Markets and rent dissipation in regulated open access fisheries

ARTICLE IN PRESS

Journal of Environmental Economics and Management 49 (2005) 381–404 www.elsevier.com/locate/jeem

Markets and rent dissipation in regulated open access fisheries$ Frances R. Homansa, James E. Wilenb, b

a Department of Applied Economics, University of Minnesota, St. Paul, MN, USA Department of Agricultural and Resource Economics, University of California at Davis, USA

Received 13 April 2000; received in revised form 6 August 2003; accepted 3 December 2003 Available online 23 November 2004

Abstract Using a model of regulated open access resource use with markets, we illustrate the potential complexity of interactions between markets, product quality, excess effort, and regulatory behavior in fisheries. Our model assumes two product types, one of which is processed and storable. The model describes the equilibrium that results when effort responds to open access incentives and when regulators respond to effort growth by mitigating its potential harmful effects on biomass safety. We use numerical simulations to demonstrate a mechanism by which market growth leads to the diversion of raw inputs into the inherently inferior market. The result is a scenario in which rents are dissipated not only because excessive inputs raise costs, but also because inferior product types reduce revenues. We conclude with an illustrative decomposition of rent gains into revenue and cost savings gains from rationalizing a hypothetical fishery, demonstrating the potential significance of market-side rents. r 2004 Elsevier Inc. All rights reserved. Keywords: Regulated open access; Rent dissipation; Markets

$

The genesis of this paper was originally presented at the 40th Anniversary Conference honoring the Norwegian School of Economics and Business’ Institute of Fisheries Economics in 1993. Since then, helpful comments by Jim Anderson, Frank Asche, Lee Anderson, Gardner Brown, Gordon Munro, Ragnar Tveteras, and Kjell Salvannes have improved the paper. Thanks also go to the Minnesota Agricultural Experiment station and the Giannini Foundation for funding support, and to the Associate Editor and anonymous reviewers for helpful comments. Corresponding author. E-mail address: [email protected] (J.E. Wilen). 0095-0696/$ - see front matter r 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.jeem.2003.12.008

ARTICLE IN PRESS 382

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

1. Introduction It has been 50 years since Gordon’s [20] seminal paper on open access exploitation was published, a paper that has proven one of the most enduring and widely cited in natural resource economics. Gordon used a simple input-demand based model to lay out the cause, the process, and the consequences of overexploitation of fisheries resources. He identified the fundamental cause of the problem as the absence of resource ownership, or ill-defined property rights. The result of ill-defined property rights is a process whereby excessive inputs are attracted by uncollected resource rents. Gordon saw several implications of this process of excessive entry, including inefficient spatial distribution of effort, resource degradation, and the dissipation of potential rents in fisheries. Gordon’s paper has been instrumental in framing the way economists have come to view the ‘‘fisheries problem’’ and its solution over much of the past half century. By identifying the absence of property rights as the cause of the problem, Gordon drew attention to potential solutions that involve creating real or proxy property rights to the productivity of fisheries resources. The kinds of institutions economists have recommended are several, but all have in common the notion of establishing some kind of dedicated use privilege defined in terms of specific quantities of allowable harvest. The most common rights-based institutions in practice involve individual transferable quotas (ITQs) whereby fishermen hold a right to a share of a biologically determined total allowable catch (TAC). The recent past has also seen development of community development quotas and harvester cooperatives that grant fixed quota rights to groups rather than individuals. Dramatic changes in fisheries have occurred under rights-based management, and there is a growing list of programs beginning with the first ITQ systems that were implemented in Iceland and New Zealand in the early 1980s. To date there are over 60 of these programs in existence [32] and important empirical evidence is beginning to accumulate about how large foregone rents have been. They appear to be large. A good empirical rule of thumb is that annual lease prices for a pound of traded quota in ITQ markets are generally at least 50% (and sometimes as much as 80%) of contemporaneous exvessel prices, with quota sale prices averaging 10–12 times the lease prices. But one of the biggest surprises emerging in the empirical evidence from rationalizing fisheries is that the first noticeable changes often occur on the market side of the ledger via higher exvessel prices rather than on the cost side.1 We highlight this as a ‘‘surprise’’ because we would 1

For example, Casey et al. [7] report that in the British Columbia halibut fishery, the season lengthened from 5 days to 8 months after ITQs. As a result, wholesalers were able to sell 94% fresh after ITQs compared with 42% before, and exvessel prices rose over 50% in the first couple of years. Geen and Nayar [17], reporting on the Australian Southern Bluefin Tuna, conclude that prices doubled from about AUS$988 to AUS$2000 per metric ton within three years of the ITQ program’s inception. This occurred as fishermen switched from targeting small tuna in nearshore regions to larger more mature tuna off the continental shelf. Before ITQs only 13% of tuna were greater than 15 kg; 2 years after ITQs the percentage had risen above 35%. In Iceland, Arnason [3] reports that a 1985 study conducted soon after ITQs were adopted showed benefits from reduced effort worth US$14 million and US$6 million from improved product quality. Gauvin et al. [16] reported two years after the wreckfish fishery converted to ITQs that prices had risen to about $1.85 per pound with considerably more stability from pre-ITQ levels fluctuating between $0.90–$1.55. Their interviews with processors confirmed that this was due to the elimination of supply gluts and the practice of dumping raw product into less valuable frozen and discount supermarket outlets.

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

383

contend that, prior to the implementation of rights-based regulations, many fisheries economists anticipated that rationalization would be reflected mostly through input reductions and cost savings rather than revenue enhancement. These anticipations reflected the influence of Gordon’s depiction of the process of open access exploitation as one driven by excessive inputs, a depiction exemplified in the oft-used characterization of open access fisheries as having ‘‘too many boats chasing too few fish.’’ The initial improvements in product attributes and revenue increases witnessed soon after rationalization in many fisheries call for some explanation not found in Gordon’s original inputbased story. Aside from questions about how large these gains associated with market side changes are (and how they compare with input cost savings gains), a more fundamental question raised by these observations is: what mechanisms led to the output quality reduction and product attribute distortion during rent dissipation in the first place? This is an important question because what we are witnessing as new rents are generated from fisheries rationalization should be a sort of mirror image of the original rent dissipation process. Understanding the texture of rent generation under the creation of new rights based systems may thus give us some richer insights into the mechanisms behind dissipation that Gordon brilliantly introduced 50 years ago. We believe that the key to understanding the observed market gains in recently rationalized fisheries lies with understanding how the regulatory sector interacts with the market. Our contention is that the character of modern fisheries on both its market and production side is heavily influenced by the nature of regulations and the manner in which these unfold over time. Importantly, regulations are influenced by, and have impacts on, product attributes and quality on the market side of the ledger as well as on inputs and efficiency on the cost side of the ledger. In this paper, we develop this theme by adding a stylized market to a model of joint industry/ regulator interaction. The aim of the paper is to better understand the interrelationships between markets, inputs, and the rent dissipation process, in order to explain recent events and to understand forces that have led to the preconditions that we see prior to rationalization. Our model enables us to accomplish two tasks. The first is that we are able to characterize how exogenous changes in the market affect the regulated equilibrium. For example, we show how income- and population-induced growth of markets may influence the level of over-capitalization and the stringency of regulatory instruments in regulated fisheries. Second, we characterize how the institutional structure of regulated open access fisheries may, itself, affect the development and emergence of particular market niches. Hence we are able to demonstrate, for example, mechanisms by which most of the raw product in a regulated open access fishery might be diverted into lower valued product markets rather than higher valued markets. This gives us some new insights into how the market-side revenue increases that are currently being unlocked by ITQs were originally dissipated.2 2

Rents are generally defined as the difference between revenues and input opportunity costs. From the open access status quo, rents can be generated both by saving on input costs through a reduction in inputs and by earning additional revenues through an increase in the exvessel price. In an early draft we used the term ‘‘market-side rents’’ to refer to these additional revenues. As a reviewer notes, however, an increase in consumer demand by itself cannot lead to rent generation: the production externality is the reason that rents are dissipated and so must be resolved for rents to be generated. In this paper we are emphasizing the idea that market as well as the production structure can respond to a change in the regulatory structure through the mechanisms we outline, and these revenue increases add to conventionally considered cost savings in the calculation of rent gains from rationalization.

