Used Clothes as Development Aid: The Political Economy of Rags (SIDA) Part II-B: Analysis of the effects of subsidizing used-clothes imports Chapter 8: Theoretical welfare effects of subsidized imports (introduction...) Introduction of a freight subsidy The positive externality (infant industry) argument again Less than fully functioning markets: Unemployment yet again Distributional effects: Benefiting the poor Import subsidy effects on exporting, and benefits If there is no domestic clothes production Dumping, and other cautions regarding who gets the subsidy, and how Conclusions Chapter 9: Alternative costs and best use of cash and clothes (introduction...) The cost of the freight subsidy The alternative cost of the freight subsidy: Cash Best use of the cash Best use of the clothes Situations where freight subsidies would be warranted: Catastrophes, no supply Conclusions

Used Clothes as Development Aid: The Political Economy of Rags (SIDA)

Part II-B: Analysis of the effects of subsidizing used-clothes imports

Chapter 8: Theoretical welfare effects of subsidized imports

The usual case that can be made for a subsidy, other than simply to put cash in someone's pocket or to allow someone to do something that they like to do, is that there is some public good which is not being provided because of a market failure which can be remedied via the subsidy. For instance, one could argue regarding used clothes that there is a public good in having a better-clothed and more productive workforce, better-clothed and healthier mothers and children, etc., and that in LDCs the lack of adequate clothing may be particularly acute, thus justifying subsidized development projects to provide used clothes. It might also be thought that the incidence of power has created a very unequal income distribution, causing a market failure - in that particular groups of potential consumers have no access to the market - and thus requiring remediation by subsidy.

In Chapter 5 we analyzed theoretically the effects of used-clothes imports in general, first under simple "ideal" conditions, and then with a variety of more realistic complications. We will now consider the specific effects of subsidizing used-clothes imports. We will start with the same set of markets (domestic new clothes, imported new clothes, imported used clothes, and all other goods and services) and the same set of original conditions as before (perfect markets and free trade). Thus we assume that no production subsidies are in place, and that there are no import tariffs or bans. We start from the same assumed equilibrium as before.

Introduction of a freight subsidy

What happens if we introduce a freight subsidy for used clothes? As seen in Diagram 2_u, this will lower the domestic price of used clothes by the amount of the subsidy (s), from p_u to p_u'. Quantity will increase from q_u to q_u'.

Diagram 2_u: Imported used clothes (showing welfare gain with freight subsidy, and its cost)

Because more used clothes are now available more cheaply than before, there should be a reduction in demand for both domestic and imported new clothes, although there will be a positive income effect that would tend to counteract this tendency. The size of the negative demand shifts for domestic and imported new clothes (in Diagrams 2_d and 2_i below) will depend on the relevant elasticities - which describe how demand for new clothes changes when the price of used clothes changes - but the shifts will in any case be in the negative direction, downward and to the left, to D_d' and D_i'; and quantities sold will shift from q_d to q_d' and from q_i to q_i'.

Diagram 2_d: Domestic new clothes (showing welfare loss with used-clothes freight subsidy)

Diagram 2_i: Imported new clothes (showing welfare loss with used-clothes freight subsidy)

What the effect would be on the rest of the economy is hard to work out: It depends on how much of the resources released from the new-clothes sectors is absorbed in the used-clothes sector; one may assume that the other goods and services sector will get some extra resources, when employment in the new-clothing sectors falls.

The welfare effect in the used-clothes market (Diagram 2_u above) is clearly positive: Consumer surplus increases by the shaded area, and there is thus a welfare gain labeled WG_u. The domestic clothing industry (Diagram 2_d above) faces a fall in demand, which leads to a lowering of price from p_d to p_d', and a smaller quantity sold, q_d' rather than q_d. This implies a reduction in consumer surplus, and a reduction in producer surplus as well, for a net welfare loss equal to the shaded area labeled WL_d. In the imported new-clothes sector (Diagram 2_i above), there is also a fall in demand, and therefore also a fall in consumer surplus of the shaded area, a welfare loss labeled WL_i. Finally, the other goods and services sector may absorb the production factors (labor, capital, etc.) released by the domestic clothing industry, thereby increasing production and generating some increase in consumer as well as producer surplus (WG_o, not shown).

What does this add up to then? Assuming (as we have done) that there are no distortions in the domestic economy, and that factor markets are fully functioning, then market prices would properly reflect social scarcities. The total welfare effect on the economy of the freight subsidy on used clothes is then equal to +WG_u -WL_d -WL_i +WG_o, which must be positive overall: The recipient economy gets goods coming in which are cheaper than before and, with working markets and in the absence of distortions, this must be good. That is, under these assumptions, although there is some restructuring required, a freight subsidy definitely helps the recipient.

