Analysis of Chilean Fisheries Expansion Alternatives

This dissertation investigates the expansion of the Chilean fisheries focusing on the processing and marketing activities for seven fish species that represent over 90% of the catch and for two groups of the remaining fish species. A review is made of the current condition of the Chilean fisheries, its expansion potential and some possible expansion alternatives. It is established that there exists an ample potential for the expansion of the catch and physical production of fishery commodities. A model is developed to characterize the economic system formed by the fisheries extractive and related activities and assess the profitability of investment under different expansion alternatives. This model includes (a) the random generation of a catch, that is normally distributed, for 9 fish species or groups of fish species; (b) the optimal allocation of fish harvested to 25 fish processing and 125 marketing activities, by means of a separable programming model that simulates the behavior of fish processors and distributors in the country's five fishing zones. The results of the model of the fishery economic system indicate static rates of return on investment in the different zones and determines the intertemporal internal rate of return for the investment in different zones. The major conclusion of the study is that it is economically feasible to expand the output of both processed and unprocessed fishery commodities, particularly in Zone II. The study concludes with a discussion of future directions for research, including improvement of the quality of the basic data used, and to a lesser degree refinement of the model, to improve the accuracy and reliability of the results.

model that simulates the behavior of fish processors and distributors in the country's five fishing zones.
The results of the model of the fishery economic system indicate static rates of return on investment in the different zones and determines the intertemporal internal rate of return for the investment in different zones.
The major conclusion of the study is that it is economically feasible to expand the output of both processed and unprocessed fishery commodities, particularly in Zone II.
The study concludes with a discussion of future directions for research, including improvement of the quality of the basic data used, and to a lesser degree refinement of the model, to improve the accuracy and reliability of the results.
To my parents, my wife and our children Table   1. 2.
3.   despite the access to marine food resources, a significant part of the population suffers from some degree of malnutrition.

LIST OF TABLES
-1 An indication of the evolution of the primary fishing activities in Chile is presented in Figure 1 and Table 1.  The fact that expansion has not come about is used as grounds to contend that either the physical potential does not really exist or is not by itself enough to make the expansion of the fisheries sector as attractive as the expansion of other sectors of the economy.
The Chilean government, accepting IFOP' s estimates of available resources, has postulated the following objectives for fisheries development: 2 (1) to make the fisheries (a) a source of sustained food re serves for the country (b) an important factor of the national economy that may permit the satisfaction of the basic needs of a constantly increasing (domestic) population and may contribute to the solution of the world food crisis.
Towards this end, the utilization of current catches must be -5 improved and subsequently the extraction of currently exploited species increased rationally in the cases where this is possible and the exploitation of potential resources and culture must be encouraged.
(2) to expand the per capita consumption of fishery products as a solution to the food and nutrition problem of the (domestic) population, through campaigns to promote the consumption of these products, in fresh or processed form.
(3) to make a positive impact on the balance of trade, generating a market for Chilean fishery products of quality standards that are internationally accepted.
( 4) to generate significant employment opportunities, through the establishment of units linked to the marine and continental fisheries that are effectively productive.
The State, through its pertinent agencies, will have the responsibility to (a) insure a rational exploitation of the fishery resources, (b) direct and supervise the development of the sector, (c) plan fi.sherie s economic activity giving due consideration to the right of private activity but will not accept that the renewable natural resources be irrationally exploited.
The above stated objectives may be summarized as: 1.
The maximization of the sustained production of food from fishery resources, both for internal consumption as -6 well as exports; 2. The maximization of productive employment opportunities for the population.
These two goals are different and are not normally reached simultaneously. Under the current condition of the Chilean economy, however, there appears to exist a relative abundance of labor, making the second objective not inconsistent with the fir st. The relative abundance of natural resources on the other hand permits the achievement of both objectives at a relatively low opportunity cost to the national economy.
Clearly though, even after accepting IFOP' s assessment of potential physical yields, there remain questions to be answered concerning the economic feasibility of expanding the production of fishery commodities, and on the implications of that expansion.
This study will seek to answer two of the most important among these questions, i.e., (1) Whether there exists, m economic terms, a potential for fisheries development in Chile, and (2) What are some of the implications of the expansion of the fisheries output.
To assess the economic potential, the fisheries economic system will be assumed to behave as if it were maximizing profits, knowing fully well it may not be entirely so, as in reality the system ts mixed, and the net revenues as soc iated with that behavior will be evaluated.
The internal rate of return on investment and the net benefit -7 to the economy will be used to evaluate alternative approaches to the expansion of the sector. b) the nature of the marketing system and the alternative adjustments required from it to absorb the increased catch.
The objectives of this Chapter are to describe the Chilean fisheries and related activities, to characterize the operation of these activities in an economic system, and to outline alternatives for the expansion of the contribution of the sector to the national economy.
1. The Chilean Fishery Re sources.
Chile has over 2, 500 miles of mainland sea coast, and claims a 200-mile fishery zone. Although the narrowness of the continental shelf off the Chilean coast limits in part the productivLty of the primary fishing activity, it is still considerable enough to place Chile among the ten major fishing nations of the world.
The vast physical extent to the Chilean fishing grounds can be divided into six zones. These are shown in Figure 2. The main conventional resources in this zone are the pelagic species --anchoveta (engraulins ringen anchovy ), jurel (trachurus murphyi -mackerel), and sardina ( strangomera benticki, sardin ops sag ax rnusica --pilchard) .
Potential resources yet unexploited include mainly the agujilla species (scomberesox stolatus -saury), also a pelagic fish. The anchoveta and sardina re sources of this zone have been heavily exploited over the pa st ten years. -10 The second fishery zone runs from south of the port of Taltal  -17 resources, most of which goes to reduction, the artisanal fishermen fish for the finer species of white fish, which fetch the best prices in the fresh market.
The total catch during 1974, with its percent distribution for each species or group of species that comes from each subsector is shown in Table 4 .
The average catch of the most important species of fish and shellfish during the 1970-1973 period is shown in Table 5 .

