Abstract
In 2010, management of New England multispecies groundfish transitioned from input restrictions on harvester effort to collective rights-based management. Faced with a large reduction in harvesting days, 432 active vessels, representing 98% of historical landings, joined one of the 17 sectors allocated catch shares. The incentives presented under sector management, combined with regulations of several separately managed, revenue-important species, led to changes in harvest strategies and the timing of landings for both multispecies groundfish and many other species targeted by the sector vessels. Temporally modified landings altered the exvessel market mix of a range of species throughout the fishing year, significantly affecting prices received as well as annual harvester revenues. Two counterfactual individual harvester landings' timing scenarios for 25 species are combined with independent fixed effects models of inverse dealer demand in estimating the revenue effects of catch shares during their first year. Aggregate gains of over $30 000 000 were found to result from advantageous market timing changes brought on by more flexible catch share management.
Introduction
Catch share management is a broadly defined fishery regulatory mechanism in which individuals or groups are given annual harvesting privileges to a portion of the total allowable catch (TAC). The use of catch shares globally—typically in the form of individual transferable quotas (ITQs)—has been steadily increasing in recent decades, though their impact on fishery resources, fishing communities and individual fishermen has been a topic of ongoing empirical and theoretical debate. [For a brief overview of this discussion as it pertains to the use of ITQs, see Sumaila (2010).] Applications in multispecies fisheries can raise additional challenges, as low allocations for depressed stocks may restrict harvest of abundant species and incentivize discard of limiting species (Copes, 1986; Turner, 1997; Squires et al., 1998). Nevertheless, the National Oceanic and Atmospheric Administration (NOAA) declared catch shares a crucial tool in revitalization of biologically and economically underperforming US fisheries, initiating a recent wave of catch share programs (NOAA, 2010). Empirical evaluation of these programs is vital in connecting circumstance with biological, economic and social results so that existing programs can be refined while new programs may be designed more effectively.
The Northeast (NE) multispecies groundfish fishery, an historically prominent New England resource, was recently transitioned from days-at-sea (DAS) to catch share management under Amendment 16 to its Fishery Management Plan (FMP). Overlapping habitat and the use of non-selective fishing gear had motivated individual effort control with DAS management, but lack of avoidance incentives exacerbated overfishing on depressed stocks and resulted in a perpetual shrinking of allocated days, which were calibrated by allowable catch of the weakest stocks. Faced with strict stock rebuilding timetables and required by law to implement hard TACs, managers provided multispecies harvesters with two options: join a harvesting cooperative (“sector”) and receive group allocations for individual multispecies stocks, or face severe cuts in individual DAS while harvesting under common pool TAC. Nearly all active permit holders with substantial historical landings joined sectors.
Evaluating the economic performance of the NE multispecies catch share program requires understanding of the induced changes in harvest and landings strategy as well as the concomitant effects on revenues and costs. An initial performance report (NEFSC, 2011) indicated NE groundfish stocks yielded more nominal value per pound at lower effort levels during the first year of catch shares than in any of the previous three years under DAS. This finding comports with observations that tradable quota-based programs increase profitability (e.g. Arnason, 1993; Gauvin et al., 1994; Casey et al., 1995; Dupont and Grafton, 2000): more exvessel revenue can be drawn from a similar or smaller quantity of fish harvested at potentially lower cost. As of 2011, exvessel price improvements have been documented for species managed under all but two of the 15 US catch share programs (NOAA, 2011).
In these fisheries, management-based revenue improvements have been driven by two mechanisms: enhanced quality made possible by mitigating the race-to-fish (Homans and Wilen, 2005), and price increases due to improved market timing (Scheld et al., 2012). Competitive harvest pressure producing the race-to-fish results from common pool TAC management and is rarely seen in fisheries managed with input restrictions such as DAS. Under multispecies DAS, a large volume of high quality fish regularly supplied the premium market, limiting opportunity for revenue gain through quality improvements. Additionally, market flooding, producing low average prices, was relatively infrequent for multispecies stocks under DAS given the lack of race incentives. From whence, then, did the observed price increases following the introduction of catch shares come?
Historically reactive and piecemeal, US fisheries management has given rise to many instances of separate, very different, and disjoint regulations for various species exploited by the same group of fishermen. New England groundfish harvesters target and land over 20 different species, which are regulated by nearly a dozen different management plans. Many of these species are harvested jointly with non-selective trawl gear. Technical interaction among separately regulated stocks forces harvesters to make trip-level targeting and/or avoidance decisions, evaluating tradeoffs of participation in one fishery versus another. Multispecies catch shares introduced new management to a subset of target species; given the presence of numerous separately regulated, jointly harvested stocks, it is believed the transition from DAS altered exploitation patterns for the broader complex harvested by the fleet. “Spillover” effects (Asche et al., 2007; Branch, 2009), where management action leads to increased exploitation of unregulated stocks, are a related, though distinct, phenomenon.
Prior to multispecies catch shares, harvesters were allocated a limited number of DAS during which to land multispecies stocks. Scarcity of this harvest privilege promoted strategic use: if a DAS were to be used, it would be in the harvester's best interest to be sure to land a large quantity of multispecies fish. Trips landing primarily non-multispecies stocks, along with small amounts (but exceeding incidental catch limits) of multispecies fish, carried a large opportunity cost in foregone multispecies landings. This trip-level constraint on harvest and landings of non-multispecies fish was relaxed under catch shares, which allow harvesters to continuously vary the ratio of multispecies to non-multispecies fish according to market prices, harvest costs, and biological availability. The introduction of catch share management therefore enhanced flexibility in landings, increasing the range of profitable multispecies/non-multispecies landing combinations.
