - TOT indexed: 1.27
    - GDP indexed: 1.06
    - Export indexed: 1.54
  - Interpretation: Doubling aid leaves volatility measures relatively unchanged; indexing still yields substantial welfare gains and potentially larger absolute gains because permanent consumption levels rise.

### Benchmark Parameter Values (selected)
- Time Preference β: 0.95
- Share of tradable Consumption π: 0.50
- Share of Exportable good in tradable good consumption ς: 0.15
- Elasticity of substitution between tradable and non-tradable consumption (1/(1+μ)): 1.279
- Depreciation Rate δ: 0.09
- Capital Share in Exportable Sector α: 0.30
- Capital Share in Non-tradable sector η: 0.40
- Risk Aversion Coefficient σ: 5.00
- Elasticity of substitution between tradable and non-tradable capital (1/(1+ν)): -10
- Productivity TN_AA = 1
- Capital Adjustment Cost φ: 0.028

### Policy-Relevant Findings and Recommendations
- Volatility and pro-cyclicality of aid produce substantial welfare losses in aid-dependent economies lacking access to international capital markets.
- Welfare costs of aid volatility are significantly larger when capital adjustment costs and terms of trade shocks are present; ignoring capital adjustment costs leads to substantial underestimation of losses.
- Rearranging aid flows to insure recipient countries against exogenous terms of trade shocks materially improves welfare by lowering consumption volatility.
  - For reasonable values of risk aversion, gains in welfare are between 2 to 4 percent of permanent consumption.
  - Alternatively, donors can reduce average aid flows by 1.5 to 3 percent without lowering recipient welfare by indexing aid to exogenous shocks.
- Additional considerations:
  - Model simplifications (omitted conditionality, fungibility, absorption constraints, productivity spillovers) likely imply that computed welfare gains are a lower bound.
  - As donor countries consider increasing aid flows to meet Millennium Development Goals, designing aid flow structures that promote growth and minimize volatility is crucial.

*Source: _wp08101 - 0.9 if 1 .  . (IMF PDF content provided).*

### References

### _wp08101 - References

### Aid, Aid Volatility, and Development Finance
- Arellano, C., Bulir, A., L. Lipshitz, and Timothy, L., 2005, "The Dynamic Implications of Foreign Aid and Its Variability", Journal of Development Economics (forthcoming), Available at http://dx.doi.org/10.1016/j.jdeveco.2008.01.005  
- Atta-Mensah, J., 2004,. "Commodity Linked Bonds: A Potential Means for Less Developed Countries to Raise Foreign Capital", Bank of Canada Working Paper, 2004-20, (June 2004)  
- Berlage, L., Cassimon, D., Dreze, J. and Reding P., 2000, "Prospective Aid and Indebtedness Relief: A Proposal", CORE Discussion Paper, 2000/32  
- Birdsall, N., 2005, "Seven Deadly Sins: Reflections on Donor Failings", Center for Global Development, Working Paper No. 50  
- Bulir, A. and A.J., Hamann, 2003, "Aid Volatility: An Empirical Assessment", IMF Staff Papers, Vol. 50, No. 1, pp. 64-89  
- Bulir, A. and A.J., Hamann, 2006, "Volatility of Development Aid: From the Frying Pan into the Fire", March 2006, IMF Staff Paper, Vol. 54, No. 07/04  
- Burnside, C. and Dollar, D, 2000, "AID, Policies and Growth", American Economic Review, September 2000, Vol. 90, No. 4, pp. 847-68  
- Berube, C., Pallage, S. and M.A. Robe, 2004, "On Potential of Foreign Aid as Insurance", CIRPEE Working Paper 04-40  
- Gemmell, N. and M. McGillivray, 1998, "Aid and Tax Instability and the Government Budget Constraints in Developing Countries", CREDIT Working Paper No. 98/1 (Nottingham, England: University of Nottingham Center for Research in Economic Development and International Trade)  
- Sevensson, J., 2000, "When is Foreign Aid Policy Credible? Aid Dependence and Conditionality", Journal of Development Economics, Vol. 61, pp. 61-84  
- World Bank, 2005, "Managing the Debt Risk of Exogenous Shocks in Low Income Countries", Washington D.C., World Bank

### Commodity Prices, Terms of Trade, and Related Instruments
- Atta-Mensah, J., 2004,. "Commodity Linked Bonds: A Potential Means for Less Developed Countries to Raise Foreign Capital", Bank of Canada Working Paper, 2004-20, (June 2004)  
- Cashin, P., Liang, H. and McDermott, C., 2000, "How Persistent Are Shocks to World Commodity Prices?", IMF Staff Papers, Vol. 47, No. 2  
- Cashin, P. and McDermott, C., 2002, "The Long Run Behavior of Commodity Prices: Small Trends and Big Variability", IMF Staff Papers, Vol. 49, No. 2  
- Cuddington, John T., 1992, "Long-Run Trends in 26 Primary Commodity Prices", Journal of Development Economics, Vol. 39, pp. 207-27  
- Gilbert, C.L. and A. Tabova, 2003, "Realignment of debt service obligations and ability to pay in concessional lending", World Bank, HIPC debt sustainability progress report  
- Gilbert, C.L. and A. Tabova, 2004, "Commodity Prices and Debt Sustainability", Discussion Paper 4, Economica, University of Trento  
- Reinhart, Carmen M. and Wickham, P., 1994, "Commodity Prices: Cyclical Weakness or Secular Decline?", IMF Staff Papers, Vol. 41 (June), pp. 175-213

