Often environmental policies are enacted without regard to their trade-altering effects. Researchers and policy-makers have begun to consider the linkages between agricultural trade and pesticide regulation. We derive the welfare effects of the U.S. unilateral ban on Methyl Bromide for domestic producers and consumers from a disaggregated perspective. We examine whether the proposed ban will achieve its environmental goal. We calculate the difference for Florida and California vegetable and fruit growers considering that Mexican farmers, their principal competitors may continue to use this important input. In addition, we examine how alternative policy options affect the welfare changes, use of Methyl Bromide, and the distribution of losses and gains.

The debate preceding the passage of the North American Free Trade Agreement (NAFTA) illustrates the potential for conflict between environmental protection and trade liberalization. Environmentalists see trade liberalization and trade agreements leading to less stringent environmental standards and more difficulty in achieving new regulations as domestic producers demand a "level- playing field". Proponents of trade liberalization claim that policies to protect the environment are disguised trade barriers.
In recent articles involving trade and the environment, researchers have focused on how trade agreements and trade mechanisms are affecting the environment either by increasing production, shifting production sites, or eliminating production. In addition, some attention has been paid to how trade instruments such as a ban on exports or imports can be used to achieve a reduction in pollution. The flip side to this analysis is how environment regulations affect trade purposely or unwittingly. By affecting trade, these environmental regulations may or may not achieve their intended goal or may do so in a way which imposes large welfare losses by distorting resource allocation.

In this paper, we examine what type of policies the U.S. can use to reduce ozone depletion caused by Methyl Bromide (MB), a broad spectrum pesticide that is used primarily as a pre-plant soil fumigant. MB has been classified as a Class I ozone depletor (a global pollutant). As such, the international community is examining ways to decrease its use under the Montreal Protocol. Under the U.S. Clean Air Act, MB's production and importation must be phased out by the January 1, 2001 in the United States. Horticulturists in California and Florida are the prime users of MB in the United States. Their primary competition during different periods of the year comes from 3 states of Mexico: Baja California, Sinaloa, and Sonora . Any policy adopted by the United States must take into account the possibility of increased use of MB in these regions as well as the trade impacts. The ban on MB will force many U.S. growers to change their production technology. Sunding et al. (1993) estimate the losses to California agriculture to be $162,096,000. Deepak, Spreen, and VanSickle (1994) estimate that Florida revenues for seven crops will decline 54 percent from $1.144 billion to $524 million. They find that most of the lost revenue in Florida will be gained by Mexico whose revenue is projected to increase by 65 percent. Neither paper looks as the increase in MB use in Mexico -- thus whether the proposed ban achieves its environmental goal.

From 1984 to 1990, world production and consumption of MB increased at a rate of almost 5 percent per year (UNEP, 1992). In 1990, the total world use was approximately 66,000 metric tons (Handy, 1992). In 1992, Mexico was the largest importer of U.S. produced MB spending $ 1,999,000. In addition, it imported MB from israeli and European companies. Industrialized countries use the vast majority of MB as shown in Chart 1. In 1992, North America accounted for 41% (27,000 mt) of the world's consumption (UNEP, 1995). Eighty-one percent of the global consumption of MB in 1990 was used for soil fumigation of which 44 percent occurred in North America. We can see in Chart 2 that post-harvest commodity fumigation accounted for another thirteen percent. Exporting countries use MB fumigation to disinfect commodities because the importing country requires the treatment. MB might be used for grains, cereals, flowers, fruits, vegetables, or logs.

In November 1990, the U.S. Congress amended the Clean Air Act requiring that all substances with an ozone depleting potential greater than 0.2 be phased out within 7 years of being listed. After being sued by several environmental groups to enforce its regulation, in 1993, the Environmental Protection Agency (EPA) announced that MB would be phased out. Under the Montreal Protocol, countries agreed that MB production is frozen at 1991 levels as of January 1994. In the United States, all production and importation of MB will be stopped in 2001. Consumption of existing MB supplies will be allowed (EPA, 1993).