The resulting changes in multispecies and non-multispecies landings altered the mix of species being processed and sold down the supply chain at each point in time, ultimately affecting exvessel prices as dealers reacted to modified market conditions. The extent to which adjusted landings by one individual responding to new regulatory incentives affect other harvesters will depend on price response to quantity changes, or price flexibility. By determining the change in timing of individual harvests and identifying market relationships between quantities and prices, controlling for exogenous price determinants, a robust prediction of the revenue effects resulting from such a policy change may be estimated.
The analysis presented in this article investigates the impact of multispecies catch shares on individual permit revenues, across all harvested species. Limited access multispecies harvesting permits are attached to fishing vessels. In what follows, the terms “vessel” and “harvester” will be considered synonymous with “permit”. Two counterfactual sets of individual daily landings for 25 revenue-important species were constructed to predict what harvesters would have landed each day, absent multispecies catch share management. In addition to the nine catch-share-allocated and two small mesh multispecies species, this analysis modelled multispecies harvester landings for: monkfish (Lophius americanus), black sea bass (Centropristis striata), butterfish (Peprilus triacanthus), Atlantic herring (Clupea harengus), longfin squid (Loligo pealeii), American lobster (Homarus americanus), Atlantic mackerel (Scomber scombrus), Atlantic Sea scallop (Placopecten magellanicus), scup (Stenotomus chrysops), northern shrimp (Pandalus borealis), little skate (Leucoraja erinacea), winter skate (Leucoraja ocellata), summer flounder (Paralichthys dentatus), and spiny dogfish (Squalus acanthias). Little and winter skate landings were modelled as generic wings and rounds to align with management delineation. The top 25 revenue-important species made up on average 97% of annual fleet revenues. Species-level models of exvessel dealer demand were used to predict prices given actual or counterfactual landings. The difference between actual and constructed counterfactual predicted revenues was used as an estimate of multispecies catch shares' exvessel revenue effect for each individual harvester.
New England fisheries management
The NE multispecies FMP manages 15 species: Atlantic cod (Gadus morhua), haddock (Melanogrammus aeglefinus), pollock (Pollachius virens), Acadian redfish (Sebastes fasciatus), yellowtail flounder (Limanda ferruginea), American plaice (Hippoglossoides platessoides), witch flounder (Glyptocephalus cynoglossus), winter flounder (Pseudopleuronectes americanus), white hake (Urophycis tenuis), Atlantic halibut (Hippoglossus hippoglossus), windowpane flounder (Scophthalmus aquosus), ocean pout (Zoarces americanus), Atlantic wolffish (Anarhichas lupus), whiting (Merluccius bilinearis), and red hake (Urophycis chuss).
Since the New England Fishery Management Council's establishment in 1976, multispecies management has been a succession of relative failures (Apollonio and Dykstra, 2008). The NE multispecies FMP has at this time 18 amendments and 50 framework adjustments, representing major and minor changes, respectively. Intense bureaucracy and diverse stakeholder interests have produced regulations that are often too little too late, cultivating a sentiment of inept management among industry and local communities while simultaneously squandering multispecies populations and the profits they could produce (Acheson and Gardner, 2011). Increasingly restrictive management has forced many harvesters originally solely dependent on multispecies groundfish to diversify, creating a complex web of harvesting alternatives. The remaining fleet is a heterogeneous group of harvesters who target a variety of stocks. Vessels range from small hook-and-line operations that fish near shore on single daytrips to moderate-scale (20–30 m) trawlers that fish grounds over a hundred miles offshore and may be at sea for over a week.
Beginning in 1994, multispecies groundfish were regulated with limited entry, joint stock DAS. Under this management regime, harvesters were charged with using one of their allocated DAS if landed multispecies volume was above incidental catch allowances. In general, a per stock incidental catch of 5% total landed weight was allowed without expending a DAS. The collapse of many multispecies stocks coincided with an emergence of new export and domestic markets for demersal species like spiny dogfish, monkfish and skate, which commix with, and are often caught when targeting, multispecies groundfish. The newfound marketability of these stocks, which had previously been discarded to free up hold space for valuable species like cod and yellowtail flounder, presented harvesters with an important trip-level decision regarding targeting, retained catch, and DAS use. It should be noted that regulatory non-compliance throughout this period was commonplace, calling into question the degree to which regulation incentivized behaviour (King and Sutinen, 2010). Still, since 2004 an active DAS lease market has ensured harvesters operate on the margin with respect to DAS use.
On 1 May 2010—the first day of fishing year (FY) 2010—NOAA allocated catch shares for 14 stocks of nine groundfish species, collectively referred to as “multispecies” throughout this article. Vessels were allowed incidental bycatch of one halibut per trip, while there was a prohibition on landings of windowpane flounder, ocean pout, and Atlantic wolffish. Whiting and red hake were small-mesh multispecies regulated in 2010 with gear restrictions and trip limits. The 17 sectors, or self-identifying groups of harvesters, were allocated collective quota based on members' landing histories. The 432 participating vessels represented 55% of the eligible limited access multispecies fishing permits and 98% of historical multispecies landings. Though afforded considerable management autonomy, all approved harvesting sectors operated with a loose form of ITQs, where individual vessel allocations were determined by contributions to the group aggregate. Quota was tradable both within and between sectors. Figure 1 captures an important behavioural consequence of multispecies DAS management not present following the transition to catch shares: under DAS, small landings of multispecies stocks made jointly with commixing non-multispecies fish were avoided or illegally discarded. Catch share management facilitated proportionately more low-quantity multispecies landings when targeting non-multispecies stocks, an indication of increased flexibility in landings that may lead to greater profitability.