### Debt Instruments, Debt Management, and GDP-Indexed Proposals
- Barro, R., 1995, "Optimal Debt Management", NBER Working Paper 5327  
- Borzentine, E. and P. Mauro, 2004, "The case for GDP-indexed bonds", Economic Policy, (April 2004), pp. 165-216  
- Hausmann, R. and R. Rigobon, 2003, "IDA in UF: On the Benefits of Changing the Currency Denomination of Concessional Lending to Low-Income Countries", World Bank Policy Paper  
- International Monetary Fund, 2004, "Sovereign Debt Structure for Crisis prevention", July 2004  
- Vostroknutova, E., 2005, "Concessional Loans Indexed to GDP Growth: Costs, Benefits and International Risk Sharing", Background Paper, Washington D.C., World Bank  
- Vostroknutova, E. and S. Yi, 2005, "On Costs and Benefits of Indexing Concessional Loans to Low Income Countries on Changes in Real Exchange Rates", Background Paper, Washington D.C., World Bank

### Macroeconomic Volatility, Business Cycles, and Exchange Rate Policy
- Athanasoulis,S. and E. van Wincoop, 1997, "Growth uncertainty and risk-sharing", Economic report series, No. 41, (February 1997)  
- Athanasoulis, S., R.J. Shiller, and E. van Wincoop, 1999, "Macro markets and financial security", Economic policy review, (April 1999), Federal Reserve Bank of New York  
- Cashin, P., Liang, H. and McDermott, C., 2000, "How Persistent Are Shocks to World Commodity Prices?", IMF Staff Papers, Vol. 47, No. 2  
- Hamilton, J.D., 1989, "A New Approach to the Economic Analysis of Non-stationary Time Series and the Business Cycle", Econometrica, Vol. 57, No. 2., pp. 357-384  
- Lucas, R., 1987, "Models of Business Cycles", (New York: Basil Blackwell)  
- Mendoza, E.G., 1991, " Real Business Cycles in a Small Open Economy", American Economic Review, Vol. 81, pp. 797-818  
- Mendoza, E.G., 1995, "The Terms of Trade, The Real Exchange Rate, and Economic Fluctuations", International Economic Review, Vol. 36, No.1, pp. 101-137  
- Mendoza, E.G. and Urine, M., 2000, "Devaluation Risk and The Business-Cycle Implications of Exchange Rate Management", Carnegie-Rochester Conference Series on Public Policy 53, Vol. 53, pp. 239-296.  
- Pallage, S. and M.A. Robe, 2001, "Foreign Aid and Business Cycle", Review of International Economics, Vol. 9, No. 4, pp. 641-672  
- Pallage, S. and M.A. Robe, 2003, "On Welfare Cost of Economic Fluctuations in Developing Countries", International Economic Review, Vol. 44, No. 2, pp. 677-698  
- Ramey, G. and V.A. Ramey, 1994, "Cross-country Evidence on the Link Between Volatility and Growth", NBER Working Paper No. W4959  
- Reinhart, Carmen M. and Vegh, C.A., 1995, "Real Effects of Exchange-Rate Based Stabilization: An Analysis of Competing Theories", in B.S. Bernanke and J. Rotemberg (eds.) NBER Macroeconomics Annual, (MIT Press, Cambridge), 125-174.

### Financial Innovation, Macro Markets, and Risk-Sharing
- Shiller, R.J., 1993, "Macro Markets: Creating Institutions for Managing Society’s Largest Economic Risks", Clarendon Press, Oxford  
- Shiller, R.J., 2003, "The New Financial Order: Risk in the 21st century", Princeton University Press, Princeton, N.J.  
- Shiller, R.J., 2004, "Radical Financial Innovation", Cowles foundation discussion paper No. 1461, April 2004  
- Athanasoulis, S., R.J. Shiller, and E. van Wincoop, 1999, "Macro markets and financial security", Economic policy review, (April 1999), Federal Reserve Bank of New York  
- Pallage, S. and M.A. Robe, 2001, "Foreign Aid and Business Cycle", Review of International Economics, Vol. 9, No. 4, pp. 641-672  
- Berube, C., Pallage, S. and M.A. Robe, 2004, "On Potential of Foreign Aid as Insurance", CIRPEE Working Paper 04-40

*Source: _wp08101 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08101.pdf_
