## _wp06129 — Sections 1–2

## Source details

**Canonical URL:** [_wp06129 — Sections 1–2](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06129.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06129.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06129.pdf.json)

---

### Introduction
- Research question: Whether foreign aid in the form of an increase in capital used only in the nontraded sector can reduce recipient real income (immiserize the country) in the presence of tariff distortions.
- Key claim: The welfare effect of such aid depends on how aid affects the price of the nontraded good and on whether imports and the nontraded good are substitutes or complements in demand.
- Context and critique of prior literature:
  - Past literature (Johnson (1967); Bhagwati, Brecher, and Hatta (1983); Jones (1975)) shows factor accumulation can reduce welfare under distortions.
  - Yano and Nugent (1999) proposed a “transfer paradox” via nontraded sector adjustments; this paper identifies two main errors in their analysis:
    - Their model structure precluded any adjustment in the price of nontraded goods (pN).
    - Their assumptions about how increases in a factor endowment affect sectoral outputs were inconsistent with standard Rybczynski effects.

### The Yano and Nugent model — structure and core critique
- Model specification:
  - Three goods: exports (E), imports (M), nontraded (N).
  - Two factors: labor and capital, mobile domestically but not internationally.
  - Small country: traded-good world prices exogenous: *p1 and *p2.
  - Initial ad-valorem tariff on imports: t.
- Domestic prices (as stated in the source):
  - Exports: p1 = *p1.
  - Imports: p2 = *p2 (1 + t).
- Core critique:
  - Zero-profit conditions imply wage (w) and rental rate (r) are determined by the two traded-goods prices, so the nontraded price pN is determined by costs and therefore is exogenous in this model structure.
  - Because pN is pinned down by traded-good prices, pN cannot adjust to clear the nontraded market; instead nontraded output XN adjusts to equate demand and supply.
  - Consequence: with d pN = 0 the welfare effect of aid reduces to the direct effect plus the “Johnson effect” (change in value of production at world prices), driven solely by Rybczynski effects.
  - Yano and Nugent’s assumed combinations of sectoral output changes violate the Rybczynski theorem: in a three-sector, two-factor model with both factors used in all sectors, an increase in capital must cause some outputs to rise and some to fall; it is impossible for both the import-competing and the nontraded sector outputs to remain unchanged when capital specific to N rises.

### Alternative models that allow pN to adjust — overview
- Two model modifications allow pN to be endogenous:
  1. Specific-factors model: each sector uses sector-specific capital and labor is mobile across sectors (analyzed in detail).
  2. Model with three mobile factors and all factors mobile across sectors (discussed briefly).
- Terminology:
  - “Short-run paradox”: immiserization in a specific-factors model (sector-specific capital).
  - “Long-run paradox”: immiserization in a model with all factors mobile.

### A specific-factors model — structure and welfare accounting
- Zero-profit and full-employment conditions (source notation):
  - Sector-specific returns rj and a common wage w (equations (11)–(13) and (14)–(17) in source).
  - Labor market clearing: L = aE L XE + aM L XM + aN L XN.
  - Sector-specific capital constraints: KE = aE K XE, KM = aM K XM, KN = aN K XN.
- pN determination:
  - pN is endogenous and determined by the market-clearing condition E_N(pE, pM, pN, V) = G_N(pE, pM, pN, V) (equation (18)).
- Welfare effect with aid as an increase in KN (equation (19) notation):
  - dU = -UMM EtP - EG tP G dV + [terms involving EG d pN] (source notation).
- Necessary directional consequences:
  - An increase in KN must cause output of the imported good to fall at constant prices (MV G < 0 in equation (21) per source text).
  - Immiserization requires the right-hand side of equation (21) to be positive and greater than *(VMMV)GtP (i.e., the reduction in value of production at world prices from traded sectors).

### Intuition for immiserization in the specific-factors model
- Immiserization can occur only when MN EG and d pN have opposite signs. Concretely (as given in source notation):
  - Case (i): d pN > 0 and imports and the nontraded good are complements in demand (MN EG − < 0).
  - Case (ii): d pN < 0 and imports and the nontraded good are substitutes in demand (MN EG − > 0).
- Mechanism: In both cases, aid leads to a reduction in demand for imports, which exacerbates the distortion from the tariff and can yield a net welfare loss if the required sign and magnitude conditions hold.