Kernel density plots of individual vessel multispecies landings made when also landing commixing non-multispecies stocks under DAS and catch shares. Catch shares density (translucent white) overlays DAS (dark grey). Only landings <5 000 lbs are shown. Incidental catch under DAS has been removed.
Under DAS, limited access multispecies permit holders developed harvesting patterns that varied in their dependence on the multispecies complex; in aggregate, revenues from multispecies groundfish made up slightly less than half of the average annual sector fleet total during the three years preceding catch share management (Figure 2). Throughout this period, lobster and scallop made up the majority of revenue shares for 15% of permit holders. These vessels landed multispecies groundfish as bycatch or secondary targets, relying on the species complex for <10% of annual revenues. The financial importance of multispecies groundfish varied regionally for the remaining 85% of the fleet. The majority of multispecies permits (60–70%) were held by harvesters in northern New England (Maine, New Hampshire, and Massachusetts), who relied heavily on multispecies stocks (70% of annual revenues) and, to a lesser extent, monkfish (10–20%) and skate (5%). Permit holders south of Massachusetts were more diverse, having annual revenues made up of longfin squid (30–40%), whiting (15–20%), multispecies groundfish (10–20%), summer flounder (10%), monkfish (5%) and scup (5%).
Average annual total revenues by species for multispecies sector vessels before catch shares (FYs 2007–2009). Bars are ordered according to species revenue ranking after catch shares (FY 2010).
Following the introduction of catch share management, the sector fleet did not dramatically modify annual harvest strategies. Figure 2 indicates that revenue rankings during the first year of catch share management were similar to their prior ordering under DAS. As evidenced by a mismatch in descending order and bar length, longfin squid supplanted monkfish to become the third most important revenue source for sector harvesters after catch shares. Other shifts in revenue importance were more subtle, and an overwhelming majority of species saw no considerable change. At the root of this consistency were fairly stable individual vessel annual landings before and after multispecies catch shares. t-tests indicate that, on average, vessels landed significantly more lobster and summer flounder—species which saw increases in abundance—while less skate. Additionally, three of nine multispecies fish experienced statistically significant changes in average annual exvessel volume. For all modelled species however, shifts in interannual sector vessel landings were not statistically different from those experienced by limited access multispecies vessels that did not join sectors. Though it appears that, in their first year, multispecies catch shares did not meaningfully affect the amount of fish landed by individual harvesters, as previously noted, incentives dictating how and when particular species were pursued and landed did importantly change.
The behavioural effects of multispecies catch shares were driven by landings of non-multispecies stocks under disjoint management regimes. For example, early in the season there was a noticeable increase in harvest of possession-limit-regulated groundfish species jointly landed with small numbers of multispecies groundfish, a targeting choice that is less costly under the new catch share programme. Under DAS, these low-multispecies-landing trips were either foregone in favour of less profitable trips or supplemented with additional multispecies tows, or multispecies catch in excess of the incidental threshold was discarded. Similarly, there was an increased pace in exploitation of spiny dogfish, summer flounder, and scup—three separately managed race-to-fish demersal stocks. In FY 2010, during each respective derby, these three commixing species were landed by sector vessels in higher daily volumes than was observed under DAS. Here again, the ability to profitably land a small quantity of multispecies groundfish appears to have modified harvest and landings strategies, facilitating greater participation in the race-to-fish for these commixing non-multispecies stocks. Finally, there were changes in sector landings of stocks that do not commix, or are harvested with different technologies, likely motivated by exogenous market and non-market forces such as non-sector harvests, weather, and regulatory structure.
Analysis
To estimate the individual harvester revenue impacts of the NE multispecies catch share program, we took a three-phase approach following Scheld et al. (2012). First, a 25-equation price model was fit using exvessel landings data. Second, two counterfactual environments were constructed to predict the daily landings of each sector harvester. Third, the price model was used to predict exvessel revenues under actual and counterfactual landings. The difference in revenue predictions is a measure of the catch shares' exvessel revenue effect.
Data
In the northeast US, all business entities buying federally managed fish directly from vessels are required to electronically report the species, weight, market category, grade, and transaction value of each exvessel sale. The data used in fitting price models consisted of 1.6 million dealer-reported exvessel species landing observations from 1 January 2007 through 30 April 2010. There were ∼375 000 individual species landing observations by multispecies sector vessels during FYs 2009 and 2010, which were used in assessing the effects of catch share management.
Price model
The price model is designed to predict exvessel prices at given landed quantities, facilitating comparison of actual and counterfactual revenues. That price and quantity are jointly determined in competitive market equilibrium is perhaps the most fundamental concept of microeconomic theory. Empirical studies of exvessel markets have historically relied on models of inverse demand, where price is the dependent variable and quantities the independent variables (Bell, 1968; Barten and Bettendorf, 1989). Under this framework the modeller must identify a market's spatial scope as well as substitute and complementary products (for consumers or processors) for which markets are integrated. There is evidence for statistically significant market integration of certain seafood products at national (Gordon and Hannesson, 1996; Asche et al., 1997) and international scales (Asche et al., 1999; Asche et al., 2002), as well as long-run substitutability between different species in wholesale markets (Gordon et al., 1993; Bose and McIlgorm, 1996; Shabbar et al., 1999).