3.
Marketing and Distribution of Fishery Products.
Fish may be marketed for direct human consumption in fresh or preserved form (canned, frozen, dry salted, smoked, etc.) or for industrial consumption in the form of semi-processed goods to be used in the production of other goods.
A breakdown of the Chilean fishery production during 1970 to 1973 is shown in Table 6.

Fresh Fish Marketing
The marketing channels for fresh fish are shown in Figure 3.
Over two-thirds of the fish and shellfish marketed in fresh form is supplied by artisanal fishermen. On its way to the consumer this fraction of the supply may pass directly from the fisherman to the household, or through one or two middlemen in a local market to three or more in the case of a great demand center where passage through a large wholesale market (terminal) is required.
Different types of fish merchants buy at the terminals and eventually sell to the consumer either in established retail outlets *Utilizing the waste from the production of main products.    Note: This subsystem will not be treated in the analysis that is performed on the subsequent chapters.

Constraints on Fisheries Performance
The primary constraints that may limit the output of the fisheries is naturally the availability of fishery resources.
Considering the fact that these are renewable, the size of fish populations and their reproduction rate determine the maximum sustainable yield for each species, i.e. max·imum amount that can be extracted during a period (usually stated in annual terms) without decreasing the size and yield of the population for the following periods. This constraint is one that rational exploitation of the fisheries would have to adhere to in order to maximize long run yield. Within these constraints, the extractive capacity of the fleet and the landings infrastructure provide additional constraints.
Given the availability of fishery resources, another set of constraints appear in the capacity to transfer and/ or distribute fishery commodities. These capacities are in the short run fixed, but over long periods of time may be changed in response to shifts m the demand for these commodities. Both in the short and long run, however, the availability of other production factors may become another constraining factor.
Finally, the size of markets may constitute an additional constraint.

Expansion Alternatives for the Chilean Fisheries
The principal sources for the expansion of the con tr i but ion of the Chilean fisheries to the economy are : per year.
A breakdown of the current and potential catch by species is given in Table 7 . It shows that the possibilities of expansion of the catch from conventional resources may come mainly from: jurel, which could yield an additional 110, 000 m. t. / yr. sierra (thyrsites atun -mackerel) merluza and other white fish cholgas (aulacomya ater-mussels) choritos (mytilus edulis -mussels)  ?9 In addition, by improved management of currently over- The resulting average is some 200 thousand metric tons below the estimated sustainable yield of the resource under more rational exploitation.