### Effects of aid on the price of the nontraded good and welfare (Section 2 summary)
- Totally differentiating the nontraded market-clearing condition (equation (18)) yields a relationship (equation (22)) between dP_N, dV, dU, and dG.
- Substituting dU from equation (7) into (22) yields equation (23): the effect of aid on pN.
- Substituting equation (23) into equation (19) gives equation (24): the welfare effect of foreign aid expressed as a change in recipient factor endowments, dV.
- In stable models, *()0 UMMU EtPE−>* so the effect of aid on welfare depends on the sign of the right-hand side of equation (24).
- With a tariff and no terms-of-trade effects:
  - The tariff initially reduces welfare by reducing imports below the optimum (raising domestic production and reducing domestic consumption of the importable good).
  - A transfer that increases imports and tariff revenue raises welfare; if imports and tariff revenue decline, welfare falls.
- Immiserization is only possible when dP_N and (* )MNMN EG− are of opposite sign.

### Cases in which the price of the nontraded good rises (dP_N > 0)
- For dP_N > 0, equation (23) must be positive; numerator and denominator of (23) have the same sign.
- For aid to immiserize when dP_N rises, equation (21) and the following must hold (source equation notation):
  - Condition (25): ()()0 NVUMMUNUMMVV GEtP EEtP GG−+ −>
  - Condition (26): ()()( )0 NNNNUMMUNUMMNMN EGEtPE EtPE G−−+ −>
  - Condition (27): ()() VMMVMMNMNN GtP GdVtP EGdP−<−−
- Logical implications and conclusions in the source:
  - Equation (27) requires (* )MNMN EG−< since 0 N dP> and the left-hand side is positive.
  - Equation (26) requires (* )MNMN EG−> because (* )NNNN EG−<: an increase in pN must reduce excess demand for N in stable markets.
  - Therefore, immiserization is not possible when dP_N rises under these sign conditions.
- Alternative sign case (numerator and denominator of (23) both negative): immiserization requires
  - Condition (28): ()()0 NVUMMUNUMMVV GEtP EEtP GG−+ −<
  - Condition (29): ()()( )0 NNNNUMMUNUMMNMN EGEtPE EtPE G−−+ −<
  - Condition (30): ()() VMMVMMNMNN GtP GdVtP EGdP−<−−
- Derived inequalities in the source:
  - (31): * * ()() () NNNNUMMU MNMN MNU EGEtPE EG tP E −− − −<
  - (32): * * ()() () VMMV MNMN MN GtPGdV EG tP dP − −<
- Interpretation (source conclusions):
  - Right-hand side of (31) is positive, so any negative value for (* )MNMN EG− will satisfy it.
  - Right-hand side of (32) is negative (since 0 N dP>). Thus values of (* )MNMN EG− that satisfy (32) will satisfy both (31) and (32).
  - Conclusion: immiserization is possible in this case provided the degree of complementarity between imports and the nontraded good is sufficiently high (must satisfy (32)). Mere complementarity is not sufficient.

### Cases in which the price of the nontraded good falls (dP_N < 0)
- Aid as increased capital used only in the nontraded sector can cause dP_N < 0 if the Rybczynski effect outweighs increased demand (via equation (23)).
- For dP_N < 0 numerator and denominator of (23) must have opposite signs. For immiserization (equation (21)) consider:

Case A — numerator positive, denominator negative:
- Required conditions (source notation):
  - (33): ()()0 NVUMMUNUMMVV GEtP EEtP GG−+ −>
  - (34): ()()( )0 NNNNUMMUNUMMNMN EGEtPE EtPE G−−+ −<
  - (35): ()() VMMVMMNMNN GtP GdVtP EGdP−<−−
- Derived requirements:
  - (36): * * ()() () NNNNUMMU MNMN MNU EGEtPE EG tP E −− − −<
  - (37): * * ()() () VMMV MNMN MN GtPGdV EG tP dP − −<
- Notes and interpretation:
  - Right-hand sides of (36) and (37) are both positive.
  - Immiserization can occur if imports and the nontraded good are substitutes in demand when 0 N dP<, but the degree of substitutability is constrained by (36) and (37).