In New England, many secondary fish dealers, processors, restaurateurs and grocers look to maintain a steady supply of fish, often purchasing available products at centralized markets supplied by exvessel dealers or having them shipped from regional ports. [For a detailed description of this market, which, aside from a degree of consolidation, has not dramatically changed in structure for several decades, see Wilson (1980).] This suggests that exvessel prices are determined by landings at the individual dealer, and also influenced by total same- and different-species quantities in the port, across the state, and within the region. The broad spatial and product scope of this model is needed for several reasons. First, though specialization exists to a degree, New England exvessel wholesale dealers are product generalists: in FY 2009 the top 100 revenue grossing dealers bought (and sold) 17 species on average. Processing and distribution require use of scarce inputs (e.g. skilled labour, refrigeration, fuel, etc.), creating production tradeoffs, or possibly complementarities, between species. Second, different seafood products are often similar in protein content, taste, texture and availability and may be substituted for one another by end consumers. Third, a sizable volume of fish is transacted in two large daily auctions that distribute price reports to all dealers, providing reliable market signals throughout the region. Note, however, that transportation costs, time and perishability, as well as local availability and cost of production inputs, limit coast-wide market integration, preventing the law of one price from holding ubiquitously (Linnemann, 1966).
Scheld et al. (2012) identified statistically significant cross- and own-species price flexibilities for daily aggregate statewide landings in New England, indicating extensive market relationships among a variety of exvessel products. For each of the 25 species pursued by sector vessels, our price model estimates each landing's exvessel price to be a function of regional, port, dealer, and individual harvester landed quantity aggregations across all modelled species. (See Supplementary data for further discussion and a technical description.)
Counterfactuals
Because multiple states of nature cannot be simultaneously observed (Holland, 1986), policy analysts are presented with two options in estimating the relationship between cause and effect: randomized controlled experiments, where the investigator is able to impose variation of conditions and use statistical inference to draw conclusions, or quasi-experimental procedures, where statistical methods use existing data, before and after the treatment, to extract information about the effect (Greenstone and Gayer, 2008). As a result of complex context and often expost analysis, evaluation of environmental policy and management actions tend to rely on the methods of the latter (Bennear and Coglianese, 2005; Ferraro, 2009).
Until recently, evaluation of fishery management decisions using robust counterfactuals has largely been absent from the literature (Smith et al., 2006). Early work investigating the economic impacts of rights-based management used associational evidence (Batstone and Sharp, 1999; Casey et al., 1995; Dupont and Grafton, 2000), analysed quota market performance (Newell et al., 2005), or extrapolated aggregate behaviour (Herrmann, 2005; Homans and Wilen, 2005). Not controlling for exogenous factors (e.g. climate, ecology and macroeconomy) and/or muting fleet heterogeneity, such strategies may fail to isolate a policy's true treatment effect and instead produce results driven by spurious correlations or weakened through aggregation.
These relationships offer alternative ways of identifying qijt10CF, harvester i's landings of species j at time t in FY 2010 under the counterfactual DAS policy. Equation (2) models counterfactual landings to be equivalent in pace with those during FY 2009, under DAS management. Specifically, on each day t of FY 2010, the sum of counterfactual daily landings (q) for harvester i of species j in proportion to the total FY 2010 landings by harvester i of species j was set equal to a corresponding ratio of FY 2009 landings. Rearranging equation (2) and solving for , one obtains harvester i's counterfactual cumulative landings of species j at each day t, from which counterfactual daily landings are calculated. If harvester i did not make any landings on the same day t in both years, counterfactual landings were assigned to the next landing day observed. Harvesters landed on six fewer days on average during FY 2010. For robustness, a second method was also used. Equation (3) provided counterfactual estimates equivalent to those derived from equation (2) weighted by the pace of aggregate daily landings (Q) by multispecies permit holders who did not join sectors and thus fished in the common pool (CP) under DAS. While this group of ∼350 harvesters had somewhat insignificant multispecies groundfish landings, they made up significant components of other revenue-important fisheries. The second right-hand ratio term in equation (3) measures the relative change in timing of aggregate common pool DAS landings. If species j was being landed relatively faster in FY 2010 by these vessels, this term would be >1, increasing the pace of species j counterfactual landings for each sector harvester. The addition of this term was used to control for exogenous climatic, ecological or regulatory shocks affecting all harvesters in FY 2010 but not in FY 2009. Note that both methods control for general macroeconomic effects.
Results
Price and counterfactual models
An Fjm value between zero and negative one indicates the price received for species j decreases when species m quantity increases, though the change in price is less than proportionate to that in quantity.
Total price flexibilities were constructed from the estimated inverse demand model by summing individual price flexibilities across quantity aggregations. During summation, port, dealer, and individual harvester quantity aggregations were multiplied by (1 - CVjX), or one minus the coefficient of variation in normalized spatial or individual landings distributions for species j at quantity aggregation X. This weighting accounts for variation in exvessel sales across space, dealers and harvesters. Total price flexibility therefore indicates the average percentage change in species j exvessel price from a 1% increase in same day landings, inventory levels, and expectations of species m quantity. Total price flexibilities were constructed for illustrative purposes only—price predictions relied on disaggregate values. Own total price flexibilities were predominantly large and negative, while cross total price flexibilities were comparatively smaller and a mixture of negative and positive (Figure 3). The former implies downward sloping dealer demands (i.e. when more of a product is landed, exvessel prices decrease), while the latter captures the intricacies and complexities of the exvessel market structure. [Note that for graphical interpretative purposes, skate bait price flexibilities are absent from Figure 3. Skate bait is an extremely low-value product; small price changes are therefore quite large in percentage. To better gauge colour scale these price flexibilities were excluded.] All estimated price flexibilities were between positive and negative one (, indicating exvessel prices are relatively inelastic. Statistically significant cross-species price flexibilities may arise from competing or complementary use of scarce processing resources or end consumer substitution.