ARTICLE IN PRESS 384

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

In the next section, we develop a model of exvessel price determination in a regulated open access fishery with two product types. The equilibrium in this setting is complicated for several reasons. First, the harvesting season length (and conversely the length of the processed (frozen) product marketing period) is determined endogenously. Regulators set season lengths primarily to ensure stock safety, but based on the expected effort level committed by the industry. The effort level chosen by industry, in turn, is based partly on expected exvessel prices, which are determined by the manner in which the whole marketing year is divided up between a fresh market and a frozen market period. Exvessel prices must reflect relative demand conditions for fresh fish during the fresh market period and also derived demand for additions to inventories. The derived demand for inventories must reflect expectations of the market for processed product to be supplied out of storage for the remainder of the marketing year. Hence in the final analysis, total effort, the harvesting season length, the period over which fresh fish are sold, the period over which frozen fish are sold, exvessel price paths, and inventory market price paths are all determined endogenously. After we develop the theoretical model, we explore its implications by calibrating the model with parameters drawn from the North Pacific Halibut fishery. We resort to numerical analysis in the end because the system developed here is too complicated to simply examine using analytical methods such as comparative statics analysis. The main goal of the numerical exercise is to explore how exogenous forces (such as population and income growth) that influence fisheries markets might influence the structure of any generic regulated open access fishery over time, including input use, regulations chosen, and product mix. We then use the simulation to depict rationalization of a fishery, and we decompose hypothetical rent gains into those associated with product form and market changes and those associated with excessive inputs.3

2. A model of markets in a regulated open access fishery In this model, the fishing industry is assumed to commit to a fixed level of harvesting effort, E, for the upcoming season. The regulatory authority sets the fishing season length, t, in order to ensure that the total harvest over the season does not exceed a predetermined TAC. Fish are per pound of landed throughout the fishing season and fishermen are paid an exvessel price PEV t whole fish. As the raw product is landed, it is allocated into either the fresh wholesale market or into the frozen wholesale market or both. Inventory accumulation for the frozen market may start at the outset of the harvest season (time 0), or may be delayed until later in the season. Once the harvest season is over (at time t), the inventory can be sold as frozen fish until the end of the inventory dissipation period, time T. The beginning and ending dates of the marketing year (0 and T) are assumed to be exogenous, but the length of the inventory accumulation period and the length of the harvest season are endogenously determined. Fishing effort (E) and the wholesale price that prevails in the frozen market at the end of the marketing period (PZ T ) are also determined within the model. 3

We should emphasize that even though our simulation mirrors qualitative properties of the North Pacific Halibut fishery, we are not conducting a formal hypothesis test of that fishery per se. Instead, our intent is to simulate market/ regulatory interactions in a generic fishery, with parameters chosen to scale realistically to a real fishery rather than completely ad hoc parameters.

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

385

2.1. Industry behavior Firms are assumed to commit to a constant level of fishing effort, E, for the season. Here, E is defined as a generic input flow that is employed in a standard density-dependent Shaefer production function along with biomass, X , to determine the harvest rate, H:4 H t ¼ qEX t : Biomass falls throughout the season due to harvest:5 X_ t ¼ qEX t : Integrating this differential equation, and noting that biomass at the beginning of the season is X0, we get the level of biomass at any instant: X t ¼ X 0 eqEt . The harvest rate at any date t will thus be: H t ¼ qEX 0 eqEt ;

ð1Þ

which also declines throughout the season along with biomass. Because of open access incentives, we assume that entry proceeds until total rents anticipated over the season fall to zero. The expression for discounted seasonal rents is Z t p¼ ert ½PEV t H t  vE dt  fE; 0

where PEV t is the exvessel price at time t, v is the instantaneous cost of employing effort and f is the fixed per unit cost of effort that is paid each season. We have assumed a simplified production technology for this model. Harvesting effort, E, is a flow that is produced in a fixed proportions technology   with capital service flows, S, and variable factors of production, R, as inputs: E ¼ min gS; jR . Capital service flows are, in turn, produced in proportion to the amount of capital, K : S ¼ aK: Therefore, the industry’s choice of an effort level implies a choice of capital stock along with a rate of variable input use. We assume that capital is specialized to this fishery with no alternative uses. We then interpret f as a composite fixed cost term that embodies both the annualized rental and any other fixed annual set up costs of capital. If c is the annual rental and set up cost of capital, then the annual fixed cost component of effort, f , equals c=ðgaÞ. Similarly, if w is the cost of the variable factor of production, R, then v equals w=j. Variable costs are incurred throughout the season and depend upon season length, whereas capital costs are independent of season length.6 4

The biological part of this model of within-season harvesting was first developed by Beverton and Holt [5]. An early attempt to develop an economic model of a within-season fishery is in Bradley [6], with subsequent work by Clark [8,9] and Clark et al. [10] attempting generalizing in a dynamic bioeconomic framework. The regulated open access framework is from Homans and Wilen [26]. 5 For analytical simplicity, we ignore natural mortality, although the qualitative conclusions would not be affected by including it. In addition, this model is a within-season model and hence we ignore between-season dynamics. Homans and Wilen [26] discuss the approach path in a fully dynamic regulated open access model. 6 The fixed cost term must apply to a stock, and effort, E, is an input flow. The parameter f is therefore reflective of, but not identical to, per unit capital costs. We thank the reviewer for suggesting this discussion that clarifies our assumptions about the production technology. These are, of course, simplifying assumptions that permit us to focus on the market side. There is a substantial literature that examines the complexity of fishing technology. See Kirkley and Squires [27] for a recent review.

ARTICLE IN PRESS 386

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

The rent dissipating level of effort is determined by the integral equation: Z t qEt ert ½PEV  vE dt  fE ¼ 0: t qEX 0 e

ð2Þ

0

Because effort appears in both exponential and linear terms, this cannot be solved explicitly for the rent dissipating effort level. The equilibrium level of effort depends implicitly on the season length selected by regulators, the path of the exvessel price, and the cost and production function parameters. 2.2. Regulator behavior We assume that regulators set the season length to achieve a given total allowable catch (TAC) target.7 The season length will depend upon the level of effort anticipated, since regulators know that total harvest over the season is an increasing function of effort: Z t H t dt ¼ X 0 ð1  eqEt Þ: ð3Þ 0

We assume regulators, act as if they are solving (3) for the desired season length by setting the seasonal harvest equal to the targeted TAC to get:   1 X0 ln : ð4Þ t¼ qE X 0  TAC This expression for the regulators’ choice of policy instrument (Eq. (4)) together with the rent dissipation equation (2) comprises the basic regulated open access model [26], with the simple addition of a time varying path for exvessel prices. In the sections below, we close the model by solving for the equilibrium exvessel price path under a variety of assumptions about the market. 2.3. The market—fresh only Assume, for the time being, that the only market outlet for raw fish is a fresh wholesale market. In this case, fishermen deliver raw fish to a daily market, where they are paid an exvessel price PEV t : The raw fish are then immediately resold, after a moderate amount of conversion, into a fresh wholesale market. Before proceeding, we need to reconcile the units of measurement in both the exvessel and wholesale markets. Generally, exvessel prices are paid for fish at the dock and are paid per pound of dressed (headed and gutted) fish. In contrast, wholesale prices are usually for partially processed fish (trimmed and partially butchered). We thus need to introduce a conversion or product loss factor to account for differences in the units of the product at the two levels of the market chain. For simplicity, we assume that an exvessel-to-wholesale conversion 7

For simplicity, we ignore the determination of the TAC in this paper. Homans and Wilen [26] discuss how regulatory agencies typically choose a TAC in order to maintain biomass at some ‘‘safe’’ level. We also ignore the complications that would be introduced by making the model stochastic. In the model presented here, we assume a deterministic world in which the equilibrium is reached by some tatonnement process. Alternatively, we could assume that the equilibrium is reached with each player operating in a Nash setting, forming expectations of the other’s actions, etc. Again, the qualitative conclusions are not changed appreciably by alternative and more complicated equilibrium mechanisms.