Case B — numerator negative, denominator positive:
- Required conditions (source notation):
  - (38): ()()0 NVUMMUNUMMVV GEtP EEtP GG−+ −<
  - (39): ()()( )0 NNNNUMMUNUMMNMN EGEtPE EtPE G−−+ −>
  - (40): ()() VMMVMMNMNN GtP GdVtP EGdP−<−−
- Derived implications:
  - (41): * * ()() () NNNNUMMU MNMN MNU EGEtPE EG tP E −− − −>
  - (42): * * ()() () VMMV MNMN MN GtPGdV EG tP dP − −<
  - Combined requirement (43): ()()( )0() NNNNUMMUVMMV MNMN MNUM N EGEtPEGtPG EG tP EtP dP −− −− <<−<
- Interpretation:
  - Aid can immiserize in this case provided imports and the nontraded good are substitutes, but only for values of (* )MNMN EG− that satisfy (43).
  - There is no guarantee such a value exists because * *()() NNNNUMMU MNU EGEtPE tP E −− − might be greater than * *() VMMV MN GtPG tP dP − −.

### A model with all factors mobile — comparison with specific-factors case
- With three goods and three mobile factors, pN will adjust in response to aid; the condition for immiserization is exactly the same as equation (21) from the specific-factors model.
- Key difference:
  - In the specific-factors model, the left-hand side of equation (21) is positive because output of the importable good must fall when aid increases capital specific to N (0 MV G< when ˆ 0 N K>), hence *()0 VMMV GtPG−>.
  - With all factors mobile, factor intensities determine the sign of MV G; therefore *() VMMV GtPG− could be positive or negative depending on factor intensities across sectors.
- Otherwise, the likelihood analysis of immiserization follows the same logic as in the specific-factors model.

### Conclusion — main findings and implications
- Yano and Nugent (1999) erred by:
  - Using a model where pN cannot adjust because traded-good prices determine factor returns and thus pN independently of demand.
  - Incorrectly assuming aid could result in no sectoral output reductions, contrary to Rybczynski implications.
- This paper’s results:
  - In a small open economy with a tariff and an adjustable nontraded good price, the welfare effect of aid depends crucially on how aid affects pN and whether the imported good is a substitute or complement for the nontraded good in demand.
  - If aid (capital specific to the nontraded sector) raises pN:
    - Immiserization can occur only if the imported good is a complement in demand and the degree of complementarity satisfies equation (32).
    - Immiserization is not possible if the imported good is a substitute when pN increases.
  - If aid lowers pN:
    - Immiserization can occur only if the imported good is a substitute, with the degree of substitutability limited by equations (36) and (37).
    - This falling-price case may present the greater chance of immiserization because it requires imports and nontraded goods to be substitutes.
  - Empirical evidence cited by Yano and Nugent suggests imports and nontraded goods are substitutes, but this paper shows the degree of substitutability must meet specific restrictions for immiserization to occur.
- Additional intuition:
  - “Overexpansion” of the nontraded sector can engender immiserization. In a specific-factors model, increasing capital used only in the nontraded sector must raise nontraded output and reduce output of all other goods (including imports). Including a nontraded good may reduce the chances of immiserization from aid specific to the nontraded sector because it causes output of the tariff-distorted import sector to decline, which is welfare-improving.

*Source: IMF Working Paper WP/06/129, “Immiserizing Foreign Aid: The Roles of Tariffs and Nontraded Goods,” Stephen Tokarick, May 2006.*

### Section 1

### _wp06129 - Section 1

### Introduction
- Research question: Whether foreign aid in the form of an increase in capital used only in the nontraded sector can reduce recipient real income (immiserize the country) in the presence of tariff distortions.
- Key claim: The welfare effect of such aid depends on how aid affects the price of the nontraded good and on whether imports and the nontraded good are substitutes or complements in demand.
- Context: Past literature (e.g., Johnson (1967); Bhagwati, Brecher, and Hatta (1983); Jones (1975)) shows factor accumulation can reduce welfare under distortions; Yano and Nugent (1999) proposed a “transfer paradox” via nontraded sector adjustments, but their analysis is shown here to be incorrect for two main reasons:
  - Their model structure precluded any adjustment in the price of nontraded goods.
  - Their assumptions about how increases in a factor endowment affect sectoral outputs were inconsistent with standard results (Rybczynski effects).