Total price flexibilities are calculated as a weighted sum of statistically significant (p ≤ 0.05) individual price flexibilities. Red (green) indicates negative (positive) total price flexibility; colour is scaled to the maximum absolute total flexibility value (Fpollock, pollock = –0.4860). Species are ordered left-to-right/bottom-to-top by FY 2010 revenue importance among multispecies groundfish (southwest corner) and other species.
Counterfactual landings were modelled to reflect how individual sector harvesters would have landed each included species throughout FY 2010 had catch share management not been introduced. Methods used altered only the individual pace of actual FY 2010 landings and, therefore, do not consider changes in harvest scale, which were found to be minimal. Divergence in timing of actual and counterfactual landings by the fleet varied from species to species, though spiny dogfish and whiting, two high volume non-multispecies stocks, were landed at a substantially faster initial pace than predicted by counterfactuals. Additionally, haddock—the healthiest multispecies stock for which ample allocation was provided—was landed earlier in the season than predicted, while yellowtail flounder and pollock were landed later (see Supplementary data for further discussion of price and counterfactual model results).
Individual average and aggregate estimated impacts
To estimate the effect of temporal shifts in landings on individual harvester revenues, inverse demand and counterfactual models were combined to produce three sets of revenues: ReviA, harvester i's predicted revenues with actual landings; ReviCRCF, harvester i's predicted revenues with counterfactual landings using equation (2), referred to as catch rate counterfactual (CRCF); and ReviCPCF, harvester i's predicted revenues with counterfactual landings using equation (3), referred to as common pool counterfactual (CPCF). The difference between ReviA and each ReviCF is a measure of the impact multispecies catch shares had on harvester i's annual revenues. The summation of this difference over i provides an estimate of total program effects.
Following the introduction of catch shares, change on the intensive margins of production was found to be enormously beneficial for the multispecies sector fleet. Statistically significant estimated total gains of $33 200 000 and $40 200 000, with individual harvester averages of $79 202 and $96 010, were predicted by CRCF and CPCF, respectively (Table 1). During FY 2010, the total volume landed by multispecies sector harvesters was only 1% above the prior three-year average, yet actual exvessel revenues rose by ∼18%, from ∼170 to 200 million US dollars. The models presented here predict FY 2010 exvessel revenues would have fallen had DAS remained in place, without even accounting for the potentially large loss in total landed volume brought on through necessary DAS reductions. This analysis suggests observed price and revenue increases for the multispecies sector fleet were the direct result of landings' timing and compositional changes made subsequent to the introduction of catch share management for a subset of target species.
Average and total estimated revenue effects of multispecies catch shares for sector vessels from CRCF and CPCF
| Effect ($) | Multispecies (CRCF) | Other (CRCF) | Multispecies (CPCF) | Other (CPCF) |
|---|---|---|---|---|
| Average | +37 020 | +42 182 | +54 581 | +41 429 |
| (2 585) | (6 179) | (2 742) | (7 977) | |
| Total | +1.55 × 107 | +1.77 × 107 | +2.29 × 107 | +1.74 × 107 |
| (1.08 × 106) | (2.59 × 106) | (1.15 × 106) | (3.34 × 106) |
| Effect ($) | Multispecies (CRCF) | Other (CRCF) | Multispecies (CPCF) | Other (CPCF) |
|---|---|---|---|---|
| Average | +37 020 | +42 182 | +54 581 | +41 429 |
| (2 585) | (6 179) | (2 742) | (7 977) | |
| Total | +1.55 × 107 | +1.77 × 107 | +2.29 × 107 | +1.74 × 107 |
| (1.08 × 106) | (2.59 × 106) | (1.15 × 106) | (3.34 × 106) |
Standard errors were derived from 1 000 draws of β ∼ N(β,σ2) and are in parentheses beneath estimates. All estimates are significant at a 99% confidence level.
Average and total estimated revenue effects of multispecies catch shares for sector vessels from CRCF and CPCF
| Effect ($) | Multispecies (CRCF) | Other (CRCF) | Multispecies (CPCF) | Other (CPCF) |
|---|---|---|---|---|
| Average | +37 020 | +42 182 | +54 581 | +41 429 |
| (2 585) | (6 179) | (2 742) | (7 977) | |
| Total | +1.55 × 107 | +1.77 × 107 | +2.29 × 107 | +1.74 × 107 |
| (1.08 × 106) | (2.59 × 106) | (1.15 × 106) | (3.34 × 106) |
| Effect ($) | Multispecies (CRCF) | Other (CRCF) | Multispecies (CPCF) | Other (CPCF) |
|---|---|---|---|---|
| Average | +37 020 | +42 182 | +54 581 | +41 429 |
| (2 585) | (6 179) | (2 742) | (7 977) | |
| Total | +1.55 × 107 | +1.77 × 107 | +2.29 × 107 | +1.74 × 107 |
| (1.08 × 106) | (2.59 × 106) | (1.15 × 106) | (3.34 × 106) |
Standard errors were derived from 1 000 draws of β ∼ N(β,σ2) and are in parentheses beneath estimates. All estimates are significant at a 99% confidence level.
Distribution of estimated impacts
Positive impacts were spread throughout the multispecies sector fleet, with gains estimated for 75–80% of individual vessels (Figure 4). Both counterfactual methods predict similar distributional results: a high concentration of slightly positive effects with a long positive tail. The 20–25% of sector harvesters predicted to have suffered losses were vessels from Maine and Massachusetts who relied heavily on monkfish and lobster in addition to multispecies groundfish. Conversely, the ∼10% predicted to have gains in excess of $350 000 were large southern New England operations harvesting primarily scallop, squid, whiting, cod and haddock.