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

387

factor of k applies, so that each pound of landed weight converts into k pounds of wholesale quantity, QF (fresh). How will the market clear under these simplified circumstances? Since fish are only marketed as fresh fish during the open harvest season, we can use the conversion factor and the expression for harvest (Eq. (1)), to determine the quantity of dressed fish entering the fresh wholesale market at each instant, namely: QFt ¼ kH t ¼ kqEX 0 eqEt :

ð5Þ

For tractability we specify an isoelastic instantaneous wholesale demand function for fresh fish: QFt ¼ BðPFt Þb :

ð6Þ

Then, we can insert the exvessel supply Eq. (5) into the demand function (Eq. (6)) and, expressing in inverse form, we get:  1=b kqEX 0 eqEt F Pt ¼ : B This is the within-season equilibrium wholesale price path, which rises at rate qE=b as the quantity landed falls with biomass depletion during the season. Whether prices actually rise noticeably is an empirical question, of course, and dependent on catchability and demand elasticity. The effort level in this inverse wholesale demand curve is endogenous and dependent upon exvessel prices rather than wholesale prices. In this simplified model, exvessel prices can be found by multiplying the wholesale price by the conversion factor: PEV ¼ kPF , so that the exvessel price path during this phase can be written as   qEt 1=b ðb1Þ=b qEX 0 e EV F : ð7Þ Pt ¼ kPt ¼ k B This expression for the within-season path of exvessel prices can be inserted into the rent dissipation equation (2) so that, with the regulatory decision rule in (4), we have two Eq. (2) and (4) in two unknowns [E and t]U Thus if the entire harvest is allocated into a fresh market only, the joint exvessel/wholesale market, regulator, and effort equilibrium will be one in which: total exvessel demand over the season just equals the TAC; the equilibrium exvessel price path just induces the amount of effort to harvest the targeted amount over the targeted season length t, with zero rents. 2.4. The market—fresh and frozen The above scenario is straightforward and only a modest extension of the basic regulated open access model. But the implications of including a more realistic depiction of the market are more profound than suggested in this simple model. In particular, when we close the regulated open access model with a processed market sector, it becomes clear that the market may accentuate the problems arising from the rent dissipation process. To outline one scenario, as the market grows, for example due to population and income growth, exvessel prices will be bid up, attracting more potential effort (capital and variable inputs) to the fishery. But then regulators will have to constrict seasons (or tighten other efficiency-stifling regulations) in order to ensure that the TAC

ARTICLE IN PRESS 388

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

is not exceeded. Thus over the long run, a fishery like the one depicted here is likely to find itself backed into a shorter and shorter season in the face of pressures from expanding markets. Shorter seasons are, in fact, what we see in many regulated open access fisheries dominated by season length restrictions. Classic examples include both the halibut and sablefish fisheries in the North Pacific, both of which were reduced to 5-day seasons by the end of the 1980s. These cases are not isolated; other examples with artificially short seasons include salmon and roe herring on the West Coast, and Chesapeake Bay oysters, and clams and quahogs on the East Coast. The phenomenon of increasingly compressed seasons was not predicted by the Gordon model since he ignored both market complexities and regulations in his simple pure open access model. A second important implication of this depiction of regulated open access with markets is that it helps explain how alternative (and lower valued) processed markets might develop in response to shortening seasons. As the fresh market grows with income and population growth, prices will rise, potential effort will be attracted, and the season will shrink. But in a paradoxical manner, shorter seasons open up enhanced opportunities for markets that ultimately substitute for the fresh market. This is also what we see in many regulated open access fisheries, namely a gradual conversion of the fishery from one serving exclusively a fresh market to one serving (almost exclusively) a frozen or otherwise processed market. Modeling the possibility of a frozen or otherwise processed product that is storable adds considerable complexity to the regulated open access model. This is because, unlike the case with fresh fish, the daily market for frozen fish must clear according to stock and flow demands for an inventory market. With inventories of a stored product, we can divide a typical year into a harvest season and an inventory dissipation period. In order to accumulate inventories of frozen fish to sell after the season closes, inventory holders must begin buying fish to store during the harvesting period. This implies that the exvessel market will be influenced by the joint willingness to pay of both fresh and frozen fish wholesalers. Fig. 1 shows one possible configuration of harvesting and marketing activities during a typical season. As the figure is drawn, there is an initial fresh phase during which all of the harvested fish are sold into a fresh market. This occurs until time s, at which time some of the harvest is added to inventory in an accumulation phase for later sales after the season is closed. During the period between s and t, raw fish are bid for by wholesalers serving both the fresh market and the market for additions to inventory. At the end of the harvest season at time t, there is a third dissipation phase in which the frozen market is served out of accumulated inventories, a period that lasts until the end of the inventory dissipation period, date T.

Harvest Season 0

τ

s Inventory Accumulation

Fresh Market

T Inventory Dissipation

Frozen Market

Fig. 1. Phases of the harvest/marketing year.

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

389

In the following sections, we characterize the price and quantity paths that emerge in each of the three possible marketing phases. In particular, we need to find the exvessel price path in the fresh and accumulation phases in order to express the zero rent equation (Eq. (2)) in terms of the four endogenous variables. We do this by deriving equilibrium exvessel prices that reflect the seasonal wholesale demands for fish, in both the fresh market and the market for inventory. These prices reflect the elimination of arbitrage opportunities between wholesalers buying for fresh and frozen markets during the fishing season, and also among wholesalers selling inventoried fish after the season is completed. 2.4.1. The fresh phase As discussed above, in the phase in which all harvest flows into the fresh fish market, the wholesale and exvessel price paths increase since harvest is declining as the season progresses. During this exclusively fresh market phase, total exvessel demand for fresh fish DF1 must equal total supply. But total supply is the integral of harvest (Eq. (1)), between 0 and s, hence: DF1 ¼ X 0 ð1  eqEs Þ:

ð8Þ

2.4.2. The accumulation phase We define the accumulation phase as one in which at least some of the harvest is allocated to the frozen wholesale market for storage and subsequent sales.8 It may also be the case that some harvest is allocated simultaneously to the fresh market, in which case the exvessel price path will be determined by competition between fresh market wholesalers and frozen market wholesalers. In equilibrium, the exvessel price path during the accumulation phase will partially reflect the same markdown of wholesale fresh prices as in (7) above. Importantly, however, the exvessel price path must also reflect the shadow value path of additions to inventory, which depends on the plans of inventory holders in the competitive frozen wholesale market. In a storage equilibrium, inventory holders must be indifferent between selling their stored product at any time during the post-season marketing period. In addition, their willingness to pay during the accumulation phase or buying period should reflect their expected sales prices during the post-season marketing period and costs of storage. For simplicity, we ignore holding costs other than the opportunity cost of inventories.9 Under these assumptions, in order for buyers to be willing to buy and hold stocks of inventory, they must anticipate that the net wholesale price will rise at a rate to cover their opportunity costs. This implies that the wholesale frozen price PZ must rise exponentially, or Z P_ ¼ rPZ ; 8

We assume that frozen fish are only consumed after the season closes. This assumption seems reasonable, particularly for short seasons. 9 This is an assumption we make for analytical convenience since it allows us to solve for a relatively simpler closed form solution than would be the case with, for example, fixed holding costs. Empirically, it may not be too inaccurate to assume that the most important holding costs are interest costs and deterioration costs, both of which can be depicted with a simple geometric rate.

