| Boiling Point No. 17 - December 1988 |
by Simon Burne, ITDG
In every country in the world you can find development programmes which include a large amount of subsidy. Just look at any project or programme concerned with primary health care, education or drinking water, for example. Surely, then, there is little point in re-opening the argument for subsidy. Sadly, perhaps, more and more is the very act of subsidy being called into question, and it is important to explore just what subsidies can and cannot deliver. This is especially important for Stove Programmes which, in many ways, are now standing at a cross-roads.
It is worthwhile, first, reciting the arguments against subsidy in general terms. Subsidy is an economic mechanism for achieving a social goal: it is a means whereby all or some people avoid paying the full economic price for any good or service.
Figure 1 shows the economic effect of a subsidy. Conventional economics holds that demand will fall as price rises, whereas production (supply) will increase. The market price (that is the full economic price) is determined where supply and demand are in balance. Thus in Figure 1, Q1 stoves will be supplied and purchased at price P1. A subsidy, however, forcing prices down to P2 would increase demand to Q2. Since producers would still only be receiving price P1, production would stay at Q1, creating a shortage in the market. To satisfy that shortage, producers would have to be paid price P3, which is way above the market price. That surplus demand in practice is usually met by the development of a black market, and a black market price.
Black market prices tend to encourage people to find ways of delivering the goods to those with money and poorer people often prefer the money from selling their goods rather than the good itself. This is obviously only true for goods that are easily movable. A mud stove or a ceramic stove fixed in mud obviously has little transfer value and in these cases a subsidy probably rests with those it is meant to assist. But a lot depends on who gets the subsidy. In India, for example, the subsidy goes to the stove maker/installer. Because of the scale of the programme, quality control is not all that it might be and the consequence is that a number of unscrupulous people are building sub-standard stoves and pocketing the subsidy. In these circumstances poorer people can often be worse off at the end of the exercise.
Subsidies, there is no doubt, are extremely difficult to introduce effectively. There is, however, a range of circumstances where subsidies can be and have been - effectively used to speed up a programme and thus maximise impact.
A number of arguments have been advanced as to why subsidy is necessary in stove programmes. Firstly, it has been argued that without a subsidy the stove would not reach poorer groups. This argument is probably the weakest, since it has long term and broad implications seldom thought through by planners. If a subsidy is introduced in these circumstances it will need to be permanent and is likely to be very expensive and widely abused. Its permanence comes from the inescapable logic that if people cannot afford their first stove they are highly unlikely to be in a position to afford a replacement and thus the only way a stove programme could achieve permanent change under those circumstances would be to extend subsidy indefinitely.
The subsidies are likely to be costly because any system which targets poorer people specifically will require administration and means testing of some form. If a general subsidy is applied, the administration costs are much less, but the total amount of money spent on subsidy will obviously be greater. In practice, most subsidised programmes take the latter approach. Only in countries where existing mechanisms for reaching poorer people exists (like the state shops for harrijan untouchables in India) can a targetted subsidy hope to succeed.
Any targetted subsidy system is open to abuse. This is most especially true where vouchers are used, since vouchers then attain an exchange value themselves; or where government administrators have to decide on entitlement, which opens numerous opportunities for bribery.
Blanket subsidies often fail simply because they are usually restricted by time. Most programmes plan for subsidies to be removed after two or three years, by which time the programme should be "self-sustaining". What often happens is that better-off people benefit from the subsidy, as-they are the people who can take the (albeit reduced) risk of purchasing a new product. The market expands at the subsidised price, although usually not much beyond 10-20% of the total market. With the removal of the subsidy, the price rises at the same time as the better-off market reaches saturation. Poorer people, now more confident in the product, suddenly find they can no longer afford it. The bottom drops out of the market, producers become disillusioned and the programme grinds to a complete halt. The question we must ask, therefore, before introducing a subsidy on poverty-focus grounds, is why can't people afford the stove at the full market price? If it is because people are destitute - say as a result of famine or because they are refugees and face a severe food shortage, then a hefty subsidy can be justified. Indeed, in those circumstances, it is difficult to see the point, or the justice, of charging anything at all.
But in "normal" circumstances, it is likely that a subsidy is being used to conceal the fact that the basic stove design is inappropriate. If a stove is properly designed through interactive research with both users and producers, it should meet users' needs and be affordable by them. If the users are outside the cash economy, or have only marginal cash resources (the case for the majority of the World's rural poor) then a stove with a money price is obviously pointless and stoves would have to be bartered with artisans or built by the users themselves.
If we look at just two programmes where subsidy has been introduced specifically to target the poor, we can see the problems. In India, as mentioned earlier, subsidy has led to corruption in the form of poorly built stoves yielding big profits to the unscrupulous builders, with little come- back for poorer people except to stop using the stove. In Hambantota in Sri Lanka, the evidence shows that it is still the middle class who have adopted the stove first, and they could have afforded the full price of the stove. Many people even offered extra money so that they could get their stove sooner. Now the subsidy is ending and it is unclear where that will leave the poor.