The positive externality (infant industry) argument again

However, we have already noted that there may be a plausible infant industry argument based on possible positive externalities associated with clothes production. The negative welfare effect that is measured in our analysis for the domestically-produced clothes market (WL_d in Diagram 2_d) would then understate the loss to society of out-competing part of the sector with second-hand imports. Instead of rectifying the distortion, we have aggravated it. If these externalities are significant, the overall effect on the economy of subsidized used-clothes imports might be negative.

Less than fully functioning markets: Unemployment yet again

Similarly, if factor markets are not functioning, so that it is difficult for resources (including labor) to find alternative employment, then the negative effect measured in our analysis would again understate the true loss. Once again, instead of rectifying the distortion, we have aggravated it.

Distributional effects: Benefiting the poor

But are there no other externalities that should be considered? Yes, it is often argued that there are positive distributional effects associated with the sale of used clothes - that it primarily benefits the poor. This could be the case, and it would strengthen the argument for subsidizing imports of used clothes. In fact, if used clothes are exclusively (or primarily) purchased by people who could not afford to enter the market for new clothes - or if subsidized used clothes were given away exclusively to such people - then the last three terms of our welfare formula would drop out, so that +WG_u -WL_d -WL_i +WG_o would become just +WG_u, an obvious gain. In addition, there might be positive externality (distributional) effects, in that, for instance, reducing income inequalities might reduce social tensions and the risk of political instability, or raising the incomes of the poorest population segments might significantly improve their productive capacity.

Import subsidy effects on exporting, and benefits

For simplicity in the discussion of subsidized imports of used-clothes, we have so far assumed that domestically-produced clothes are non-tradable, that is, that the domestic clothing industry does not export. How would the argument change if domestically-produced clothes were exported, and prices were thus set on the world market? Diagrams 2_i and 2_u (for imported new and used clothes, respectively) would be unchanged, of course, but Diagram 2_d (domestic new clothes) would have to be replaced by Diagram 3_d, below. Here, domestic demand for domestically-produced clothes is D_d, and the quantity purchased domestically at price p_d is q_d^d. However, we now assume that we have comparative advantages in this sector, and are able to produce a total of q_d^t, given the world market price p_w = p_d. The difference between domestic supply and domestic demand (q_d^t-q_d^d) is then sold on the international market.

In this case, a subsidy on used-clothes imports would still reduce demand for domestically-produced clothes from D_d to D_d', but this would not affect production. Instead, exports would increase by the amount (q_d^d-q_d^d'). The negative effect of the subsidy would therefore be less. We have now assumed that international demand for the exported clothes is infinitely elastic - that is, that the international market can and will absorb whatever excess production remains after domestic demand is satisfied. This may be too extreme an assumption, so it may well be that there would be some negative effects on production even in the case where there are exports. But just as we saw in the analysis of used-clothes imports in general, exporting ameliorates any negative effects.

Diagram 3_d: Domestic new clothes (with exports, and used-clothes freight subsidy)

If there is no domestic clothes production

If there does not exist any domestic clothes production at all, we would obviously not see a cutback in the production of that sector due to imports of cheap used clothes, and the negative secondary effects would thus be less. However, the introduction of subsidized imported used clothes would still constitute a problem for potential future development of the sector. This sector is recognized as one where many LDCs may be competitive, and where relatively little capital is required to get started. For instance, Toyne (1984) notes that "textiles is one of the first internationally competitive industries generally developed by developing countries." Subsidizing used-clothes imports could therefore have negative long-term effects on development by precluding or limiting potential future development of domestic production.

Dumping, and other cautions regarding who gets the subsidy, and how

Subsidies can take a number of forms and operate in a number of ways, which we need to look at. Who is helped by freight subsidies for used-clothes imports? Is it clear who is getting the benefit, and are they getting the maximum benefit possible? And is the subsidy legal?

If the clothes are being provided to individuals below market cost, then of course those individuals benefit. The International Textiles and Clothing Bureau in Geneva points out that such subsidized sales constitute dumping, and that "a subsidy to exports of used clothing for commercial purposes [is probably not] consistent with the Agreement on Subsidies and Countervailing Measures of the Uruguay Round" of GATT, because it represents unfair competition. Dumping itself (apart from the question of its legality) is not necessarily anything to be worried about: If someone wants to sell us goods at prices lower than costs, it is generally to our advantage, provided that we can use our resources fully in other activities, and provided that the dumping is not a temporary move to knock out our industries, to be followed later by higher prices. While the latter possibility is not so likely in the case we are considering here, it may be difficult - as we have discussed - for a typical LDC to find good alternative uses of resources, so that there may be substantial unemployment.