### The Yano and Nugent Model
- Model specification:
  - Three goods: exports (E), imports (M), nontraded (N).
  - Two factors: labor and capital, mobile domestically but not internationally.
  - Small country: traded good world prices exogenous: *p1 and *p2.
  - Initial ad-valorem tariff on imports: t.
- Domestic prices (as stated in the source):
  - Exports: p1 = *p1 (equation (1) notation in source).
  - Imports: p2 = *p2 (1 + t) (equation (2) notation in source).
- Core critique:
  - Zero-profit conditions (equations (3)–(5) in source) imply wage (w) and rental rate (r) are determined by the two traded-goods prices, so the nontraded price pN is determined by costs (equation (5)) and therefore is exogenous in this model structure.
  - Because pN is pinned down by traded-good prices, pN cannot adjust to clear the nontraded market; instead nontraded output XN adjusts to equate demand and supply. Thus Yano and Nugent’s model cannot capture a price response in pN to aid.
- Consequence for welfare analysis:
  - With pN fixed (d pN = 0), the welfare effect of aid reduces to the direct effect plus the “Johnson effect” (change in value of production at world prices), driven solely by Rybczynski effects (sectoral output responses to factor endowment changes).
  - Condition for aid to reduce welfare (as presented in the source, equation (9)/(10) notation): aid reduces welfare if it reduces the value of production measured at world prices for traded goods and the exogenous market price of the nontraded good. This is analogous to Johnson (1967) and Caves and Jones (1973) results.
- Additional critique:
  - Yano and Nugent’s assumed combinations of sectoral output changes violate the Rybczynski theorem: in a three-sector, two-factor model with both factors employed in all sectors, an increase in capital must cause some outputs to rise and some to fall; it is impossible for both the import-competing and the nontraded sector outputs to remain unchanged in response to an increase in capital specific to the nontraded sector.

### An Alternative Model with pN Flexible — Overview
- Two model modifications allow pN to adjust:
  1. Specific-factors model: each sector uses sector-specific capital and labor is mobile across sectors (analyzed in detail).
  2. Model with three mobile factors and all factors mobile across sectors (discussed briefly in source).
- Terminology:
  - “Short-run paradox”: immiserization in a specific-factors model (sector-specific capital).
  - “Long-run paradox”: immiserization in a model with all factors mobile.

### A Specific-Factors Model (detailed)
- Modified zero-profit conditions (equations (11)–(13) in source) use sector-specific returns rj and a common wage w.
- Full-employment conditions (equations (14)–(17) in source) specify labor market clearing and sector-specific capital constraints:
  - L = aE L XE + aM L XM + aN L XN (equation (14) notation).
  - KE = aE K XE, KM = aM K XM, KN = aN K XN (equations (15)–(17) notation).
- pN determination: pN is determined by market-clearing condition E_N(pE, pM, pN, V) = G_N(pE, pM, pN, V) (equation (18) notation), so pN is endogenous and adjusts to equate demand and supply for the nontraded good.
- Welfare effect with aid in the form of an increase in KN (capital used only in nontraded sector) (equation (19) notation):
  - dU = -UMM EtP - EG tP G dV + [terms involving EG d pN] (presented in source notation).
- Conditions for immiserization (source equations (20) and (21) notation):
  - Since an increase in KN must cause output of the imported good to fall at constant prices, MV G < 0 in equation (21) (source states 0 MV G <).
  - Immiserization requires the right-hand side of equation (21) to be positive and greater than *(VMMV)GtP (i.e., the reduction in value of production at world prices from traded sectors).
- Intuition and necessary conditions (direct quotes/paraphrase of source results, preserving signs and relations exactly as described):
  - Immiserization can occur only when MN EG and d pN have opposite signs. Concretely, immiserization occurs if:
    - Case (i): d pN > 0 (price of nontraded good increases) and imports and the nontraded good are complements in demand (MN EG − < 0 in source notation); or
    - Case (ii): d pN < 0 (price of nontraded good decreases) and imports and the nontraded good are substitutes in demand (MN EG − > 0 in source notation).
  - In both cases, aid leads to a reduction in demand for imports, which exacerbates the distortions caused by the tariff and can lead to a net welfare loss.

### Conclusion (as in source)
- Yano and Nugent’s conclusions that aid may immiserize via nontraded sector adjustments are incorrect under their model structure because pN is exogenous there and because of inconsistent application of Rybczynski effects.
- When pN is allowed to adjust (e.g., specific-factors model), aid in the form of sector-specific capital can immiserize a tariff-distorted small country, but only under clear conditions: the sign of the change in pN and the sign of the cross-price demand effect between imports and the nontraded good must be opposite (i.e., complementarity or substitutability in demand interacts with the direction of pN change).
- The larger the contraction in importable output induced by the aid (more negative MV G), the less likely immiserization is, because contracting importable output offsets the tariff distortion.