In a first-best world, we would use a directly linked policy instrument which results in the internalization of the production externality to achieve the optimal use level of the chemical. A Pigouvian tax, subsidization of abatement technology, or possibly a marketable permit scheme might be the optimal instruments to impose on MB users. We demonstrate how some of these mechanisms would be welfare improving relative to the current policy of an unilateral ban. Since MB is a global pollutant, its location of use is less important than the level of use.

In a first-best situation, without any other policy distortions, a typical market equilibrium framework for analyzing the welfare effects of a predetermined level of abatement could be represented in Figure 1. Let S0 be the private supply curve in the first-best world with a closed economy and where the externality is the only market failure. Thus, we assume some pollution is the byproduct of the production process and increases with production. The value of the marginal damage that the pollution causes can be calculated as the health effects, property damage and other negative environmental effects. Since competitive firms do not pay the costs associated with their pollution, they underestimate the true costs of production resulting in too much production and too much pollution. To internalize the externality, i.e. to force firms to consider the marginal pollution costs of their production, an optimal tax is imposed. This tax shifts the supply curve to S' which now represents the vertical summation of the private supply curve and the marginal pollution cost curve. The marginal pollution costs curve reflects the amount of money the pollution victims would have to receive to be indifferent between tolerating the pollution and being compensated, and not having to face the pollution caused by the last unit of production. D0 is the market demand curve. We find the welfare before the tax is A+B (producer and consumer surplus) minus the cost of pollution; the net welfare is A+B-F. After the tax is imposed, societal welfare is A+B. Output has decreased to the socially optimal level given that the social costs of production are now being taken into account. The welfare change is F which is always positive (Just, Hueth, and Schmitz, 1982). In addition, although in many cases private firms face reduced profits following environmental regulations, sometimes producers can benefit from chemical regulations and taxes as it decreases supply and with an inelastic demand may increase their revenue.

With an open economy, however the results are somewhat different. In an open economy, the domestic firms produce a lower quantity and actually may be producing close to the optimal level. Thus, if a tax is imposed without consideration of the trading partners' response, the domestic welfare may decrease. As shown in Figure 2, the domestic producers lose significantly as foreign producers are able to expand production to satisfy a large percentage of the market. The domestic producers' welfare falls from the dark shaded area to the smaller striped area. The foreign producers on the other hand gain both from the higher price and from the expanded production to increase welfare to the dark shaded area on the exporter side. In the case of local pollution, the domestic consumers benefit by obtaining lower cost commodities that do carry with them the pollution costs. This result does not hold though if the pollution has a global effect. Then the pollution does not decrease optimally when a domestic tax is imposed.

We however do not live in a first-best world. Calculating the benefits of reducing MB use is very difficult especially since experts disagree as to the degree human-use of MB affects the ozone depletion. Ozone depletion potential (ODP) is based on a number of factors that affect atmospheric lifetime. These factors then are used to determine activity on a mass per kilogram basis compared with the standard, chlorofluorocarbon-11, which has been assigned an ODP of 1. The debate continues as to the degree MB effects the ozone level as well as whether the bromine they find in the ozone layer is from natural sources or from human-use. In addition, emissions may vary across growers due to tarping, to different quality of plastic tarping, care taken at the ends of beds/fields, and different styles of fumigation: hot gas or injection. However, measuring emissions from growers would be costly and difficult, a classic example of a non-point source pollution problem.

Our research focuses on the interaction between trade and environmental policy. We want to know how alternative environmental policies would change the welfare costs and trade patterns. We look at the value of the marginal product of each pound of MB resulting from a regulation affecting technology choice and other production decisions.
The value of marginal product (VMP) can be calculated for annual crops as

(1)

where the change in Yij is the output per acre in region j and crop i from adopting MB technology or adopting the next-best technology. Pij is the grower price for crop i in region j. The change in Cij is the difference in production costs per acres for MB or the next best technology. In Table 1, each region's best alternative technology is depicted. LRij is the label rate or recommended input usage per acre for crop i in region j. For perennial crops, the VMP is calculated as

(2)

where t indexes time, t is the discount factor for period t and T is the last year of production under the alternative technology. Figure 3 contains the estimated VMP's for both California and Florida. We can see how other types of policies might affect welfare changes and distributional effects of the environmental regulation. For example, it would require a tax of over $10.00 per pound of MB to reduce the level of use 50 percent. On the other hand if Mexico's VMP is of lower value, it might be willing to sell its right to use MB to the United States' growers.