-31
The state of the art m this field is, however, at a very early stage and much biological re search remains to be done before £inane ial analysis of commercial undertakings can be performed, and the necessary investment is attracted.
Estimates of the potential yield from shellfish culture, (mainly mussels) are made at 81,600 metric tons per year.
Research on the costs and benefits associated with this culture indicate an annual rate of return of 18. 6 % vis a vis the 1 7. 3 % that the exploi-1 tation of natural choritos yields.

Utilization of Fishery Resources
Fishery resources may be used to feed people or animals and also to produce goods that satisfy other human needs. Among the latter, fish oil is used in the production of many non-food commodities, and other fish parts have been used to produce goods that are not food. By far the most important use for fishery resources is as food, either for direct human consumption or, via another step in the food chain, as cattle or poultry feed to produce other animal food for human consumption, or just to feed animals that render services that satisfy other human needs. On one hand, given the organic composition of each species, a fish may yield different proportions of their landed weight in different products.
As an example, jurel may give around 19. 8 % of fishmeal and 3. 5 % of oil in the reduction process; 36.1 % of fillet; 47.8% of minced flesh; 44% in smoked goods; 40.2% in canned goods. The organic composition also determines organoleptic character is tics (the way each product looks, smells and tastes) for the different products that can be made from each species of f ish.
The other set of forces is provided by consumers' tastes and is manifested through their preferences and willingness to pay for each product.
The interaction of these forces over a long period of time results in an infrastructure for the extraction and transformation of fishery re sources. This infrastructure provides a set of 11 capacity 11 constraints that limit in the short run the amounts of each product that may be produced, and cannot be expanded from one period to the next, but may be expanded by efforts that are sustained over several periods.
The suitability of some of the species under conventional exploitation for different forms of marketing is shown in Table 8 .
In the case of the unexploited species it would appear that the agujilla may make a good canned product, while the merluza de cola seems to have most promise in the form of frozen minced fish flesh, or as input in the production of fish protein concentrate.

Improvement in the Utilization of Currently Extracted Resources
The average utilization of the Chilean fishery resources during 1970-73 is shown in Table 9 . It can be seen from this Table that during the period observed, 99% of the anchoveta, 39% of the merluza, 93% of the sardinas and 76% of the jurel were directed to the production of fishmeal.
The average contribution to the gross value of the production during the same period of one metric ton of each of these species in the different forms that they may be marketed is shown in This product is not considered in this study, as its production and marketing do not appear yet to be commercially feasible, unless in the context of governmental spending programs. For further details see references 3, and 53 .
1 Table 9 Utilization  if output of fishery commodities is significantly increased, the impact of this expansion on price may be such that price may drop beyond the point needed for the producer to meet costs.
Beyond some level of sales, then, this makes the domestic market prospects look unpromising for Chilean producers. In the exports markets, however, this is not the case as the entire potential Chilean production is not likely to affect prices.
In the light of the above discussion, the Chilean fish processors have not been irrational in making their decisions. They The utilization of fishery resources is not itself a controllable variable, but rather the result from the interaction of the different components of the economic system de scribed in Section 4, each of them formed by numbers of variables.
To change the current utilization of Chilean fishery resources changes must occur fir st in the variables that determine the economic system, and those changes must sustain themselves long enough to permit the adjustment of the processing and marketing infra structure.