*Source: IMF Working Paper WP/06/129, “Immiserizing Foreign Aid: The Roles of Tariffs and Nontraded Goods,” Stephen Tokarick, May 2006.*

### Section 2

### _wp06129 - Section 2

### Effects of aid on the price of the nontraded good and welfare
- Totally differentiating equation (18) yields equation (22), relating dP_N, dV, dU, and dG.
- Substituting dU from equation (7) into (22) yields equation (23), the effect of aid on the price of the nontraded good.
- Substituting equation (23) for dP_N into equation (19) gives equation (24), the welfare effect of foreign aid expressed as a change in recipient factor endowments, dV.
- In stable models, *()0 UMMU EtPE−>* so the effect of aid on welfare depends on the sign of the right-hand side of equation (24).
- With a tariff distortion and no terms-of-trade effects:
  - The tariff initially reduces welfare by reducing imports below the optimum (raising domestic production and reducing domestic consumption of the importable good).
  - A transfer that increases imports and tariff revenue raises welfare; if imports and tariff revenue decline, welfare falls.
- Immiserization is only possible when dP_N and (* )MNMN EG− are of opposite sign (text reference).

### Cases in which the price of the nontraded good rises
- For aid to raise the price of the nontraded good, equation (23) must be positive; this requires numerator and denominator of (23) to have the same sign.
- For aid to immiserize when dP_N rises, equation (21) must also hold. The three conditions listed are:
  - Condition (25): **()()0 NVUMMUNUMMVV GEtP EEtP GG−+ −>** (equation (25))
  - Condition (26): **()()( )0 NNNNUMMUNUMMNMN EGEtPE EtPE G−−+ −>** (equation (26))
  - Condition (27): **()() VMMVMMNMNN GtP GdVtP EGdP−<−−** (equation (27))
- Logical implications:
  - Equation (27) requires (* )MNMN EG−< since 0 N dP> and the left-hand side is positive.
  - Equation (26) requires (* )MNMN EG−> because (* )NNNN EG−<: an increase in the price of the nontraded good must reduce excess demand for the nontraded good in stable markets.
  - Therefore, immiserization is not possible when dP_N rises under these sign conditions.
- Alternative case: numerator and denominator of (23) both negative. For immiserization the conditions are:
  - Condition (28): **()()0 NVUMMUNUMMVV GEtP EEtP GG−+ −<** (equation (28))
  - Condition (29): **()()( )0 NNNNUMMUNUMMNMN EGEtPE EtPE G−−+ −<** (equation (29))
  - Condition (30): **()() VMMVMMNMNN GtP GdVtP EGdP−<−−** (equation (30))
- Additional derived inequalities:
  - Equation (29) implies (31): ** * * ()() () NNNNUMMU MNMN MNU EGEtPE EG tP E −− − −<** (equation (31))
  - Equation (30) implies (32): ** * * ()() () VMMV MNMN MN GtPGdV EG tP dP − −< −** (equation (32))
- Interpretation:
  - Right-hand side of (31) is positive, so any negative value for (* )MNMN EG− will satisfy it.
  - Right-hand side of (32) is negative (since 0 N dP>). Thus, values of (* )MNMN EG− that satisfy (32) will satisfy both (31) and (32).
  - Conclusion: immiserization is possible in this case provided the degree of complementarity between imports and the nontraded good is sufficiently high (degree of complementarity must satisfy (32)). It is not sufficient merely that they be complements; complementarity must be high enough.

### Cases in which the price of the nontraded good falls
- Aid in the form of increased capital used only in the nontraded sector can cause a decline in the nontraded good price if the Rybczynski effect outweighs increased demand (as shown by equation (23)).
- For dP_N to be negative, numerator and denominator of (23) must have opposite signs. For immiserization equation (21) must hold. Consider cases:

Case A: numerator of (23) positive and denominator negative
- Conditions required for immiserization:
  - Condition (33): **()()0 NVUMMUNUMMVV GEtP EEtP GG−+ −>** (equation (33))
  - Condition (34): **()()( )0 NNNNUMMUNUMMNMN EGEtPE EtPE G−−+ −<** (equation (34))
  - Condition (35): **()() VMMVMMNMNN GtP GdVtP EGdP−<−−** (equation (35))
- Derived requirements:
  - From (34): (36) ** * * ()() () NNNNUMMU MNMN MNU EGEtPE EG tP E −− − −<** (equation (36))
  - From (35): (37) ** * * ()() () VMMV MNMN MN GtPGdV EG tP dP − −< −** (equation (37))
- Notes:
  - Right-hand sides of (36) and (37) are both positive.
  - Immiserization can occur if imports and the nontraded good are substitutes in demand when 0 N dP<, but the degree of substitutability is limited by (36) and (37).