ARTICLE IN PRESS 390

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

where r is the discount rate. Integrating this equation from t to T yields the arbitrage equilibrium price path in the wholesale market for frozen fish: PZt ¼ PZT erðTtÞ :

ð9Þ

where PZT is the price at the last date (T) in the inventory dissipation period. Only if prices follow this path from the start of the inventory accumulation period (s in Fig. 1) to the end of the inventory dissipation period, T, will wholesalers be willing to hold some raw product in inventory for later sales. Between the beginning of accumulation in period s and the end of the harvest season t, the shadow prices for additions to inventory will follow this path, but after t, inventory holders will receive these actual prices in the wholesale market for frozen fish. The exvessel price path during the harvest season will be a markdown of the net wholesale shadow price path:10 PEV ¼ kPZt ¼ kPZT erðTtÞ : t

ð10Þ

During the inventory accumulation phase, wholesalers in the fresh market may also be competing to buy raw fish, and hence must also offer a competitive price in order to obtain raw fish to serve the fresh market.11 If they offer an exvessel price below that offered by wholesalers purchasing for inventory, they will be unable to operate in the market. If wholesalers in the fresh market can offer an exvessel price above that determined by the inventory shadow price, then they will be able to purchase the entire harvest for the fresh market. Thus, in an arbitrage equilibrium in which both fresh and frozen wholesalers are buying, the exvessel price paid by fresh market wholesalers must equal the exvessel price paid by accumulators of inventory. The net wholesale price for fresh fish will then equal the net wholesale price for frozen fish: Z Z rðTtÞ : PFt ¼ PEV t =k ¼ Pt ¼ PT e

ð11Þ

This allows us to calculate the total wholesale demand for fresh fish between s and t: Z t  b BerbT  rbs B PZT erðTtÞ dt ¼  b  erbt : e s PZ T rb This is the cumulative fresh wholesale demand during the accumulation phase [s; t], given that fresh wholesalers must pay the same price that inventory holders are also paying. The total exvessel demand for fish destined for the fresh market in the accumulation phase can then be computed by dividing by k to get landed weight, or BerbT  rbs e  erbt : DF2 ¼  b k PZT rb

ð12Þ

The remainder of the harvest in this phase is bought by inventory holders as inventory supply destined for the frozen wholesale market. 10 We assume, for notational parsimony, that the landed to wholesale conversion factor for frozen is k, or the same as for landed to fresh wholesale. 11 A reviewer points out that there is some debate about whether the exvessel market is indeed competitive (see, e.g., Love et al. [30]). A longer season may help to dilute any monopsonistic behavior among wholesalers. This may be an additional means by which exvessel prices increase once ITQs are introduced.

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

391

2.4.3. The inventory dissipation phase During the inventory dissipation phase from t to T, there is no harvesting since the season is closed and the inventory is sold until it is fully depleted.12 If we assume that instantaneous wholesale demand for frozen fish is also isoelastic: QZt ¼ AðPZt Þa : Since prices follow Eq. (9), the frozen sales path is given by  a QZt ¼ A PZT erðTtÞ : Total wholesale demand for frozen inventory can be derived by integrating this quantity between t and T: Z T  raðTtÞ   A rðTtÞ a e A PZ dt ¼  Z a 1 : Te PT ra t Finally, total demand for frozen inventory in landed weight is computed by dividing by the conversion factor to get:  raðTtÞ A 1 : ð13Þ DZ e 3 ¼  Z a k PT ra

2.4.4. Inventory supply and demand The total supply of raw fish placed in inventory during the accumulation phase must exactly balance with the demand for inventory for ultimate sales over the subsequent marketing or dissipation period. The supply of fish added to inventory must equal the total supply of fish over the harvesting season (Eq. (3)) less total fresh demand in both the fresh and accumulation phases (Eqs. (8) and (12)): SZ ¼ X 0 ð1  eqEt Þ  DF1  DF2 : Demand for inventory is DZ3 is given in Eq. (13), and hence the inventory-clearing equation that must be satisfied is that S Z ¼ DZ3 , or  raðTtÞ BerbT  rbs A e  erbt   Z a  1 ¼ 0: X 0 ðeqEs  eqEt Þ   b e k PT ra k PZ T rb

ð14Þ

12 There are several ways to close this model. One is to assume that there is an exogenously given choke price. Then the end date of the marketing period T will need to be determined endogenously. A second way to close the model is to assume T is given and then allow the final price at that date to be endogenous and determined. We close the model in this second manner in this paper since it seems to fit most of the cases we are familiar with better. Most markets with a semi-perishable product form (such as frozen) operate on a year-to-year basis, with the end of the marketing period coinciding with the new season opening. In this way inventory holders attempt to sell all of their accumulated inventories by the start of the new season. For processed products that are more durable (such as canned) inventories are carried over. Carryover complicates the story considerably by linking optimal inventory decisions within the current marketing period to expectations of all future harvests, etc. (see [25]).

ARTICLE IN PRESS 392

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

2.5. Equilibrium Now that expressions for the exvessel price paths in both phases have been developed, we can rewrite the zero rent condition (Eq. (2)) to include these prices:   Z S qEt 1=b rt ðb1Þ=b qEX 0 e e k qEX 0 eqEt dt B 0 Z t Z t rt Z rðTtÞ qEt þ e kPT e qEX 0 e dt  ert vE dt  fE ¼ 0: S

0

This expression, when integrated out, gives us:  B1=b ðkqEX 0 Þðb1Þ=b b ðqEðð1bÞ=bÞrÞs 1 e qEð1  bÞ  rb   vE rT qEs þ kPZ e  eqEt  ð15Þ ð1  ert Þ  fE ¼ 0: T X 0e r This zero rent equation, the inventory clearing Eq. (14), and the regulator equilibrium equation (4) give us three equations, and four unknowns: E, s, t, and PZT . The remaining equation relates the choke price to an arbitrage condition that must hold at either the beginning or the end of the inventory accumulation period. There are two qualitatively different scenarios possible, depending upon whether the equilibrium price path for fresh fish rises more quickly or more slowly than the discount rate. Recall that the fresh fish price path rises at a rate qE=b. If this rate (evaluated at the equilibrium effort level) is smaller than r, then the shadow price path of inventory will cross the fresh fish wholesale price from below. Under these conditions, the market will be characterized sequentially by an exclusively fresh phase, an accumulation phase with both fresh and frozen inventory buyers, and a dissipation phase. This pattern is the case that is developed above in Eqs. (8)–(16), illustrated in Fig. 2. In Fig. 2, fresh wholesaler buyers outbid frozen wholesalers during the first phase, but in the second (accumulation) phase, inventory buyers pay the shadow price of additions to inventory and some raw product goes into each of the fresh and frozen markets.13 To the equilibrium Eqs. (4), (14), and (15), we add:   qEs 1=b ðb1Þ=b qEX e k ¼ kPZT erðTsÞ ; ð16Þ B which enables us to find s, or the date at which inventory accumulation begins. This arbitrage condition requires that the exvessel price paid by fresh market wholesalers for the last unit purchased in the exclusively fresh phase is just equal to the price paid for the first unit of raw product purchased for inventory. Another qualitatively different pattern of market sequences may arise when the fresh fish market price rises faster than the discount rate, in which case the shadow price path for inventory 13

The dotted price path below the actual bold joint wholesale price represents the price that wholesalers would be paying for raw fish destined for the fresh market, if they were buying all of it, at the equilibrium effort level. Since they must pay the price frozen wholesalers are paying, the quantity actually destined for the fresh market is less, by the amount entering frozen inventories.