Histograms of individual harvester program effects from CRCF (dark grey) and CPCF (translucent white). Bin width is set to $10 000 and predicted effects <$150 000 and >$350 000 are collapsed into “Less” and “More”, respectively.
Though this analysis focused on estimating program impacts for harvesters transitioned from DAS to catch share management, program effects for non-catch share vessels were also estimated (Figure 5). Vessels landing in the northeast US during FY 2010 not operating under multispecies catch shares were one of two types: multispecies common pool DAS and non-multispecies harvesters. The former group was of ∼350 vessels, comparable with sector harvesters in scale and geographic distribution, while the latter group was comprised of over 3200 generally smaller operations. Both groups relied primarily on monkfish, lobster, scallop, summer flounder, whiting and herring for exvessel revenues. On average and in total, both groups were predicted to have been slightly worse off.
Histograms of individual harvester program effects for multispecies DAS common pool (a) and non-multispecies (b) vessels from CRCF (dark grey) and CPCF (translucent white). Bin width is set to $5000 and effects <−$50 000 and >$50 000 are collapsed into “Less” and “More”, respectively.
Effect of diversity in species landed
One factor influencing the extent of individual benefit from catch share management was the vessel's diversification in landings from different fisheries. A Herfindahl–Hirschman index, also called the Simpson index, is a bounded sum of squares measure that increases with decreasing diversity (Simpson, 1949; Hirschman, 1964). Only recently has the index been applied in fisheries economics, estimating the relationship between harvester income diversity, with respect to participation in multiple fisheries, and risk, as measured by the coefficient of variation in annual revenues (Kasperski and Holland, 2013). To investigate the relationship between annual landings species diversity and predicted program impacts, individual sector harvester estimated effects were regressed on FY 2010 vessel revenues, a Herfindahl–Hirschman index of individual annual landings species diversity, and dummy variables indicating state of permit registration (Table 2). Results of this regression reveal that, controlling for vessel scale and home state, increased species diversity in annual landings was significantly correlated with predicted gains, suggesting harvesters more flexible in their targeting strategy had greater success in timing exvessel markets.
Select output from OLS regressions of Effecti (equation 1) on harvester FY 2010 total revenue, a measure of species diversity in annual harvest, and dummy variables indicating state of permit registration
| Variable | Coefficient (CRCF) | p-value | Coefficient (CPCF) | p-value |
|---|---|---|---|---|
| Intercept | 68 321 | 0.001 | 71 381 | 0.004 |
| (21 132) | (24 407) | |||
| Total Revenue | 0.1528 | 0.000 | 0.2232 | 0.000 |
| (0.0177) | (0.0205) | |||
| Herf. Index | −106 109 | 0.000 | −134 316 | 0.000 |
| (29 164) | (33 684) |
| Variable | Coefficient (CRCF) | p-value | Coefficient (CPCF) | p-value |
|---|---|---|---|---|
| Intercept | 68 321 | 0.001 | 71 381 | 0.004 |
| (21 132) | (24 407) | |||
| Total Revenue | 0.1528 | 0.000 | 0.2232 | 0.000 |
| (0.0177) | (0.0205) | |||
| Herf. Index | −106 109 | 0.000 | −134 316 | 0.000 |
| (29 164) | (33 684) |
Species diversity is calculated as where θij is the share of harvester i's annual landings made up by species j. Regressions were run for predicted effects from both CRCF and CPCF. Standard errors are beneath coefficient estimates.
Select output from OLS regressions of Effecti (equation 1) on harvester FY 2010 total revenue, a measure of species diversity in annual harvest, and dummy variables indicating state of permit registration
| Variable | Coefficient (CRCF) | p-value | Coefficient (CPCF) | p-value |
|---|---|---|---|---|
| Intercept | 68 321 | 0.001 | 71 381 | 0.004 |
| (21 132) | (24 407) | |||
| Total Revenue | 0.1528 | 0.000 | 0.2232 | 0.000 |
| (0.0177) | (0.0205) | |||
| Herf. Index | −106 109 | 0.000 | −134 316 | 0.000 |
| (29 164) | (33 684) |
| Variable | Coefficient (CRCF) | p-value | Coefficient (CPCF) | p-value |
|---|---|---|---|---|
| Intercept | 68 321 | 0.001 | 71 381 | 0.004 |
| (21 132) | (24 407) | |||
| Total Revenue | 0.1528 | 0.000 | 0.2232 | 0.000 |
| (0.0177) | (0.0205) | |||
| Herf. Index | −106 109 | 0.000 | −134 316 | 0.000 |
| (29 164) | (33 684) |
Species diversity is calculated as where θij is the share of harvester i's annual landings made up by species j. Regressions were run for predicted effects from both CRCF and CPCF. Standard errors are beneath coefficient estimates.
Interpreting the effects of price flexibilities
Estimated revenue effects can be interpreted as the difference in cumulative impacts of price externalities arising under each management regime and the associated distribution of landings during the season. Since each price equation included total daily coast-wide quantities of all species modelled, a shift in landings by any given harvester affected exvessel prices received by all other harvesters landing that day. The degree to which price externalities were realized by an individual was a function of market level overlap, with harvesters landing at the same dealer on the same day imposing the greatest externality on one another. Large predicted revenue gains imply that after catch shares there were either fewer spatially and temporally coincident landings of products with negative price externalities, more such landings of products with positive price externalities, or both.