While this analysis provides a general understanding of how different policies may affect MB use, it assumes constant prices. Taking price as parametric is not crucial if the output affected is a small proportion of the market. In this case, however, the output changes can be quite large and the market shares reach over three-quarters. Further analysis was done with the next model which allows for the endogeneity of prices.

In this model, the choice of MB or other technologies is incorporated as a putty-clay model with fixed proportions technology to generate supply by region. We have determined yields under different technologies which are assumed to be non-stochastic. While per acre yields for fruit and vegetables vary from year to year, we examine several technologies that are not currently being used in all regions and thus time-series data on yields was not possible to obtain.

The distributional effects of proposed environmental policies vary between different regions both within and outside the country. The direct costs of environmental and resource policies often vary markedly among regions on both producers and consumers. This heterogeneity implies a need to use regionally disaggregated models in which the regions correspond to the natural differences that exist. Pesticide use patterns vary across regions in response to both economic and environmental conditions. Humid regions with severe pest problems may use toxic insecticides extensively, while farms in high wage regions may rely on herbicides as effective substitutes for labor. In addition, certain regions may have natural abatement characteristics; others may have characteristics which lead to more pollution for a given production technology. We assume the per unit emissions resulting from the the polluting technology will not be different between regions, just the costs and effectiveness of the technology.

Pest management considerations have been important in determining the production location of specific crops. Regions may specialize in crops that are more appropriate climatically with fewer pest problems. Thus, if one region imposes extremely strict regulations against pest control, production may shift to regions that actually have more severe pest problems, which could ultimately result in an overall increase in pesticide applications. This can be seen as an agricultural example of a pollution haven .

Our modeling suggests that three possible outcomes exist of the U.S. ban: 1) Mexico increases its production and its use of MB resulting in an even greater level of MB being used that before the ban, 2) Mexican growers do not have the same soil-borne problems and do not use the same level of MB resulting in a decrease in MB use following the ban with an increase in production, and 3) Mexico is unable to dramatically increase its production with or without MB while Florida and California yields decrease resulting in increases in agricultural prices and losses in consumer welfare.

We have J regions in total producing several different crops. Regions j = 1, ... J1 are in the United States and j = J1 + 1, ... JTot in Mexico. There are I crops, i = 1 ... I in the regions that are being considered. In addition, each producer has a choice of different technologies, h=0, ... H which have varying per acre costs and per acre yields.

Crops are produced with fixed proportion production functions. We have yijh, the yield per acre of crop i in region j with technology h. The acreage used in a particular crop in a region is lij with the maximum possible acres set at Lij. The MB used per acre of crop i in region j is mij. The consumption of MB in region j on crop i is represented by mij*lij and sell for the commercial price of w per pound. For each technology, a productive capacity or maximum total production exists for crop i in region j, Lijyijh. The cost per acre of crop i in region j for technology h equals cijh with wxij equalling the additional cost of using MB.

One may break U.S. demand and Mexican demand for the goods into different demand functions or interpret Di(Pi) as one consumer market which we will do here. The model has been simplified so that the price of each commodity is a function of its own quantity alone. This simplifying assumption eliminates the integrability problem addressed by McCarl and Spreen (1980), and Peters and Spreen (1989). The demand for each crop, i, can be represented as

(3)

Thus to calculate the area of the demand curve of crop i, we have

(4)

We model the initial problems as

(5)

We maximize producer and consumer surplus given the constraints. The constraints are that all land used in all technologies does not exceed the land available for that crop in that region. The quantity available to sell does not exceed the sum of the acreage in that crop multiplied by the yield per acres. Methyl bromide use is the amount used per acres for each crop in each region multiplied times the number of acres using MB technology. Total MB use is the summation of MB used in all regions in all crops.