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

393

Price

PTz

equilibrium price path frozen price path: slope=r

fresh price path: slope=qE/β τ

s

0

T

time

Quantity

harvest path

inventory accumulation

fresh market inventory fresh market dissipation 0

τ

s

T

time

Fig. 2. Equilibrium wholesale price path: early fresh scenario.

will cross the fresh fish price from above as in Fig. 3. In this scenario, the market will be characterized by a sequence involving an accumulation phase in which raw fish are allocated to both wholesale markets, an exclusively fresh phase, and a dissipation phase. In this case, frozen wholesalers establish the price in the first (accumulation) phase and some raw product is allocated to both fresh and frozen wholesale markets. Buyers for the fresh market must pay the inventory shadow price; again the dotted line shows what they could pay if all landings were devoted to the fresh market (at the equilibrium effort level). In the second phase, fresh wholesale buyers can pay more than the shadow price of inventory and hence they receive all the raw product. The dotted line indicates the amount frozen wholesalers could pay, but they are outbid in the second phase.14 The analytics of this scenario require straightforward modifications in (14) and (15), and then (16) 14

Note the drop in wholesale prices at the end of the harvest season. This does not indicate a disequilibrium because the prices are for different wholesale products. In the last instant of the second phase, wholesalers are paying a high price for raw fish to sell into the fresh market, but in the next instant there is no fresh market and the wholesale prices clear the frozen market.

ARTICLE IN PRESS 394

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

Price

PTz

equilibrium price path

frozen price path slope=r

fresh price path: slope=qE/β τ

s

0

T

time

inventory accumulation

Quantity

harvest path

fresh market

0

inventory dissipation

fresh market

τ

s

T

time

Fig. 3. Equilibrium wholesale price path: late fresh scenario.

gives the date s at which the accumulation period ends.15 A priori, it is difficult to tell which sequence of markets should prevail in any situation. Different combinations of parameters will lead to different qualitative scenarios.16 Thus we turn in the next section to a calibration and numerical application based on explicit parameters drawn from the North Pacific Halibut fishery case. 15

The interpretation of the modified arbitrage equation in (16) would then be that it represents the switchover date at which the shadow value of the last unit added to inventory equals the value of the first unit allocated to the exclusively fresh market. 16 It should also be noted that corner solutions are possible. For example, it would be possible for accumulation to take place over the entire season. It would also be possible for parameters to be such as to allow only the fresh market to be viable. It is also possible, in principle, to have circumstances in which the shadow price of inventories is always greater than that for fresh fish, although this would invite some odd arbitrage opportunities whereby fresh fish are bought and immediately converted to frozen. For the models we present in this paper, with isoelastic demand curves, the solution is always interior in the sense that it never pays to devote all of the raw product to only one market because the price in the other market would then approach infinity.

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

395

3. A numerical application The system developed above to describe the workings of the exvessel and wholesale markets in a regulated open access fishery is complicated and difficult to examine using analytical methods such as comparative statics. In the first place, there are several kinds of qualitative equilibria possible, with a range of interior (and corner) solution outcomes depending upon parameters in the harvesting, regulator, and marketing components. In addition, the order of the marketing, accumulation and dissipation phases depends upon which of two regimes one is in, and the equilibrium regime is itself endogenous. Finally, we are most interested in tracing out how the system might evolve under changing market conditions, requiring a dynamic simulation. In order to explore some of the model’s predictions, we calibrate it with parameters drawn from data for the North Pacific Halibut Fishery.17 While the calibration is meant simply to be illustrative and to illuminate possibilities in any generic fishery, the halibut fishery has exhibited a long history of interesting and relatively unexplored interactions between the market and the fishery.18 During the early periods of the fishery’s development (1890–1920), the fishery was conducted year long under completely open access conditions and most fish were sold into local fresh and moderately processed markets. In the 1920s, rail completions and refrigerated shipping opened up eastern markets and resultant high prices continued to draw in effort and deplete stocks. After a decade of negotiation, the US and Canada agreed in 1930 to jointly manage the fishery, using scientifically determined TACs designed to rebuild the stocks to healthy levels. The success of the rebuilding program attracted entry throughout the 1930–1960 period, and this was met with steadily reduced season lengths, falling to less than two months in the 1950s. During the 1960s and 1970s, world demand for fish boomed, fueling price increases and pressures on most stocks, regulated or not. In the halibut fishery, the frozen market was buoyed by strong demand in the whitefish fillet market that emerged in response to the growth of the fast food business and the promotion of fish as a healthy protein source.19 By the 1970s, open harvesting seasons in both the Alaskan and Canadian portions of the halibut fishery had fallen to a week or less. Using data described in the appendix, we calibrated our conceptual model in order to simulate the evolution of a regulated open access fishery with a realistic market structure. Our aim with these simulations is, first and foremost, to explore how exogenous market forces might influence the evolution of the structure of any generic regulated open access fishery over time. Since we have calibrated the model with halibut fishery parameters in particular, a secondary aim is to ask 17 This fishery has been described and analyzed elsewhere, but the important point for our purposes is that it fits the conceptual structure outlined in the model presented here almost exactly up until it was rationalized in the early 1990s. First, it was managed from 1930 until recently almost exclusively with season length controls. Second, regulators set TACs on a yearly basis in accordance with implicit and explicit goals designed to keep the stock as safe levels. Third, the market consists of a fresh and frozen market. In the frozen market fish are inventoried and removed during the marketing season and sold into higher levels of the marketing chain. 18 The exception to this relative neglect of market analysis is the remarkable study by Crutchfield and Zellner [12] originally published in 1962. Crutchfield and Zellner devoted several chapters to the workings of the halibut market at the wholesale and exvessel levels. They note that, by the 1950s, regional markets like the Seattle market were serving both a fresh wholesale as well as an inventory market for frozen fish, and that exvessel prices were determined by the simultaneous operation of both forces. Other work on the halibut fishery includes Lin et al. [29] and the recent analysis by Herrmann [24] that examines the impact of ITQs on the Canadian halibut market. 19 See Edwards [14] for an analysis of demand growth in fisheries.

ARTICLE IN PRESS 396

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

Table 1 Impacts of fresh market growth Fresh market scale parameter

Season length

Effort

Exvessel price at 0

Exvessel price at t

B

t

E

PEV 0

PEV t

1.66 2 3 4 5 10 15

12.00 10.71 8.41 7.13 6.28 4.29 3.46

12.21 13.69 17.42 20.58 23.35 34.16 42.38

0.34 0.36 0.42 0.47 0.51 0.68 0.81

0.39 0.41 0.48 0.54 0.59 0.78 0.93

whether what we have broadly witnessed in the halibut fishery could plausibly be reflecting market growth as well as technical change and rent-dissipating input growth. We caveat this, however, because we are not attempting a formal hypothesis test of this issue; for that we would need a more comprehensive econometric investigation of the multiple drivers of capacity growth including technical change and production input growth. The simulations we report here show how market growth can result in a steady and selfreinforcing progression of a fishery into shorter and shorter seasons, much like what we have observed in many season-length regulated fisheries. The simulations in Table 1 are calibrated with halibut cost and catchability parameters. The demand elasticities are consistent with empirical studies of the fishery, and we use demand shifters as free parameters, initially chosen to mimic the very early phases of a fishery selling into a predominately fresh market and then chosen to simulate generic market growth. The simulation begins by depicting a regulated open access fishery characterized by a reasonably long season, modest amounts of effort, and an exclusively fresh market. In the first row, the fresh market is relatively small and the rent dissipating effort level is just enough to harvest the TAC in a 12-month period. As a fishery matures, we would expect growth of the fresh market, driven by income and population growth and transportation/ handling innovations. These changes are simulated by a shift in the wholesale demand curve, the implications of which can be seen in subsequent rows of Table 1.20 This hypothesized growth of the market causes fresh demand to rise, causing exvessel prices to rise, attracting more potential effort to dissipate the potential rents that grow as the market expands. Unlike the pure open access Gordon model (which would predict a progressively lower open access biomass level and lower harvests with market growth), the regulated open access model predicts that biomass and harvest remain fixed by regulatory actions. The primary effects of exogenous market growth then are growing potential effort and a season length that progressively falls over time. At some point in this pattern of effort growth and shorter seasons, it becomes profitable for a wholesale processed market to emerge. In particular, with a larger and larger fraction of the year left with no fish as the fresh market is compressed, it eventually pays to buy fish for storage and 20

For simplicity, we simulate market growth by increasing the scale constant in the wholesale demand curves. We might expect elasticities to change also, although mixing shifts and elasticity changes muddies the overall results.