Sensitivity of the overall results to counterfactual predictions can be assessed by predicting prices after replacing counterfactual with actual landings, one species at a time. Differencing predicted revenue effects produced by this set of one actual and 24 counterfactual landings from those predicted by a full counterfactual provides a measure of gains by individual species timing change. The effects of counterfactual inclusion captured by this method are only partial; predicted revenue effects were the result of multiple simultaneous timing changes. Figure 6 indicates that predicted program impacts result from a combination of own- and cross-species market interactions. A primarily green diagonal signifies timing changes were generally own-species beneficial, while numerous green off-diagonal points highlight the benefits derived from cross-species price externalities. Interestingly, timing changes for a number of species were detrimental to haddock revenues. This result may be a product of haddock's non-binding allocation and lack of quota value providing limited incentive to efficiently time the market and take advantage of haddock price externalities.
Graphical representation of counterfactual sensitivity results. Each point depicts the effect of including column species counterfactual landings on row species FY 2010 aggregate sector revenues. Colour is scaled to the maximum absolute effect. Darker green (red) indicates larger predicted gains (losses) resulting from species' counterfactual inclusion. Species are ordered left-to-right/bottom-to-top by FY 2010 revenue importance among multispecies groundfish (southwest corner) and other species.
Following the introduction of multispecies catch shares, harvesters seasonally intensified targeting of many stocks. Multispecies sector vessels typically land around 90% of total white hake, pollock, yellowtail and winter flounder exvessel volume. All four species have large and negative total own price flexibilities, indicating an increase in daily volume decreases contemporaneous prices. During FY 2010, white hake, pollock, yellowtail and winter flounder all experienced periods of intensified targeting and, counterintuitively, higher prices. This result implies cross-species price flexibilities played a dominant role in revenue outcomes for these particular multispecies stocks. For example, decreased mid- and late-season landings of winter flounder resulted in lower levels of market mix with price impairing summer flounder (−0.0436), plaice (−0.0504), witch flounder (−0.0064), and yellowtail flounder (−0.0492). Despite the price depressing effect of higher daily winter flounder landings (−0.1004), harvesters gained by concentrating those landings away from species the market considers substitutes. All price flexibilities included in the text are total price flexibilities calculated as previously described.
Together the nine multispecies groundfish accounted for 45–52% of aggregate sector fleet gains. Cod exvessel price (15% of gains) benefited from a slightly more constant supply throughout the season and a large negative total own price flexibility (−0.3176), as well as reduced spiny dogfish (−0.0040), winter flounder (−0.0292), and skate (−0.0244) landings during autumn and winter. Though spiny dogfish, winter flounder, and skate would not generally be thought of as substitutable for cod in end consumer markets, processing, shipping, and inventory tradeoffs likely exist as large quantities of spiny dogfish, skate, and winter flounder are frozen and may be exported, while cod is processed fresh and supplies local markets. When cod landings are homogenous, dealers can focus on cleaning it well and bringing it to the best markets, but if landed with large amounts of dogfish, skate, or winter flounder it must be unloaded quickly, so efforts can focus instead on the freezing line. Flatfish species (25% of gains) are highly substitutable to consumers (average cross-price flexibility of −0.0493), and when individually landed more homogenously throughout the season they yielded higher exvessel prices. Yellowtail and summer flounder gains were partially derived from increased market mix and a positive cross-price flexibility in landings of the former on prices of the latter (0.0196), which outweigh negative effects of the opposite relation (−0.0156). This interesting and statistically significant market relationship perhaps results from a combination of supply chain functioning and unidirectional substitutability.
Scallop, longfin squid, and whiting (31–37% of gains)—three economically important non-multispecies stocks—saw revenues increase from timing changes of many different species, highlighting the value of targeting flexibility afforded through catch share management. Despite some lack of temporal flexibility in these fisheries arising from opening times, weather, or derby incentives, the fleet used catch shares to advantageously alter concurrent landings of other stocks. Increases in concomitant landings of scallop with haddock (0.0034), lobster (0.0499), and plaice (0.0031), in addition to decreases with pollock (−0.0094), enhanced scallop prices. Scallop, lobster, haddock and plaice all generally serve high-end restaurant and grocery markets whose proprietors likely place a positive value on product diversity. Pollock on the other hand is typically considered low-grade groundfish, potentially competing with scallop for processing and shipping resources. [Exvessel scallop prices were found to depend substantially more on landings of other species than own market supply. In FY 2010, both sector and non-sector vessels saw their average price/lb for scallop increase by >25%. Our models suggest that this dramatic price improvement arose from shifts in sector vessel landings of non-scallop species, contributing to 13% of total estimated gains. This result may be slightly confounded by a simultaneous introduction of the scallop general category individual fishing quota (IFQ) program.] Whiting and longfin squid are high-volume species that are commonly frozen when landed due to the high perishability of the former and minimal quality reduction of the latter. Increases in concurrent landings between these two species, which have positive cross-price flexibilities suggesting processing complementarity, improved exvessel prices.
Discussion
Improved exvessel revenues following the introduction of multispecies catch shares arose from advantageous market timing changes by sector vessels. This conclusion suggests DAS created inefficiencies, restricting harvesters' ability to exploit price externalities and to respond to market incentives. The observed dramatic shift in sector vessel joint landing behaviour of multispecies groundfish with commixing, separately regulated stocks additionally suggests DAS may severely constrain catch and landing opportunities when vessels participate in overlapping fisheries with different management systems. This analysis combined counterfactual predictions of individual harvester landings timing, reflective of prior management, with models of exvessel inverse demand in estimating total revenue gains of >$30 000 000.
