The D.C. bottle initiative, which went down to defeat on Tuesday, was well intentioned but fundamentally flawed. First, it attempted to solve a national problem with a local ordinance that would have shifted a large volume of sales (and tax receipts) from the District to Maryland. Litter in D.C. might have been only modestly reduced as residents evaded the deposit by purchasing soft drinks and beer in the suburbs. Second, the initiative would have imposed excessive costs on small stores that lack sufficient space to store empty bottles.

A successful response to the problem of litter created by bottles should be based on two principles: first, it should be national rather than local in order to avoid distorting purchases from one jurisdiction to another. Second, it should start from the simple concept of a user fee, under which people compensate the government for particular costs that their activities impose on society. The costs in this case are litter and trash disposal.

A federal user fee (i.e., tax -- but one cannot use that word until early 1989) should be imposed on the production of new cans and bottles and on their importation from abroad. The level of the fee should be the estimated costs to society of trash collection and litter. No additional fee would be collected on bottles that are refilled. This would create a strong financial incentive for the recycling and reuse of bottles. The price of new bottles would increase by the amount of the fee, making it far more attractive for producers of soft drinks and beer to collect and refill old bottles.

Owners of retail stores could decide individually whether to become active in purchasing old bottles and selling them back to breweries and soft drink producers. Those with sufficient space would find it profitable to do so, but those lacking space would not have to become involved. New businesses would probably develop for the sole purpose of buying used bottles from individual homes for recycling.

The process of reusing bottles could be simplified if soft drink and beer producers were allowed to cooperate on the design of standard bottle sizes for use by many firms. A uniform beer bottle, for example, would mean that these bottles would not have to be sorted before being returned to brewers.

The incentive for recycling would be determined by the level of the fee on the production of new bottles; if glass litter remained a problem, the fee could be raised by 5 or 10 cents, which would make it even more profitable to reuse old bottles.

Since aluminum cans cannot be refilled but must instead be melted for the production of new cans, they create a modest problem for this system. The solution is to impose the fee on the production of new cans, but to provide parallel rebates for cans that are recycled, creating the same incentive for avoiding litter and reducing trash disposal costs.

The impact of this fee system on imports would be mildly protectionist, which should encourage political support by U.S. brewing firms. The fee would be collected on bottles of imported beer (and perhaps wine), but rebated on bottles sent back. Distances would make such returns of empty bottles very expensive, how-ever, putting foreign brewers at a modest disadvantage in the U.S. market, which should please local brewers.

The funds raised from fees on production and imports of containers could be returned to local governments to cover trash collection costs, which would be the fair approach, or could be retained by the federal government as part of a deficit-reduction package, which is much more likely. This should be another item for the continuing debate over how to reduce the federal budget deficit.

The fee system being proposed here is based on the simple principle that when businesses and individuals impose sizable costs on society, they ought to bear those costs through taxes or user fees. Gasoline taxes are used to finance highway construction, for example, and landing fees help pay for airports. Litter and trash disposal costs are imposed on society by producers and users of bottles and cans. The fee system puts these costs where they belong -- namely, on those who produce and use containers. It also creates a strong financial incentive for the recycling of bottles and cans, which reduces both litter and trash collection costs. Since this would be done by the federal government, there would be no distortion of purchases from, for example, D.C. to Maryland.

Bottle and cans are not the only source of litter and trash collection costs, and the same approach could be used to encourage recycling of other products. There is, for example, the major problem of discarded newspapers.

A fee should be collected for every ton of newspaper that is produced domestically or imported, but no payment should be required when old paper is recycled. This would create the same financial incentive for gathering old newspapers for reuse and for the avoidance of litter. The funds generated by this fee on paper production could be used to further reduce the federal deficit. Newspaper publishers would have to pay more for their raw materials, but their subscribers could sell bundles of old newspapers to recyclers. Besides, an increase in the cost of newsprint would be a small price to pay for a cleaner city and lower trash collection costs. Somebody is trying to pull the plug on my word proces . . .

The writer is a professor of economics at The George Washington University.