Using a linear demand curve, the inverse demand is defined as

(6)

Thus the Lagrangian is

(7)

Where the shadow prices are ij for the constraint on land lijh <= Lij (or the quasi-rents for land in region j) and for the MB. Pi is the output price. The first-order conditions then for the problem are:

(8)

We have the "shadow price" of one more unit of crop i is the market price. The first order condition for Yi reveals the inverse demand function. Since this Lagrangian does not limit the use of MB, we find the shadow price of one more unit of MB is equal to its market price, w. The first order conditions for land reveal how much land is used in each crop in each region in each technology. For example, if the revenue per acre is less than the marginal costs, no land will be in that crop in that region. Similarly, if is positive, we know all land is in production and more land in this crop would be profitable. From the last first-order condition, we can find the number of acres using technology h=0, MB use as a function of the crop price and the cost of MB:

(9)

We then can find the demand for MB:

(10)

We also can find the marginal region; a j=j such that Piyijh=cijh+mijw and

(11)

that the marginal region does not produce to maximum capacity. The alternative possibility is that there is a j=j such that Piyijh=cij0+mijw +gij and

(12)

where the demand curve intersects between regions.

Now because of the Clean Air Act, the U.S. growers can not use MB thus the new optimization problem becomes

(13)

with the first-order conditions being

(14)

We find again the inverse demand curve defines the marginal unit of crop production. Since Mexico now is the only country which can use MB, the U.S. land is constrained to alternative technologies. Once we find the land that will use MB as a function of Pi and w, we can find the demand for this input. We define the equilibrium level of MB in this case as X0.

In the third example, the ban still exists but U.S. producers are given the possibility to buy the right to use it from the Mexican growers as long as overall use, X0, does not increase. We have a technology index, h where h=0 if the growers use MB, and h=1 if they do not use MB. As before, we can use the first-order conditions to find both the demand for and supply of MB. We find that a grower will use MB if the profits earned with MB are greater than the profits earned under the alternative technology. The cost of using MB is now both the cost of MB, w, and the cost of buying the right to use it, .

(18)

Measuring Changes in Producer and Consumer Surplus

Using the model above, we obtain the price and quantity changes as well as the change in MB use. We can derive losses to consumers and to U.S. producers who are negatively effected by the ban. In addition, we can calculate the gains to Mexican producers and the U.S. producers that do not use MB. The change in nonusers producers' surplus can be estimated as

(19)

The change in consumers' surplus is

(20)

The change in current users producers' surplus can be estimated as the sum of the change in revenue, the change in costs due to change in technology, and the cost savings due to the production reduction:

(21)

The net change in domestic welfare can be obtained by summing these changes in producers' and consumers' surplus. Since the relevant policy affects both yield and cost, the change in marginal cost in the ith region becomes

(22)

where Yi/yijh denotes the percentage reduction in yield of crop i in the jth region.

We base the empirical analysis on data from 6 regions of California, 4 regions of Florida, and 3 states of Mexico (Sinaloa, Sonora, and Baja). These regions account for 90 % or more of the commodities examined in 1990. In the "market windows" that they supply, these regions are often three quarters or more of the market. There is some possibility of increased competition from other regions such as Central America, the Caribbean, and Chile, this is not being considered.

Demand curves for each commodity was obtained by calibration using an estimate of price elasticity of demand, the observed quantity produced and the price in each region. Alternative values of price elasticity will be used to examine the sensitivity of policy induced changes in welfare to the value of the elasticity of demand.

We rather focus on the commodities which are the major users of MB. Chart 3 shows the major crops affected in the U.S. and their relative percentage of MB use. The fresh sector is the principal user for tomatoes, strawberries, melons, bell peppers cucumbers, and eggplant as well as the perennial crops, peaches, nectarines, grapes, almonds and walnuts. We also will look at nursery crops and crops using post-harvest treatments of MB. Because of the seasonal nature of fresh produce, our analysis will comparing changing costs of production and yields on a crop by crop basis for the regions with concurrent shipping seasons.