If the used clothes are sold on the market directly by the NGO, then that NGO's other projects or overhead get the benefit. If the clothes are sold on the market by some governmental, quasi-governmental, or local counterpart organization, then that organization gets the benefit. In any case, clarity of accounting and of understanding would be increased if the funds were allocated directly to the NGO or its other projects, or to the other organization involved. The used clothes can be shipped without subsidy, and the sales receipts will cover the costs. Given likely inefficiencies in sorting and handling, greater benefits may be possible if the used clothes are thus handled in a more business-like way. Providing them via subsidized freight may also lead to greater quantities being imported, and sold more cheaply, than would have otherwise been the case, which again constitutes dumping.

If the clothes are simply misappropriated somewhere along the way, then of course someone still benefits, but not the intended targets, and the resulting unbusiness-like environment is clearly not desirable either.

Conclusions

Subsidies can provide benefits and cause harm; harm may be ameliorated if domestic production is exported; subsidies may be illegal nevertheless

In ideal conditions, any subsidy which provides useful goods cheaper than before is a good thing in itself; but under more complex conditions - with positive externalities and/or unemployment, for instance - subsidies tend to aggravate rather than correct distortions. Cheap used clothes that primarily benefit the poor might give positive distributional effects, however. If domestic clothes production is being exported, any damages from used-clothes imports would be ameliorated. On the other hand, if there is no domestic clothes production, one cannot assume that there is no damage, because cheap competing imports may preclude future development. Practically speaking, subsidies which result in goods being sold below normal market price constitute dumping, which is illegal under international agreements. Subsidies may also constitute indirect funding support for various organizations or projects, or even for individuals who may misappropriate the goods. If the results are intended, it is probably better to effect them directly.

Chapter 9: Alternative costs and best use of cash and clothes

Even if there is extreme concern for poverty, and a freight subsidy for used clothes would have a tremendous impact on poverty, this would not necessarily imply that freight subsidies should be given: One must obviously consider the alternative uses of the money and, for that matter, of the clothes.

It may be helpful to reexamine our basic problem at this point. We have been asked whether used-clothes exports to less-developed countries should be subsidized. This question makes two basic assumptions, and thus we might separate the question into two more basic ones:

1. There is an assumption that used clothes exist in industrial countries and are available for subsidized export. One question we might ask is, what is the best use of these clothes?

2. There is also an assumption that development funds exist in industrial countries which might be available for subsidizing used-clothes exports, because industrial countries have a desire to help - and an interest in helping - less-developed countries develop. An obvious question we might ask is, what is the best use of these funds? What is the best way that industrial countries can help LDCs develop?

There is an obvious, simple complementarity to the two conditions (the existence of used clothes in industrial countries; and the existence of a desire on the part of people in industrial countries to help the development process in LDCs), which is to use available development funds to send available used clothes from industrial countries to LDCs. But is this the best use of the clothes, or of the development funds? These are two separate questions. Let us consider the use of the funds first.

The cost of the freight subsidy

So far, we have looked at the effects of subsidized used-clothes imports only on the recipient country, without comparing these effects with the cost to the donor, or with the alternative cost to the recipient. In Diagram 2_u (in the previous chapter) we showed the cost of the subsidy: If we subsidize the transport of each unit of used clothes by s = p_u - p_u', and the total quantity is q_u', then the total cost of the subsidy (s _* q_u') is equal to the rectangular area ABCD.

The alternative cost of the freight subsidy: Cash

The cost of the freight subsidy (ABCD) is clearly larger than the total increase in consumer surplus in the used-clothes market (+WG_u), much (but perhaps not all) of which would accrue to the poor. We also remember the two other directly negative welfare effects, -WL_d and -WL_i (in Diagrams 2_d and 2_i, respectively), and also the loss of the positive externality that might be associated with clothes production, and the likelihood of increased unemployment resulting from subsidized used-clothes imports.

Best use of the cash

It would therefore be better to put the used-clothes subsidy funds directly in the hands of the poor, and then let them use the money as they see fit, rather than subsidizing used-clothes imports. Such transfers would have all of the distributional benefits of (or more than) subsidizing used-clothes imports, without the loss of efficiency implied by the fact that the cost exceeds the consumer surplus gain. If such cash transfers are not feasible, one could use the money for efficient projects that benefit the poor.

