Available via license: CC BY 4.0
Content may be subject to copyright.
Contents lists available at ScienceDirect
Fisheries Research
journal homepage: www.elsevier.com/locate/fishres
Fishing methods for Atlantic cod and haddock: Quality and price versus
costs
Geir Sogn-Grundvåg
a,
*, Dengjun Zhang
a,b
, Bent Dreyer
a
a
Norwegian Institute of Food, Fisheries and Aquaculture Research, Norway
b
Business School, University of Stavanger, N-4036, Stavanger, Norway
ARTICLE INFO
Handled by Niels Madsen
Keywords:
Fishing methods
Fish quality
Hedonic price
Sustainability
Fishery management
ABSTRACT
This study explores trade-offs between fish quality, fishing efficiency, costs and profitability across three dif-
ferent vessel groups in the Norwegian groundfish fishery, that is, vessels fishing with bottom trawls, longlines
and Danish seines. The results of hedonic price analysis at the ex-vessel level of the value chain indicate sub-
stantial differences in fish quality as Atlantic cod caught with longlines obtain price premiums of 15.0 % and
12.6 % compared with bottom trawling and Danish seining, respectively, holding other variables constant. For
haddock, longlining obtains a price premium of 20.0 % compared with Danish seining and 13.3 % compared
with bottom trawling. However, despite better quality and prices, the costs of fishing are substantially higher for
longliners than for bottom trawlers and Danish seiners, which explains the differences in profitability favoring
the more technically efficient bottom trawlers and Danish seiners. Policy implications are discussed considering
trade-offs between fish quality, ex-vessel prices and vessel profitability. In a highly regulated fishery such as the
Norwegian groundfish fishery, with individual vessel and vessel group quotas based on historical fishing rights,
policy intervention is important for optimal use of limited fish stocks but is not necessarily straightforward.
1. Introduction
It is well known that different fishing methods influence the quality
of landed fish differently. For example, Rotabakk et al. (2011) found
that Atlantic cod caught by longline had a better overall quality com-
pared with cod caught by trawl. An important quality issue for Atlantic
cod and haddock is discoloration of fillets due to poor draining of blood
(Botta et al., 1986;Margeirsson et al., 2007;Olsen et al., 2014). When
fish is caught in large hauls, which is often the case with bottom
trawling and Danish seine, the fish often dies before bleeding (Olsen
et al., 2014). In addition, it is not unusual for the fish to be kept in
storage bins for hours before bleeding (Olsen et al., 2014). This sim-
plifies the slaughtering process (Van de Vis et al., 2003), but the fish
must be bled while it is still alive to facilitate good drainage of blood.
Differences in fish quality caused by different fishing gear may lead
to different prices at the ex-vessel level of the value chain, as indicated
by studies focusing on Atlantic cod in Canada (Lee, 2014) and Norway
(Asche et al., 2015), Baltic cod in Sweden (Blomquist et al., 2015;
Hammarlund, 2015), and bluefin tuna in Hawaii (McConnell and
Strand, 2000) and Japan (Carroll et al., 2001). Different fishing
methods may also impact the environment in different ways, which
may influence the market value of the fish. Sogn-Grundvåg et al. (2013,
2014) estimated price premiums in the UK grocery retail market in the
10.4–24.6 % range for line-caught chilled and frozen Atlantic cod and
haddock compared with fish caught with other fishing methods. In
addition, Sogn-Grundvåg et al. (2019a) showed enhanced product
longevity for products of Atlantic cod and haddock with the line-caught
label compared to similar products without the label, implying reduced
costs. This suggests that longlining is preferred for Atlantic cod and
haddock, at least in the UK market. These price premiums for line-
caught fish may relate to better product quality (Rotabakk et al., 2011)
and also to the well-documented negative environmental effects of
other fishing methods, including the detrimental effect that bottom
trawling may have on the seabed and habitat (Puig et al., 2012).
Knowledge of these issues among retail management and consumers
alike may have influenced the demand for line-caught fish (Sogn-
Grundvåg et al., 2013;Zhang et al., 2018a).
Vessels using different fishing methods also differ in technical effi-
ciency, influencing the costs of fishing, which may vary between vessels
and vessel groups (Guttormsen and Roll, 2011;Asche and Roll, 2018).
And, importantly, the cost of fishing and fisher behavior may be in-
fluenced by the way fisheries are managed, for instance, by restrictions
related to vessel size and gear types to limit fishing effort and by sub-
sidizing input factors such as capital and fuel to support fishers’
https://doi.org/10.1016/j.fishres.2020.105672
Received 17 January 2020; Received in revised form 11 June 2020; Accepted 13 June 2020
⁎
Corresponding author.