The highest-profit alternative to MB was obtained through interview with growers, cooperative extension personnel and scientists. For Florida, the alternative was obtained from the report by Deepak, Spreen and VanSickle. The alternatives used are presented in Table 1 for both California and Florida. The per acre production costs and yields are presented by region by crop in Table 2. The changes following a ban on MB for California and Florida are also presented. Although for most crops, the costs increase and the yield decreases, for some crops, the costs decrease with the decrease in yields.

Mexico allocates 2.5 percent (700,000 hectares) of its 27,160,565 hectares of agricultural land to horticultural crops, with horticulture yielding over 9 percent of total agricultural production. Of the total crop land 21 percent (5,803,113 hectares) is irrigated (Instituto Nacional de Estadistica Geografia e Informatica (INEGI), 1990). Approximately 2 million hectares of the total irrigated land is concentrated in northwestern Mexico.

In the last four years, the growing regions of Mexico have suffered from various problems such as inclement weather, water shortages, overproduction, an overvalued exchange rate, limited capital, and marketing difficulties. This has resulted in a decrease in acreage in certain crops, a cost minimization mentality, and product being shifted to a growing domestic market. If the U.S. markets become more profitable, some of these problems can be overcome in the short-run and production can be increased. There continues to be acreage in several of these regions devoted to grains which could be shifted to horticultural products especially as governmental support for grains has decreased significantly and transportation infrastructure is improving. Growers in these areas by shifting technologies can also increase yields dramatically on existing acreage. Mexico continues to be advantaged by less expensive land and lower wage labor.

California supplies about 75 percent of the U.S. market for fresh strawberries. In 1990, California's 20,000 strawberry acres produced almost 500,000 tons. The total value of the California crop is $431.4 million. Mexico exports about 85% of the strawberry imports to the United States. Florida production is much smaller with about 5000 acres in strawberries.

Florida and California dominate the U.S. tomato market with Mexico providing 97.5 % of the fresh tomato imports, most from the state of Sinaloa. California produce almost 500,000 tons of fresh tomatoes on 352,000 acres. Florida produced more than 762 thousand tons. Sinaloa planted over 55,000 acres to tomatoes with Baja California cultivating almost 12,000 acres.

The preliminary results for two crops, tomatoes and strawberries show that we are able to decrease welfare costs with alternative policies but not necessarily in the U.S. producers favor. As seen in Table 3, for tomatoes, the ban results in a decrease of U.S. producer welfare of $86,574,273 and of consumer welfare of $45,621,653. Mexican growers' welfare increases $65,469,686. Florida tomato producers face much greater losses as do some regions of California. By allowing for the endogeniety of price and acreage, we find a price increase of $52.41 a ton (7.28%) with a decrease in tonnage of 164,371 (7.35%). Florida growers lose the most acreage as Mexican growers increase their production by almost 80 percent as do the non-MB using California regions. The ban results in some increase in MB use in Mexico but an overall reduction of almost 8.3 million pounds (68.8%). The MB use before the ban is approximately 12 million pounds.

Allowing for a marketable permit system that limits the MB use to the level in Mexico following the ban and gives all the property rights to the Mexican growers, we find US producer surplus decreases $63,554, 213 and consumer surplus decreases $40,087,016. These changes are smaller than under the ban; consumers are better off. Mexican growers' surplus is $49,566,788. Global welfare is greater than under the ban by almost 12 million dollars. However, U.S. producers as a whole do not benefit from the marketable permit scheme as they must also make a transfer to the Mexicans of $58,312,000. The U.S. growers revealed a willingness to pay $15.76 per pound for the right to use MB. Since the VMP of MB is so great, U.S. growers compete against one another for the right to use it, bidding the price up very high. We find that growers who buy the right are individually better off. However, quantity decreases less (6.37%), price increases less (6.34%), and the growers who benefited from the ban now do not benefit as much. Thus under the ban U.S. growers lose $86,574,273, with the marketable permit system, they "lose" both the permit cost and some surplus for a net result of -$121,866,213.