Though this analysis was only concerned with exvessel revenues, it appears likely that fleet profits were also enhanced as a result of catch share management. With variable operational costs at 38% of gross revenues—the rate used by sector managers when generating initial projections—a no profit increase would require fishing costs to have grown by over 12 million dollars. This seems unlikely, however, as vessels made 10% fewer trips and average trip costs were found to fall well within the range observed under DAS (NEFSC, 2011). Other recent empirical findings have shown tradable access rights are often associated with decreased fishing expenditures, increasing profits through cost reduction (Andersen et al., 2010; Nielsen et al., 2012). The results of this study indicate that in fisheries with flexible prices, constraining management regimes may severely limit potential profits by weakening exvessel revenues. Significant revenue increases resulting solely from a temporal redistribution of supply further question the utility of optimum yield, which has thus far remained an elusive management goal in New England groundfish (Rothschild et al., 2014).
The newfound production flexibility introduced through catch share management may appear inflated if discard of multispecies stocks was frequent. In FY 2010, only 30% of trips were observed, suggesting vessels may have had ample opportunity to discard constraining catch share stocks. This level of observer coverage did represent a dramatic increase from that under DAS management, however, and landings from unobserved trips were monitored, and vessels penalized when discard was suspected. Additionally, it is conceivable that harvesters may have been reluctant to discard during the first year of new management, where each sector member was for the first time jointly liable in fishing activities. Unaccounted discard does not affect the landings revenue results presented here, though it is expected that increased control of illegal discard might reduce revenue gains by curbing flexibility in landings.
Had catch shares not been introduced, multispecies harvesters would likely have experienced drastic cuts in DAS and multispecies fishing income as regulators responded to stock rebuilding requirements specified in the Magnuson–Stevens Fishery Conservation and Management Reauthorization Act of 2006. The two counterfactual methods used relied on information from the most recent prior management year; they are not wholly descriptive of the fishery that would have existed had regulators significantly cut DAS without providing the option to join catch share sectors. Additionally, counterfactual methods altered landings' pace within the set of actual trips taken, which, if endogenous, may confound results. Significant room exists for the development of harvester behavioural models; the work presented here methodologically supplements the current literature, extending methods used in Scheld et al. (2012).
The introduction of catch share management for multispecies stocks significantly affected seasonal exploitation patterns of several separately regulated species. For example, increased participation in the race-to-fish for spiny dogfish, summer flounder, and scup occurred perhaps as a direct result of multispecies catch share management, which made low-quantity multispecies landings profitable by reducing opportunity cost. This implies a mismatch in the scope of regulations and individual economic decisions, which may exacerbate management difficulties already complicated by joint harvest relationships (Squires, 1987; Kirkley and Strand, 1988). Understanding behavioural drivers and the tradeoffs of fishery participation, as well as identifying user groups that overlap regulatory regimes, will ultimately yield superior ex-ante program analysis, reducing management implementation uncertainty and unexpected outcomes (Fulton, 2011).
Including complex micromarket structure in the model of inverse dealer demand allowed for robust estimation of exvessel price and revenue effects resulting from changes in the daily market mix of landed species. Numerous different individual strategies, modifying landings in response to expected price externalities, likely produced the aggregate fleet level behaviour and outcomes observed and discussed. Though analysing each of potentially millions of individual harvester–species daily timing changes is an unrealistic pursuit, conclusions drawn from their combined effect suggest multispecies catch shares afforded individual harvesters increased opportunity and flexibility in supplying market demand. Large exvessel revenue advantages owing to market timing changes have been previously documented after the introduction of rights-based management (e.g. Herrmann, 2005; Scheld et al., 2012), though gains are generally seen to result from a downward-sloping demand and own-species effects. Following the introduction of multispecies catch shares, cross-species price flexibilities were found to play an important role. A counterfactual sensitivity analysis demonstrated that individual species revenue gains often arose from timing changes of several complementary or substitute species, or species groups. Intermediary processing and supply chain structure has often been of limited concern in fisheries policy analysis, though here it was a major driver of policy outcomes.
As a management institution, catch shares have been argued to prevent fishery collapse (Costello et al., 2008) and align individual and collective incentives (Grafton et al., 2006). The transition from DAS to catch shares ensures each multispecies TAC, if adequately enforced, places an upper bound on possible harvest—a necessary, though not sufficient, condition for ecological benefit (Branch, 2009). Multispecies catch shares may wield biological consequence if seasonal changes in targeting intensify harvesting pressure during periods of amplified stock vulnerability, or result in increased mortality for separately managed stocks lacking overall catch limits, shutdown provisions, and/or avoidance incentives. Additionally, multispecies catch shares may impose cultural and societal costs if intended reductions in overcapitalization or consolidation of catch share holdings dramatically alter working waterfronts, community structure, or other factors affecting quality of life for New England residents.
Supplementary data
Supplementary data are available at ICES Journal of Marine Science online.
Funding
Andrew Scheld gratefully acknowledges financial support provided through the National Marine Fisheries Service/Sea Grant Graduate Fellowship in Marine Resource Economics (#NA11OAR4170180). Initial model development was supported by NOAA grant EA133F-09-BAA-17093.
Acknowledgements
The authors would like to thank Drew Kitts and Min-Yang Lee for data assistance and helpful remarks after a presentation of initial findings. Hiro Uchida and Trevor Branch provided insightful comments on an earlier form of the manuscript.
References
Author notes
Handling editor: Claire Armstrong