Alternatives for Expansion
The three sources for the expansion of the Chilean fisheries discussed m the preceding subsections do not necessarily imply exclusive alternatives. They do, however, imply alternatives in 1 cORFO's assessment of capacity utilization in 1971 indicated utilization rates of 34%, 13% and 15% for reduction, freezing and canning capacities respectively 15 A comparison of registered capacity during 2 00 shifts against actual average production in 1970 -7 3 indicates utilization rates of 50, 20 and 25% for the same facilities. The basic alternatives which will be analyzed m the following chapters are: Alternative 1 -Efforts are made to utilize the expanded catch with the currently available facilities, expanding the fresh market opportunities to its limit and, when necessary, fishmeal production capacity until all the expected catch can be processed.
Alternative 2 -The fresh market is expanded as in Alternative 1, while processing capacity for the production of -41 fully elaborated frozen and canned fishery commodities is expanded as required to exploit currently underutilized species (jurel, sardinas) and unutilized species (agujilla, merluza de cola), and fishmeal capacity is expanded only as required to process by-products from the production of frozen and canned commodities.
The analysis of these alternatives will be carried out by focusing on the following nine fish species or group of fish species: 1. anchoveta 5. other pelagic fish species 2. jurel -sierra 6. merluza There are, in sum, a number of factors that determine fish concentration and catchability that are still beyond human control.
If we add to this the fact that fish are not easily seen and that fish schools are mobile, we can expect the results of the fishing activity to be highly variable.
In analyzing economic activities that are generated by the availability of a raw material which is so variable it is imperative to account for this variability -45 if we des ire to capture to a meaningful extent the dee is ion environment.
One way to accomplish this is the generation of a variable catch that deviates randomly from its expected value.
Section 2 --Simulation of the Extractive Activity The procedure to follow to simulate the extractive activities includes:  Table .11, along with their standard deviations.
If we assume normality m the distribution of the deviations around the mean catch for each species, and independence among the different fish species, we can simulate extractive activity by generating a normally distributed random variable C(I), i.e. the national catch of a given species in a given year.
The fir st assumption, i.e. , normality in the distribution of the deviations from the mean may not hold true when account is made of long run trends in the exploitation of each re source. In the short run, however, this assumption will generally hold true. These shares are shown in Table 12 for the actual catch during the period indicated, along with the shares of potential catch based on the estimated concentrations of fish populations in each zone, towards which the distribution of the catch in the long run will be changed.

I
1 For the purposes of this analysis fishery zones V and VI have been merged. if the processors are attempting to maximize profits, within the constraints imposed on their possibilities of choice by the production function, availability of inputs and the capacity of their installations (in the short run), they will buy inputs until the cost of the last unit equals the revenue associated with its contribution to production, and will produce commodities to the point where the cost of the last unit produced equals the revenue it generates. As there is no evidence to suggest that Chilean fish processors are not profit maximizers the assumption is made that that is, in effect, their objective.

-50
A relatively simply analytical tool that permits the examination of this type of behavior is linear programming.
Its basic structure is the following for the problem under conside ration.

Estimation of Net Revenue Functions
The net revenue function that results from the combination of cost and demand functions has been separated into five segments for each of the commodities under consideration.
The net revenue and as soc iated output at the extreme points of each segment are listed for these commodities in Table 13.
-55     (1) Based on the utilization of required facilities in 200 eight-hour shifts during the year.
(2) Base for projection of fresh fish consumption and the production capacity required to match it.   Section 5 -Evaluation of Performance The optimal solution to the linear programming model gives for each year and each zone, the optimal outputs for each commodity, the net returns as soc iated with them and the scarcity values for each capacity or resource that constrains production.
The net returns to the producer are an adequate indicator of private performances in the production or marketing of each commodity. Aggregation of zonal results leads to yearly performance, that can be stated in relation to investment. Aggregation of yearly performance leads to long run performance that can be stated in terms of internal rates of return on investment.
Social performance may be partially evaluated on the basis of output, associated net returns and contribution to employment.
A flow chart of the simulation of the operation of the Chilean fisheries and related activity is shown in Figure 7.

Initial and Projected Conditions
The initial and projected conditions for the operation of the processing subsystem are given by a) the current and potential expected catch, shown in Table   11, page 46, that determine the level of the re source constraints at a national level. b) the current and potential zonal catch distribution shown in For both alternatives the investment required to expand capacity is made at the beginning of the third year of the simulation, and the capacity expansion is effected at the beginning of the fifth period.