Frozen market scale parameter

Inventory accumulation end period

Season length

Effort

Exvessel price at 0

Exvessel price at t

B

A

s

t

E

PEV 0

PEV t

15 15 15 15 15 15 15

0 0.1 0.2 0.3 0.5 5 10

0.00 1.79 2.51 3.05 3.32 2.35 1.73

3.46 3.43 3.40 3.38 3.32 2.35 1.73

42.38 42.72 43.06 43.41 44.10 62.28 84.81

0.81 0.85 0.86 0.87 0.88 1.18 1.56

0.926 0.919 0.912 0.906 0.909 1.211 1.586

Percent frozen

Percent fresh between s and t

Wholesale price at T

PZT 0.00% 1.52% 2.99% 4.44% 7.30% 52.38% 74.09%

100.00% 46.72% 25.68% 9.53% 0% 0% 0%

1.05 1.10 1.11 1.12 1.14 1.53 2.02

ARTICLE IN PRESS

Fresh market scale parameter

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

Table 2 Impacts of processed market growth

397

ARTICLE IN PRESS 398

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

sales during the post-season period. Table 2 depicts this scenario by a gradual shift of wholesale frozen demand. We assume that frozen demand is more elastic than fresh demand, and that the market is initially dominated by the existing strong fresh market.21 But as marketers focus attention on building the frozen market, and as income and population grow, the demand for frozen fish will also shift out, just as depicted under the fresh regime. As is depicted in Table 2, this opens up an inventory market in which it proves profitable for some raw product to be stored for later sale. Using the halibut case parameters, we find that the equilibrium for this set of parameters produces the particular sequence that first involves an accumulation phase in which both fresh and inventory wholesalers buy raw product as soon as the season opens. For example, in the second row in Table 2, inventory accumulation takes place until period 1.79, at which time inventory wholesalers can no longer compete against fresh buyers. The exclusively fresh market then lasts from period 1.79 until the end of the season at 3.43, at which time the frozen product is sold out of inventory. With such a small frozen market in the early phases of the hypothetical fishery’s evolution, most of the important market activity is predicted to take place in the fresh market; 98.48% of the harvest is sold fresh, split roughly evenly between the accumulation phase and the exclusive fresh phase. As the frozen market begins to grow however, exvessel prices rise, effort is attracted, and the season length is reduced. Note the interesting phenomenon with exvessel prices. As the frozen market grows, exvessel prices at the initial date 0 are always bid up, but exvessel prices at the terminal date t fall, within some ranges of parameters. This occurs because the season must be shortened, reducing the period during which the exclusively fresh market can operate: t arrives earlier. At some point (here when the scale parameter for the frozen market reaches 0.5) the whole season is characterized by an accumulation phase in which both fresh and frozen buyers are buying in an arbitrage equilibrium. Over the long run, we would hypothesize a process in which the frozen market continues to grow, fueled by population and income growth and market development. This market growth leads to a continuation of the process exhibited during the all-fresh market phase; namely a situation in which the season length is progressively shortened as the new market grows and stimulates rising exvessel prices. Hence we see in Table 2, for example, the season dropping from three to two and finally to one month as more and more effort is attracted. By the time the frozen market has reached the stage depicted in the final row, the bulk (74.09%) of the raw product is frozen, stored, and dissipated out over a marketing period that lasts nearly a year (10.27 months). This simulation reveals a mechanism by which the market may be adversely affected in an open access system. In particular, even when fresh markets yield higher wholesale prices, ceteris paribus, the evolutionary forces associated with fresh market growth sow the seeds of their own destruction by attracting potential effort and triggering regulatory counter-reactions (shorter seasons). So it is the endogenous regulatory reaction that opens up preconditions for markets for processed commodities that may be inherently inferior uses of the raw product. And this occurs because regulators are assumed to be focused on stock safety rather than rents, and myopically responsive to symptoms (excessive inputs) rather than the cause (ill-defined property rights) of the problem. The process may be self-reinforcing, since shorter seasons cause more marketing 21

Table 2 has the frozen market growing from a trace level to higher levels, all assuming that the fresh market is still characterized by the scale given in the last row of Table 1 (namely B ¼ 15).

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

399

resources to be shifted into processed market development, hastening the constriction of the season. Eventually, with very short seasons, most of the product must be allocated to the frozen market, even if it would yield more revenues when spread over a longer period in the fresh markets. In summary, the analytical model and the simulations developed here provide an alternative or additional explanation for a common observation of short seasons in modern regulated open access fisheries. While many would be inclined to attribute progressively shorter seasons to technical change driving increased capital stuffing, we show here that exogenous market growth alone (or in combination) could just as plausibly be responsible for observed dynamics in regulated open access systems. More generally, the model shows mechanisms by which the market may both drive and react to changes in a regulated open access fishery. Market growth may drive the potential rent generation process by raising prices and bidding up raw product values. But when effort enters a regulated open access fishery in response to potential rents, it will be met by regulatory reactions, and these may influence the impact of market growth and set the process off in new path-dependent directions. In our example, as fresh markets grow with population and income growth initially, the fishery attracts entry, forcing regulators to restrict the season. But as the season is reduced, the window opens for the appearance and development of alternative processed markets. Processed market growth may then take over as a driver of the system in a selfreinforcing spiral towards the absurdity witnessed in many modern fisheries in which all the product is devoted to lower end value processed uses in an artificially intense and short season. This product switch to low valued end uses can be another dimension of rent dissipation associated with open access, although the fundamental cause of rent dissipation is still ill-defined property rights. It is not clear, a priori, how the magnitude of product conversion losses compare with the losses attributable solely to excessive inputs that Gordon focused on in his influential 1954 article. It is likely, however, that ignoring the potential for market losses will lead to an underestimation of the potential waste from open access.

4. Concluding remarks Among natural resource economists generally, and among those who study renewable resources specifically, there has not been much attention paid to resource output markets.22 This is in 22

We do not want to oversell this point or imply that economists have not provided important insights into how fisheries markets operate. There has been a considerable amount of work, particularly in the last decade or so, estimating retail and wholesale demand for fish. Examples include Anderson and Bettencourt [2], Asche et al.[4] Saalvanes and DeVoretz [33], Eales et al. [13], Hardle and Kirman [23], Gordon and Hannesson [18], Gordon et al. [19], Graddy [21] and Anderson [1], and Wessells and Anderson [37]. In addition, there has been some work describing the role of the processing sector in fisheries. For example, early work by Clark and Munro [11] and Schworm [34] looked at the manner in which processor market power might mitigate pure open access rent dissipation and Love et al. [30] examine the link between season length and market power. Matulich et al. [31] and Weninger [36] have recently modeled wholesale/exvessel linkages in a vertical market chain for fisheries products. A nice example of an empirical analysis of marketing losses associated with sub-optimal regulations in the Whiting fishery is in Larkin and Sylvia [28]. Our main point is that there has not been enough attention devoted to understanding the mechanisms linking rent dissipation, regulations, and markets.