A second argument for subsidy that has been gaining ground in recent years is that clean smoke-free kitchens are a basic human right; stove programmes are, therefore, social development programmes and should be subsidised in the same way that other social programmes are - like the immunisation or drinking water programmes mentioned earlier. Certainly Kirk Smith's excellent work (see BP 13) would suggest that smoky kitchens are a major cause of respiratory infections and cancer. Smoke-free stoves should, therefore, form a significant part of any preventative community health care programme wherever cooking is carried out in a smoky room. If these arguments are won, then governments need to address themselves to the permanent support of smokeless stoves with a built-in and continuous level of subsidy to ensure accessibility by all groups. In these circumstances, a blanket subsidy would still probably prove more efficient than a targetted one. As yet, no stove programme has successfully won this argument with funders or governments, except to a certain extent in India.
A third argument for subsidy has been that the research and development costs of a product aimed at poorer people are as high as for high-value products and that the development programmes should bear those cost so that producers and consumers only face the marginal cost of production. I don't in fact know of a single case of a stove bearing the Research and Development costs, but several papers have been written recently suggesting that these costs should be recouped. This is patently absurd. Any development programme aims to provide a stimulus to the economy it is working in. The reason development aid exists is because that stimulus cannot be generated simply by the existing economy and its mechanisms. Research and development is often particularly weak in this context. To argue that the poor should be expected to pay the support cost of programmes is the same as arguing that all aid should be loaned at commercial rates. That is one stage - and only a small one - from saying that aid as a whole should be discontinued.
There are three further arguments for subsidy. One is that often the initial costs of production are very high and it takes a high price to persuade producers to start making the stoves. The subsidy here is used to enable the stove to be sold at its long-term price from the beginning of a programme. There are two dangers here; firstly, it often means that the programme has to interpose itself between producer and consumer, buying at a higher price than it sells the stove for. This tends to make producers reluctant to find their own sales outlets and thus creates dependency. Ends the programme, ends the production. Secondly, it is very risky because it has to make assumptions about what the long-term price might be. Get that wrong and you've got trouble.
Self-build stoves are designed specifically to be free in cash terms for the user. The long-term aim is obviously to instil into the culture the new stove design so that it becomes the "normal" thing to do to build the improved stove. This requires considerable investment to build awareness and motivation for at least a generation, that is a minimum of 20 years. This investment, with its necessary subsidy can be justified from a national economic stand point because it underpins the success of the programme. To try and recoup these costs would be to undermine the very rational for the original stove strategy. The same can be said, especially in the early days, for commercial stoves, where subsidised training can ensure quality and encourage the first artisans or workshops to take the risk of changing production away from an established line to an unknown one.
Lastly, subsidy can be argued on purely commercial grounds. In much the same way as companies sell products below cost as "loss leaders", small numbers of stoves can be targetted geographically to aim awareness of the stove and create a demand. Rather than reducing the price of the stove, however, other ploys can be more effective: buy a stove and get a free bundle of firewood; or trade in your old stove and get a new one at half price, for example.
Alternatively, whole marketing campaigns can be subsidised to accelerate awareness and demand. It is unlikely that private entrepeneurs will undertake massive promotion on their own for three reasons: firstly, low-cost products imply low margins and will not lend themselves to costly promotion campaigns. Secondly, new products are only heavily promoted in the commercial sector to establish a rapid dominant position in the market, preferably a monopoly. Since programmes on the other hand tend to work with a number of producers making the same product, any one person advertising will see the benefits, at least in part, flowing to other people. The third reason is, quite simply, that most producers involved in stove programmes do not have the resources to invest in expensive advertising.
Let's, then, try and answer the five questions asked in the title:
1. To accelerate the speed of development and dissemination through subsidised - research and development - training - promotion campaigns
2. As part of a social programme of promotion in health care.
3. NOT to bring down the cost to poorer people unless clear channels already exist to do that and the stove is part of a long term social programme.
Subsidies should be blanket because of low administration costs. The stove designs should be targeted at poorer people, not the subsidy.
Because of the nature of any development programme (finite life, finite resources, finite political backing) any subsidy should be phased out over time, and that phasing out should be planned. Even in social health programmes, research should continue to bring the actual cost of the stove down, in case subsidies are removed at some future date.
Subsidies can be targeted geographically, for example to the poorest region. There is, however, a danger that if the subsidy is greater than transport costs, you will simply see stoves being shipped back out of that region to other areas.
It is essential to try and maximise control of subsidy in the hands of the programme itself to avoid abuse or misapplication. This is no problem in the case of research and development, training and promotion as costs can be paid direct by the programme. But with blanket subsidies as part of a social programme, this becomes much more serious.
Subsidising producers or installers leads to the quality problems found in India. Placing the programme in the position of intermediary between producer and consumer creates dependency and threatens the long-term viability of the programme, as has been found in Sri Lanka. One option, which to my knowledge has not been tried, would be to pay consumers subject to their having an installed working stove. Not only would this give consumers an added incentive to ensure that the stove is properly made but would also provide an excellent opportunity for follow-up training in use and monitoring of effectiveness. To enable this to work, consumers would have to purchase off approved producers. There would either have to be a credit arrangement with produces or a good credit system to users; and the programme would have to be rapid in its response to users to avoid any hardship. But it appears to offer one of the few ways of providing blanket or even targetted subsidy without the programme imposing itself permanently on the production and distribution system.
Subsidies, then, can work and can help support a more rapid development and dissemination programme. But they must be used carefully and discriminately and, above all, they must not be used to disguise a bad stove design.