Table 4 presents the results for strawberries. The ban results in a decrease of U.S. producer welfare of $313,611,416 and of consumer welfare of $70,377,518. Mexican producers gain $89,831,695. By allowing for the endogeniety of price and acreage, we find a price increase of $132.62 a ton (18.2%) with a decrease in tonnage of 131,518 (21.2%). The ban results in some increase in MB use in Mexico but an overall reduction of almost 4.4 million pounds (81.6%). The current level of use of MB is about 5.2 million pounds. After the ban, Mexico uses slightly more than one million pounds in strawberries. Allowing for marketable permits at zero transaction costs decreases U.S. producer surplus by $294,657,2009 with consumer surplus decreasing $69,013,586. Again, global welfare improves with the marketable permit scheme. In this case, we find that U.S. producer welfare also improves. U.S. growers revealing a willingness to pay $16.67/lb for the right to use it resulting in a transfer of $16,670,000 million to Mexican growers.

Choice of domestic environmental and trade policies has critical impact on social welfare. For environmental goals, the two policies explored had an equal effect on the effectiveness of reducing overall MB use. The ban imposes greater global welfare costs with Florida regions being the most negatively effected. Consumer surplus decreases under the unilateral ban. To achieve the same environmental goal, a marketable permit system that permitted U.S. growers to purchase the right to use only the quantity of MB that Mexico uses in the post-ban world would be globally welfare improving. It however has significantly different effects on U.S. producers; some benefit from the scheme and some do not. The price of the permit is bid up high both because it decreases production costs and because the price of the commodity is between seven to twenty percent greater.

This analysis demonstrates how important the assignment of property rights is to the welfare effects of different environmental policies. By unilaterally banning the chemical, the United States is essentially giving up all property rights to it; it is setting its endowment to zero. When it wishes to induce another country to "share" its rights, U.S. growers are willing to pay a very high price to obtain a right of use. If the U.S. has initially assigned the "right to use" differently, not only would global welfare be higher than under a ban but overall U.S. producer welfare would be different under a marketable permit scheme. This demonstrates why under international agreements, developing countries fight for the right of equal endowments of the right to pollute, the right to use a certain chemical. If the U.S. growers colluded to set the price of a permit, the permit scheme would be more profitable to them but there would be excess demand for permits.

The analysis using all major crops will provide more insights on the derived demand for MB under a marketable permit system. This model also allows for more accurate measure of the degree of MB adoption in Mexico and regional shifts of commodities. In addition, more sensitivity analysis is needed to determine how robust the results are. We plan to use a range of elasticity and yield changes to determine a range of results.

MB is extremely valuable to growers in warmer climates where plant pests do not die off during the winter. MB is effective at all temperatures against a wide range of pests and weeds at all stages of their life cycle and it dissipates without leaving a detectable residue. MB is simple chemical compound (CH3Br) that is found in nature, is not patented, and is made by several manufacturers in the U.S. and abroad. Packaged for fumigation use, MB cost between $1.00-2.00 per lb. in the U.S. In Mexico, it is more expensive with estimates ranging from $1.30-5.00 per lb. Any alternative to MB is thought to increase production costs and /or to decrease marketable yields. The long-run effects of not using MB include increasing infestation of pests that survive consecutive plantings. Florida fears the rebirth of "old soil" disease. MB is often planted with chloropicrin which has excellent fungicidal properties. No single alternative exists that is capable of targeting the wide range of pests that MB/Chloropicrin are able to control.

Vapam (Metam-sodium)

Vapam is thought to be one of the best chemical alternatives for many crops by both USDA and farm advisors. It controls certain fungi, acteris, weeds, nematodes, and soil insects. It can be used for a variety of crops. Many growers do not think it is as consistently effective and its phytotoxicity limits early planting dates. Vapam is under scrutiny due to potential ground-water contamination and carries the stigma of the environmental devastation following the train crash spilling Vapam into the American river. Since Vapam moves with water, it only kills pests that come in contact with the water. Application rates for Vapam are 75-100 gallons/acre, at a cost of approximately $4.40/gallon.