ARTICLE IN PRESS 400

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

contrast to what we see in agricultural economics, where there is a long-standing tradition of relatively evenly balanced interest in production and marketing topics related to food production. We have suggested a reason for this in the case of fisheries economics, namely an intellectual tradition heavily influenced by H.S. Gordon’s early paper on open access exploitation, which downplayed the market in order to focus on distortions in production inputs and output supplies. Whatever the reasons for the lack of formal attention, by underplaying the role of the output market in the rent dissipation problem in fisheries, we have failed to appreciate some of what has been happening in the important test cases of fisheries rationalization with individual transferable quotas (ITQs). Because the market responds quickly to changes in raw product quality, output prices may jump almost immediately after rationalization, leaving significant market-side rent gains in the hands of quota owners. Regardless of the size of the cost savings and rents that eventually emerge from reconfigured input combinations and reduced inputs, it is clear that immediate revenue gains can also be substantial. In this paper, we take a first cut at depicting how markets operate in regulated open access settings and how these potential revenue gains are originally lost. The model presented here is very stylized, in the sense that the fishing technology, regulatory structure, and market are all depicted under simplified assumptions. The only regulatory instrument assumed in play is the choice of season length, and the only two markets are a fresh and frozen market. Our results suggest some alternative or additional mechanisms behind the observation that many of the most important and valuable fisheries are backed into a situation of extreme excess effort, short seasons, and inferior products, even in the face of worldwide growth in demand for high quality products. In particular, our results show how, over time, an increasing fraction of raw product may be diverted to processed products. The mechanism for this is subtle, because it is partly driven by regulatory reactions to increased fishing effort, increases that, in turn, are driven by ill-defined property rights. In our model, as potential fishing power increases, effective effort must be stifled by shortening the season. Shorter seasons, in turn, produce a kind of structural change in the market by reducing the opportunity to sell into the fresh market and increasing the need to hold raw product in inventory to sell into inferior markets later in the year. This is a positive prediction, of course, with normative implications related to the social welfare losses due to the waste. But it also suggests some hypotheses about where we might expect to find large market losses vis a vis excess effort costs. In addition to predicting mechanisms showing how regulations, entry, and the market interact, the model can be used to predict what might happen once a regulated open access resource is rationalized with an instrument like ITQs. It is possible to run the model ‘‘backward’’ from the end point of the scenarios in Tables 1 and 2 depicting regulated open access equilibria with short seasons. For example, suppose that property rights are created by dividing the TAC into ITQs and distributing them to original participants. In this case, the harvesting season length would almost immediately stretch to the full year, and quota holders would plan the timing of harvests over the year to achieve the highest possible exvessel prices.23 If this can be achieved by allocating a more even and smaller average monthly flow to the fresh market, then raw product would be

23

A reviewer points out that the halibut fishing season is approximately 245 days, not a full year, due to biological considerations that prevent fishing for some portion of the winter.

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

401

Table 3 Rent decomposition after rationalization

Fixed costs Variable costs Total costs Revenues Rents

Optimal

Regulated open access

12.22 8.29 20.51 173.95 153.44

84.81 8.72 93.53 93.53 0.00

diverted from frozen to fresh markets.24 For the parameters we use in the simulations here, Table 3 shows what is predicted to happen. We begin with the regulated open access conditions depicted in the final row of Table 2, in which the fishery has attracted 85 thousand units of gear, operating over a season of less than 2 months, selling 74% of the product out of inventory as a frozen commodity. As can be seen, there are two important consequences of rationalization. The first is as predicted by Gordon, namely a reduction in the need for the excessive fixed and variable inputs attracted under open access. We see, for example, total discounted yearly input costs falling from 93 million dollars to 20 million dollars, mainly through a reduction in redundant fixed costs associated with excessive gear. Variable costs fall slightly, but the total variable costs represent costs associated with fewer fixed units of gear being used over a longer period. The other important consequence of rationalization is the gain in revenues, as the raw product is diverted back into the more valuable fresh market. With the parameters used here, the estimated market gains are associated with revenues rising from 93 million to 174 million dollars annually. While this example is stylized, it shows that it is at least possible for the gains in revenues to be equal or even larger than the cost savings induced by input reduction.25 The initial ‘‘impact effect’’ of rationalization no doubt understates the longer term effect because we would likely see new development of fresh marketing niches, essentially re-creating opportunities in the fresh market that would have been dormant under short seasons. This occurred when British Columbia adopted ITQs in their halibut fishery in 1993. Prior to ITQs, the British Columbia halibut fishery was conducted over a 5-day season, with most fish sold into frozen inventory and parceled out over the subsequent marketing season. Interviews with marketers after ITQs were adopted revealed that the fresh market network grew as wholesalers 24

This in fact happened almost immediately in British Columbia when their portion of the halibut fishery came under ITQs in 1993. See Casey et al. [7]. 25 How likely is it that revenue and market-induced changes might swamp input efficiency and cost savings in practice? This is, of course, an empirical question that will only be addressed after careful ex post analysis of rationalization cases. But a recent exhaustive study by Fox et al. [15] of the North Pacific Halibut fishery concludes: ‘‘the most striking change in the profit decompositions over all vessels occurred with respect to the output price. The increase in the price of halibut is directly attributable to the change in regulations as previous fisheries landed a frozen product caught in a total fishing season of a few daysy for the firms in the sample, increases in the price of halibut was the single biggest factor in improving performance from 1988 to 1994. Indeed, without the increases in output prices attributable to the introduction of private harvesting rights, average firm performance would not have increased over the 1988–1994 periody for this industry, it would seem that the greatest benefits associated with a shift in individual harvesting rights in 1991 has been an increase in the output price, directly attributable to a longer fishing season’’ (p. 473).

ARTICLE IN PRESS 402

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

opened up new sales opportunities for fresh fish that had never been served before. As one wholesaler put it: ‘‘when ITQs were first adopted, fresh fish prices would drop dramatically as soon as 100,000 pounds per week were landed. After 2 years, however, the fresh fish market could absorb 800,000 pounds per week before prices began to drop.’’ This kind of innovation in the product market in response to rationalization makes it even more difficult for analysts to forecast what might happen under discrete changes in property rights. The task is more difficult because fishing technology choice and the market are inextricably intertwined. To take another example, when New Zealand converted its mixed inshore trawl fisheries to species-specific ITQs in the 1980s, some fishermen completely switched gear types and fishing practices in order to access new market opportunities made possible by the slower-paced fishery. The snapper fishery is an oft-cited example in which fishermen holding rights switched away from trawl toward long line gear. This was done in order to target certain sizes to sell into the lucrative Japanese live fish trade and fishermen reportedly were able to triple their revenues. Interestingly, however, it is generally more costly to harvest with long line gear. This kind of change in response to rationalization would be completely missed if one were expecting, a la H.S. Gordon, a simple decrease in inputs in response to rationalization. If there is any central lesson to be learned from these kinds of empirical cases it is that we must be cautious about simple forecasts based on technology in use before rationalization. When discrete changes in property rights induce wholesale changes in incentives, dramatic changes can occur on both the input and output side of the ledger. If we admit that market factors may be as important as input inefficiencies in conditioning pre-ITQ technology and behavior, it becomes clear that the unraveling of a system bound up by controls and distortions can take unanticipated turns. An analyst trying to forecast the impacts of implementing ITQs in New Zealand’s snapper fishery, for example, might make predictions assuming that fishermen would use the same trawl technology, but with inputs reconfigured, only to discover that trawl fisheries disappeared after ITQs were adopted. These kinds of discrete process shifts induced by market opportunities are difficult to forecast under any circumstances, of course. What this paper suggests is that analysts ought to be aware of the possibility as regulated open access fisheries are rationalized, because rent dissipation and distortions on the marketing side of the ledger may be as important as distortions on the production or cost side of the ledger.