Short-Run Performance
The indicators of yearly performance are a) output and associated prices for each commodity b) net revenues and static return on investment c) slack available and associated shadow prices for the Source: !FOP, Secci6n Estudios Econ6micos for current investment in processing facilities and linear projections for expanded capacities.
In the case of extraction for fresh marketing, assumed to be carried out by artisanal fishermen, an average investment of US$650 per fisherman, and a productivity of 6. 5 metric tons per year is initially assumed and projected linearly for the estimation of investment required for expanded output. The optimal yearly output of each commodity is shown in      3.

Long-run Performance
The indicators of long-run performance are a) gross output b) employment c) internal rate of return on investment The total outputs and levels of employment associated with the ten year operation of the sector under the two alternatives are shown in Table 20, while a comparison of the initial and final output and employment levels are shown in Table 21, page 78 The internal rate of return on the investment made on each zone is shown in Table 22, page 79 .  Notes · (1) as function of capacity.

Output Behavior
The outputs under both alternatives do not differ in the fir st five periods. This is to be expected as the initial conditions are the same. Starting on period 5 and through to the end of the simulation differences exist in the composition of the output with a higher production of fishmeal for Alternative A and higher production of frozen and canned products for Alternative B. As the extract ion of fishery re sources is increased nearly 100% (year 5 to the end of simulation) average output of fishery commodities increases by approximately 80% in Alternative A and 100% for Alternative B, with the composition of average output broken down as shown in Table 23, page 81 As can be seen, the effect that the increase in capacity has on output is important but it is not the most important one.
-80 -81 What really allows the increased production is the increased availability of fishery resources (nearly 200% of initial availability) particularly in zones where they were a constraining element under ·nitial conditions. This is apparent from the fact that under initial conditions simulated capacity utilization is quite low, except for the canning process.
The relation between the fresh and canned fish production in the initial years (24, 350 tons to 36, 190 tons) does not correspond to the current situation of the Chilean fisheries, (a v erage fresh production is nearly 60, 000, while average canned production is nearly 12, 000). This is a sign that the relation between the prices used for these commodities over -values canned production, a clear indication of the weakness of the pr ice information utilized for those commodities.

Pr ice Behavior
The prices associated with the output of the different commodities produced, shown in It is clear that the pr ice ratio between canned and fresh commodities does not correspond to reality, and implies a composition of consumption that over states canned products by a factor of 3.
Furthermore, the price elasticity of demand for canned products, implied by the prices on Table 24 is another indication of the limited success that estimation efforts have had in these commodities. This is largely due to the quality of the aggregated information utilized which does not recognize the variety of canned fishery commodities.  0  167  0  167  167  167  2,030  166  49  283  174  283  53  283  0  283  3,432  690  3,554  688  3,433  690  3,093

Interpretation of the Shadow Prices
The shadow prices indicate the scarcity value of the constraints they are associated with, showing the potential contribution that a unitary relaxation of them would make to the objective function value. mentary slackness property that falls out from the relationship between the primal of the maximization problem under con s ideration and its dual is met throughout. This property requires that the product of the value of each slack variable and its associated shadow pr ice be zero.
In practical terms it implies that if a constraint is not binding, i.e., if it has unused slack, its potential contribution to the objective function value has to be zero.
If on t he other hand, all the resource or capacity is used up, that is the value of the slack is zero, its potential contribution to the value of the objective function, given by the shadow price, must be equal to or greater than zero.
Accordingly, the constrain ts the relaxation of which would con- -88 The apparent attractiveness of the canning process as suggested by the high shadow prices must be taken with some reservation, as this is due to the high price used for canned products in relation to the other commodities, particularly to fresh products. If as indicated earlier the relative prices are biased towards canning, this suggest ion could prove quite misleading. The validation of the prices utilized is an important pre-requirement to the application of the model's results since the accuracy of the signals given by the shadow pr ices will not be better than the accuracy of the prices used to run the model.

Long-Run Results
The simulated operation of the fisheries economic system during a ten year period yields results, shown in Table 20

Sensitivity Analysis Results
The sensitivity As in the case of single objective functions, the resulting composition of output shows an output ratio between canned and fresh products that is much greater than the ratio encountered in the operation of the real 1 ife system.