Case B: numerator of (23) negative and denominator positive
- Conditions for immiserization:
  - Condition (38): **()()0 NVUMMUNUMMVV GEtP EEtP GG−+ −<** (equation (38))
  - Condition (39): **()()( )0 NNNNUMMUNUMMNMN EGEtPE EtPE G−−+ −>** (equation (39))
  - Condition (40): **()() VMMVMMNMNN GtP GdVtP EGdP−<−−** (equation (40))
- Derived implications:
  - From (39): (41) ** * * ()() () NNNNUMMU MNMN MNU EGEtPE EG tP E −− − −>** (equation (41))
  - From (40): (42) ** * * ()() () VMMV MNMN MN GtPGdV EG tP dP − −< −** (equation (42))
  - Combined requirement (43): **()()( )0() NNNNUMMUVMMV MNMN MNUM N EGEtPEGtPG EG tP EtP dP −− −− <<−< −** (equation (43))
- Interpretation:
  - Aid can immiserize in this case provided imports and the nontraded good are substitutes, but only for values of (* )MNMN EG− that satisfy (43).
  - There is no guarantee such a value exists because * *()() NNNNUMMU MNU EGEtPE tP E −− − might be greater than * *() VMMV MN GtPG tP dP − −.

### A model with all factors mobile
- With three goods and three mobile factors, the price of the nontraded good will adjust in response to aid.
- The condition for immiserization is exactly the same as in the specific-factors model (equation (21)).
- Difference from the specific-factors model:
  - In the specific-factors model, the left-hand side of equation (21) is positive, *()0 VMMV GtPG−>, because output of the importable good must fall when aid increases capital specific to the nontraded sector (0 MV G< when ˆ 0 N K>).
  - With all factors mobile, factor intensities determine the sign of MV G; therefore *() VMMV GtPG− could be positive or negative depending on factor intensities across sectors.
- Except for this difference, the likelihood analysis of immiserization is the same as in the specific-factors model.

### Conclusion — main findings and implications
- Yano and Nugent (1999) analyzed welfare impact of aid-as-capital in a small tariff-distorted economy and concluded aid could harm recipients via adjustments in the nontraded sector, but did not specify required adjustments. Their analysis had two errors:
  - The price of the nontraded good cannot adjust in their model because traded-good prices determine wage and rental rates, pinning down the nontraded good price independently of demand.
  - They incorrectly assumed aid could result in no sector experiencing a reduction in output, contrary to theory.
- This paper’s results:
  - In a small open economy with a tariff and an adjustable nontraded good price, the welfare effect of aid depends crucially on how aid affects the price of the nontraded good and whether the imported good is a substitute or complement for the nontraded good in demand.
  - If aid (capital specific to nontraded sector) raises the nontraded good price:
    - Immiserization can occur only if the imported good is a complement in demand and the degree of complementarity satisfies equation (32).
    - Immiserization is not possible if the imported good is a substitute when the nontraded good price increases.
  - If aid lowers the nontraded good price:
    - Immiserization can occur only if the imported good is a substitute, with the degree of substitutability limited by equations (36) and (37).
    - This falling-price case may present the greatest chance of immiserization because it requires imports and nontraded goods to be substitutes.
  - Empirical evidence cited by Yano and Nugent suggests imports and nontraded goods are substitutes, but this paper shows the degree of substitutability must meet specific restrictions for immiserization to occur.
- Additional intuition:
  - “Overexpansion” of the nontraded sector can engender immiserization. In a specific-factors model, increasing capital used only in the nontraded sector must raise nontraded output and reduce output of all other goods (including imports). Including a nontraded good may reduce the chances of immiserization from aid specific to the nontraded sector because it causes output of the tariff-distorted import sector to decline, which is welfare-improving.

* _wp06129 - Section 2_*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06129.pdf_