Telone (1,3-D)

Telone is a compound with nematicidal properties; its is the best alternative to MB is Florida. Its use in California however, is currently very limited due to re-registration requirements (The suspension of Telone in California led to an additional 1.5 million lbs of MB to be used). In addition, the application rates permitted under current regulations are deemed insufficient to achieve desired control levels.
Telone has been found to enter the ground water and into ambient air. Growers think they need at least 60 gallons/acres to be efficient but the permitted levels are 20-30 gallons/acre.

Steam Sterilization

Although steam sterilization is relatively expensive, it is widely used in European greenhouse production. This prevents the cultivation of soil-borne pathogens, which often accompany strawberries and other crops when they are transferred from the greenhouses to the field. The process can be aided by the use of rockwool, a lightweight and easily sterilized fibrous mineral, as a growing medium. Steam applied for eight minutes to a depth of eight centimeters has been used for weed control outdoors in the Netherlands. Italian hothouse growers have successfully used solarization to combat nematodes and root rot for tomatoes, basil, peppers, carrots, and lettuce.

Baumol, William J., and Wallace E. Oates, The Theory of Environmental Policy, New York: Cambridge University Press, 1988.

Deepak, M.S., Thomas H. Spreen, and John J. VanSickle, An Analysis of the Impact of a Ban on Methyl Bromide on the U.S. Winter Fresh Vegetable Market, University of Florida, Florida, 1994.

Environmental Protection Agency, Protection of Stratospheric Ozone, Federal Register, 58(236): 65018-65082, December 10, 1993.

Florida Agricultural Statistics Service (FASS). "Vegetable Summary." Various issues, Orlando, Florida.

Hazell P. B. and R. Norton, Mathematical Programming for Economic Analysis in Agriculture. New York: MacMillian Publishing Company, 1986.

Instituto Nacional de Estadistica Geografia e Informatica (INEGI), 1990

Just, Hueth, and Andrew Schmitz (1982)

Dudley, Susan E., and Brian F. Mannix, "Comparing the Benefits and Costs of EPA's Proposed Phaseout of Methyl Bromide," Prepared for the Methyl Bromide Working Group, May 18, 1993.

Lichtenberg, Erik, Douglas D. Parker, and David Zilberman. "Marginal Analysis of Welfare Costs of Environmental Policies: The Case of Pesticide Regulation," American Journal of Agricultural Economics, November, 70(4):867-874, 1980.

Muth, Richard F., "The Derived Demand Curve for a Productive Factor and the Industry Supply Curve," Oxford Economic Paper, 16:221-34, 1964.

Mc Carl, B. A., and T. H. Spreen, " Price Endogenous Mathematical Programming Models as a Tool for Sector Analysis." American Journal of Agricultural Economics, 62:87-102, 1980.

Handy (1992)

National Agricultural Pesticide Impact Assessment Program, United States Department of Agriculture, "The Biologic and Economic Assessment of Methyl Bromide," Washington, D.C. April, 1993.

Peters, M. A., and T. H. Spreen, "Price Endogenous Mathematical Programming Models and Integrability: An Alternative Approach." American Journal Agricultural Economics. 71:1342, 1989.

Smith, S. A. and T. G. Taylor, "Production Costs for Selected Florida Vegetables, 1991-1992." Institute of Food and Agricultural Sciences. Florida Cooperative Extension Service. Circular 1064. University of Florida.

Sunding, David, Cherisa Yarkin, David Zilberman, Jerry Siebert and Alan Marco, Economic Impacts of Methyl Bromide Cancellation. Department of Agricultural and Resource Economics, University of California, Berkeley, CA 1993.

United Nations Environmental Program, Montreal Protocol on Substances that Deplete the Ozone Layer: 1994 Report of the Methyl Bromide Technical Options Committee, 1992

United Nations Environmental Program, Montreal Protocol on Substances that Deplete the Ozone Layer: 1994 Report of the Methyl Bromide Technical Options Committee, 1995

Van Sickle, J. J., E. Belibasis, D. Cantliffe, G. Thompson, and N. Oebker. "Competition in the Winter Fresh Vegetable Industry." USDA Economic Research Service, Ag. Economic Report # 691, 1994.