Accuracy and Reliability of Results
The application of the separable programming model to the simulation of the economic system formed by the fisheries and related activities provides a reasonably close replication to its real life operation, but has not yet become an accurate tool for predicting the behavior of the system. Its -main wekness has been ' I -92 its failure to reflect real life output composition. This is due ma inly to the quality of the input information, particularly the prices of canned commodities and their sensitivity to changes rn output, and to a lesser extent production costs. Improvement of the input data and validation of results constitute an iterative fine tuning process that is required for the improvement of the results from the application of this model.

CHAPTER IV
CONCLUSIONS AND POLICY RAMIFICATIONS 1.

Summary of Results
The most important conclusion from the first chapter is that there exists and important potential for the expansion of the physical production of the Chilean fisheries. Another conclusion -9 3 is that this physical potential has not been realized , in part due to lack of information on the profit potential of investment in such activities, and in part due to lack of some services (usually of governmental provision) and infrastructure required for the successful expansion of productive activities.
The simulated operation of the fisheries economic system has given the following results: Of these conditions the most weakly met i.s d), therefore these results must be considered with much reservation.
Their sensitivity to price relations requires that more extensive research be done on the demand and cost function used to run the model.
In its present state the results from the application of this model can be st be regarded as providing useful information on the re la--96 tive advantage of investment in different zones. However, improvements in the cost and price data utilized to run the model are required before it can be used as a tool for assessing the relative advantage of investment in the production of different commodities or predicting the system's performance.

Policy Ramifications
Given the national objectives of fishery exploitation and in the light of the results of the analysis performed, the following recommendations would seem in order : given the availability of marine fish re sources , efforts should be encouraged to exploit them rationally to their full potential, providing the fishery management and re search required to generate such activity and prevent overexploitation.
-along with the expansion of the supply of fishery commodities efforts should be made to shift the demands for these commadities in the domestic ma-rket .
-research on the costs and benefits associated with the governmental investment in the prov is ion of services and infrastructure required for the production and marketing of fishery commodities should be carried out.
-given the zonal availability of processing facilities and the concentration of fishery resources, services such as ports and transportation infrastructure should be made available so as to encourage the investment in processing facilities where required.
-97 -marketing services should be provided to ensure the timely dissemination of information on prices for both domestic and export commodities. Pricing of export commodities (fishmeal and frozen products) has been parametric.
The reader is warned, therefore, about the questionnable accuracy of the estimated costs and prices, which are determinant for the composition of the optimal outputs, and for the profitability of the different investment patterns. Also troublesome for the latter is the estimation of the current and required investments.
There is a strong possibility that the estimations of current investments are overvalued, and that needed investment is undervalued, making returns on investment appear more attractive than what they are really.

Model Structure
The structure of the separable programming model as utilized m this study implies the following assumptions: -productive activities in each zone are represented by a single profit maximizing producer, that has the option to sell fish fresh, fno zen or canned, or to reduce to fishmeal the catch available to him. Processors in all zones face the same demand functions for commodities sold in the domestic market, or parametric prices for export commodities.
-There is one domestic market where fresh and canned -99 products are consumed and one export market that buys the outputs of fishmeal and oil and frozen products.
Abstraction is made of spat i a 1 relationships between supply and demand centers, although this is partially considered in the sensitivity analysis.
-there is a single activity in the extraction of the fishery resources and their transformation into each of the 25 fishery commodities considered.

Extension of Economic Research on the Production of Fishery Commodities in Chile
After some adjustment in the input data (cost and prices) the model presented in this study can give results that are reasonably useful for governmental policy and private investment decision making.
To make this tool a more accurate simulator of economic performance and improve its reliability for the analysis of investment, the following improvements would be required: -performance in different periods could be linked in a sequential manner so that the results of the operation in one period be made use of in the next period, e.g., making investment in one period function of the shadow prices that fall out from the optimal solution of the programming problem in the previous period -consideration should be made of other inputs or re sources utilized in the production of fishery commodities, such as labor, that may or could become scarce on a regional context.