Appendix Parameter

Value

Source

Biomass

440

Quota

60

Fixed cost (f )

1

Implied equilibrium biomass, in millions of pounds, from the quota rule target, combining Areas 2 and 3. Homans and Wilen [26] Implied equilibrium quota, in millions of pounds, from the quota rule target, combining Areas 2 and 3. Homans and Wilen [26] Estimated fixed cost of employing a unit of longline gear for the season in Area 2. Homans and Wilen [26]

ARTICLE IN PRESS F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

Variable cost (v)

0.06

Frozen elasticity (a)

1.5

Fresh elasticity (b) Catchability coefficient (q)

1.1 0.001

Discount rate (r)

0.01

Conversion factor (k)

0.87

403

Estimated daily variable cost per unit of longline gear for Area 2. Homans and Wilen [26] Herrmann [24] has an import own-price elasticity of 1.68. Homans and Wilen [26] estimate 1.16 at the means Estimated catchability coefficient for Area 2. Homans and Wilen [26] Implied monthly discount rate, from a 12% annual discount rate Reported conversion factor from dressed weight (headed and gutted) to blocks, fillets, and steaks (US Dept. of Commerce [35])

References [1] J.L. Anderson, The International Seafood Trade, Woodhead Publishing Limited, Cambridge, UK, 2003. [2] J.L. Anderson, S.U. Bettencourt, A conjoint approach to model product preferences in the New England market for fresh and frozen salmon, Marine Res. Econ. 8 (1993) 31–49. [3] R. Arnason, The Icelandic individual transferable quota system: A descriptive account, Marine Res. Econ. 8 (1993) 210–218. [4] F. Asche, H. Bremmes, C.R. Wessells, Product aggregation, market integration and relationships between prices: an application to world salmon markets, Amer. J. Agri. Econ. 81 (1999) 568–581. [5] R.J.H. Beverton, S.J. Holt, On the Dynamics of Exploited Fish Populations, Fisheries Investigations Series, Vol. 2(19), Ministry of Agriculture, Fisheries, and Food, London, 1957. [6] P.G. Bradley, Some seasonal models of the fishing industry, in: A.D. Scott (Ed.), Economics of Fisheries Management: A Symposium. Institute of Animal Resource Ecology, University of British Columbia, Vancouver, 1970, pp. 33–43. [7] K.E. Casey, C.M. Dewees, B.R. Turris, J.E. Wilen, The effects of individual vessel quotas in the British Columbia Halibut fishery, Marine Res. Econ. 10 (1995) 211–230. [8] C. Clark, Mathematical Bioeconomics, second ed., Wiley, New York, 1990. [9] C. Clark, Towards a predictive model for the economic regulation of commercial fisheries, Can. J. Fish. Aquat. Sci. 37 (1980) 1111–1129. [10] C. Clark, G. Edwards, M. Friedlaender, Beverton-Holt model of a commercial fishery: optimal dynamics, Fish. Res. Board Canada 30 (1973) 1629–1640. [11] C. Clark, G. Munro, Fisheries and the processing sector: Some implications for management policy, Bell J. Econ. 11 (1980) 603–616. [12] J. Crutchfield, A. Zellner, Economic Aspects of the Pacific Halibut Fishery, US Dept. of Interior, Bureau of Commercial Fisheries, USGPO, 1962; Reprinted as J. Crutchfield, A. Zellner, The Economics of Marine Resources and Conservation Policy, University of Chicago Press, Chicago, 2003. [13] J. Eales, C. Durham, C. Wessells, Generalized models of Japanese demand for fish, Amer. J. Agr. Econ. 79 (1997) 1153–1163. [14] S. Edwards, Evidence of structural change in preferences for seafood, Marine Res. Econ. 7 (1992) 141–151. [15] K.J. Fox, R.Q. Grafton, J. Kirkley, D. Squires, Property rights in a fishery: regulatory change and firm performance, J. Environ. Econ. Manage. 46 (2003) 156–177 10.1016/S0095-0696(02)00027-X. [16] J.R. Gauvin, J.R. Ward, J.M. Burgess, E. Richardson, Description and evaluation of the Wreckfish (Polyprion Americanus) Fishery under individual transferable quotas, Marine Res. Econ. 9 (1994) 99–118.

ARTICLE IN PRESS 404

F.R. Homans, J.E. Wilen / Journal of Environmental Economics and Management 49 (2005) 381–404

[17] G. Geen, M. Nayar, Individual transferable quotas in the Southern Bluefin Tuna fishery: an economic appraisal, Marine Res. Econ. 5 (1988) 365–387. [18] D.V. Gordon, R. Hannesson, On Prices of fresh and frozen Cod fish in European and US markets, Marine Res. Econ. 11 (1996) 223–238. [19] D.V. Gordon, K. Salvannes, F. Atkins, A fish is a fish is a fish? Testing for market linkages in the Paris Fish Market, Marine Res. Econ. 8 (1993) 331–343. [20] H.S. Gordon, The economic theory of a common property resource: the fishery, J. Polit. Econ. 63 (1954) 116–124. [21] K. Graddy, Testing for imperfect competition at the Fulton Fish Market, Rand J. Econ. 26 (1995) 75–92. [23] W. Hardle, A. Kirman, Nonclassical demand: A model-free examination of price-quantity relationships in the Marseille fish market, J. Econometrics 67 (1995) 227–257 10.1016/0304-4076(94)01634-C. [24] M. Herrmann, Estimating the induced price increase for Canadian Pacific Halibut with the introduction of the individual vessel quota program, Canad. J. Agr. Econ. 44 (1996) 151–164. [25] F.R. Homans, Modeling regulated open access resource use, Ph.D. Dissertation, University of California at Davis, 1993. [26] F.R. Homans, J.E. Wilen, A model of regulated open access resource use, J. Environ. Econ. Manage. 32 (1997) 1–21 10.1006/jeem.1996.0947. [27] J. Kirkley, D. Squires, Capacity and capacity utilization in fishing industries, Discussion Paper 99–16, Department of Economics, University of California, San Diego, 1999. [28] S. Larkin, G. Sylvia, Intrinsic fish characteristics and intraseasonal production efficiency: a management-level bioeconomic analysis of a commercial fishery, Amer. J. Agr. Econ. 81 (1999) 29–43. [29] B.H. Lin, H.S. Richards, J.M. Terry, An analysis of the exvessel demand for Pacific Halibut, Marine Resource Econ 4 (1987) 305–314. [30] A. Love, D.M. Burton, G. Sylvia, S. Lei, Regulatory controls and market power exertion: a study of the Pacific Halibut industry, Natural Res. Modeling 9 (1995) 229–253. [31] S. Matulich, R. Mittelhammer, C. Reberte, Toward a more complete model of individual transferable fishing quotas: Implications of incorporating the processing sector, J. Environ. Econ. Manage 31 (1996) 112–128 10.1006/ jeem.1996.0035. [32] Organisation for Economic Co-operation and Development, The use of individual quotas in fisheries management, OECD Publications and Information Centre, Paris, 1993. [33] K. Salvannes, D. DeVoretz, Household demand for fish and meat products: separability and demographic effects, Marine Resource Econ 12 (1997) 37–55. [34] W. Schworm, Monopsonistic control of a common property renewable resource, Can. J. Econ. 16 (1983) 275–287. [35] United States Department of Commerce, Environmental impact statement, regulatory impact review, initial regulatory flexibility analysis for proposed individual fishing quota management alternatives for the halibut fisheries in the Gulf of Alaska and Bering Sea/Aleutian islands, July 1991. [36] Q. Weninger, Equilibrium prices in a vertically coordinated fishery, J. Environ. Econ. Manage. 37 (1999) 290–305 10.1006/jeem.1999.1069. [37] C.R. Wessells, J.L. Anderson, Innovations and progress in seafood demand and market analysis, Marine Resource Econ 42 (1992) 87–103.