A freight subsidy does not eliminate a market distortion, but rather introduces one, in the form of distorted prices for used clothes. It should imply some benefit for the recipient country, but even without any possible negative secondary effects, it will not increase aggregate welfare as much as a direct transfer of money, which would allow the import of whatever is most wanted and needed. This conclusion does not change when distributional effects are considered. There are certainly better ways to help the poor directly, rather than doing it indirectly, via the used-clothes market.

Best use of the clothes

If it is clear that available development funds can be put to better use than for freight subsidies for used clothes, then it follows that the used clothes themselves can also be put to better use. As we have seen, commercial markets worldwide are willing to convert used clothes into cash, and the cash can then be put to good use in development projects.

Situations where freight subsidies would be warranted: Catastrophes, no supply

Are there any situations where freight subsidies would be warranted? What about a catastrophe situation, where there is no alternative local supply, either of domestically-produced new clothes, or of imported new or used clothes? In this case, the negative side effects of cheap used-clothes imports would not exist. Thus it is possible that there are in fact market failures which NGO projects could help to remedy with Sida-subsidized used-clothes exports.

Analytically, assume we have the situation as depicted in Diagram 4_u, below: There is a demand curve for used clothes, but initially no supply. However, assume now that we were to give such a large freight subsidy that prices in the market go to zero: Then the consumer surplus gain would equal the whole triangle ABE. The cost is again equal to the rectangle ABCD, but now this may be considerably less than the gain, though this, of course, depends on the costs of transporting the goods to the country in question. The difference between this and the former case (Diagram 2_u in the previous chapter) is that now there is no initial commercial supply of used clothes at all. Before, the subsidy only changed the amount of consumer surplus; now the subsidy accounts for all of it, because the existence of any supply at all is conditional upon it. Since we assume that this is a crisis situation where goods would not be available if they were not delivered by the donor, the money transfer option would not exist as an effective alternative. Thus, in such a case, the subsidy would be warranted, basically for humanitarian reasons.

Diagram 4_u: Imported used clothes (catastrophe = no supply - showing welfare gain, and its cost)

The demand curve here is taken to show what people would be willing to pay for different quantities. In the case of a complete disaster, of course, people may be destitute, and there would not be any market demand in spite of great need. The conclusion that transfers in kind would be called for would still seem valid, although we would then have to assign some value to the utility to the population of the clothes transferred.

Disaster relief is thus an example in which people in desperate need of clothing may have no capacity to enter the market to satisfy those needs, not because they themselves are poor (although they may be destitute), but because the markets and the production supporting them have disappeared. However, we have noted that many major international NGOs generally prefer not to utilize second-hand clothes as part of emergency relief, and even the Swedish Red Cross notes that used clothes are not and should not be an important part of emergency planning. There can be a need for clothes as part of relief assistance, perhaps after a catastrophe situation has "normalized", but in such situations it is noteworthy that UFF - which has been severely criticized in the media for selling used clothes commercially in Africa - nevertheless generally gives away new clothes when it comes to relief assistance. It seems that used clothes should be used if and when it can be demonstrated that the need is urgent and cannot reasonably be met from other sources, and that the supply is most appropriate to the need.

Conclusions

Subsidies generally help less than they cost, and may in fact cause harm, although they may be necessary in responding to catastrophes

What do we conclude from this theoretical exercise with regard to freight subsidies? We conclude that subsidies help LDCs get imports at lower prices, which suggests that they are welfare-enhancing. Secondly, however, we note that the aggregate welfare gain is less than the cost of the freight subsidy. It would therefore be better to use the money for direct transfers, or for more efficient projects. This applies even in the case where there is no domestic clothing industry that is negatively affected. And it may actually be easier to target the impact of other types of interventions on the poor.

Moreover, if there are positive externalities associated with domestic clothes production, then the effect of a used-clothes freight subsidy on this sector will tend to negate the welfare gain of the transfer itself, unless the unemployed resources find as good a use in other industries. However, they may be transferred to sectors with less positive external effects, or - if factor markets are less than fully functioning - they may even remain unemployed, and then the negative effect is compounded.

The policy conclusion that can be drawn from this theoretical review is thus that, under normal circumstances, Sida should cease giving freight subsidies for used-clothes exports: This is a costly way to help the poor; the money can be put to better use. The only time where such subsidies might be warranted is in catastrophes, where markets collapse, and the population requires transfers in kind. Freight subsidies should thus only be used within the framework of catastrophe aid.