E-mail address: geir.sogn-grundvag@nofima.no (G. Sogn-Grundvåg).
Fisheries Research 230 (2020) 105672
0165-7836/ © 2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/BY/4.0/).
T
incomes (Asche et al., 2020). Thus, an issue with substantial policy
implications is the potential trade-offbetween fish quality, prices
gained and the costs of fishing caused by gear restrictions and quota
allocations between vessel groups fishing with different gear. For ex-
ample, the most efficient fishing methods and vessels may provide fish
of low quality but may still be more profitable than less technically
efficient fishing methods providing fish of high quality and at better
prices. This may be further complicated by the fact that fishing vessels
often hold quotas for several different species with sometimes over-
lapping seasons which may lead to inefficiency because fishers may
choose to fish too much of some species relative to other species, as
found by Asche and Roll (2018) for the Norwegian trawler fleet. This
indicates that fishery management excessively focused on input re-
strictions and profitability in the harvesting sector may result in land-
ings of fish of reduced quality, which may affect negatively the op-
portunities for value-adding and marketing in subsequent links in the
value chain (Larkin and Sylvia, 1999;Homans and Wilen, 2005). Thus,
regulators risk missing out on the FAO’s request to utilize fish stocks in
a way that contributes to the nutritional, social and economic value of
wild fish stocks (Food and Agriculture Organization (FAO, 2018).
The contribution of this study is to investigate the ex-vessel prices
obtained by vessel groups using bottom trawl, Danish seine and long-
lines while also considering the costs and profitability of the same
vessel groups. Covering a period of 9 years (2009–2017), the study
analyzes 26,639 transactions for Atlantic cod and 10,489 transactions
for haddock, which consist of 464,311 tons of Atlantic cod and
240,101 tons of haddock. This covers about all the frozen cod and
haddock landed by the three vessel groups under analysis. The fish is
headed and gutted and frozen at sea to preserve the quality and allow
longer trips. The fish is typically packed and frozen in 20-kilogram
boxes and sold for secondary processing in downstream markets.
Reported ex-vessel prices obtained by the three vessel groups are dis-
cussed, considering accounting data from the Norwegian Directorate of
Fisheries for the same vessel groups. In this way, important implications
for fishery management can be highlighted and discussed.
2. Material and methods
2.1. Fishery management
The Norwegian groundfish fishery is the most valuable fishery in
Norway, and by far the most important species in this fishery is Atlantic
cod, followed by haddock. The fishing fleet participating in the
groundfish fishery is diverse, ranging from small coastal vessels fishing
with jig machines, gillnets and hand-baited longlines, delivering fresh
catches to local fish plants daily, to large oceangoing bottom trawlers
and longliners, freezing their catch at sea. In addition, medium-sized
and large coastal vessels using gillnets or Danish seines deliver fresh
daily catches, and some of these also freeze a share of their catch at sea.
The Norwegian fishery management system is based on several
policy goals, including preventing overfishing and maintaining sus-
tainable coastal communities (Standal and Aarset, 2008). In addition,
resource rent and fleet profitability are key objectives of the manage-
ment system, as in many other countries (Flåten and Heen, 2004;
Hannesson, 2013;Zhang et al., 2018b).
The groundfish fishery is regulated on a single species basis with a
total allowable catch (TAC) for the main groundfish species. The yearly
TAC for Atlantic cod and haddock is set based on scientific advice from
the International Council for the Exploration of the Seas (ICES).
Although scientific recommendations are given within a range, the
quotas tend to fluctuate more or less in accordance with the variation in
biomass estimates, which fluctuated substantially during the period
studied.
The management system divides the large and diverse fishing fleet
into a number of different vessel groups based on gear type, target
species and vessel sizes. The TAC for the main species is allocated
among vessel groups based on the so-called “trawl ladder”, the objec-
tive of which is to provide predictability and stability in quotas for
smaller vessels and, as such, to contribute to regional policy goals
(Guttormsen and Roll, 2011). When fish stocks are low or modest, a
higher share of the TAC is allocated to smaller coastal vessels. Quotas
cannot be transferred between vessel groups. Quotas are, however,
transferable within vessel groups as quotas can be transferred by pur-
chasing a vessel, removing it from the fishery and transferring the quota
to the acquirer’s vessel. Vessel quotas have been raised several times to
further stimulate consolidation, reduce the overall fishing capacity and
enhance profitability. This has been successful in the sense that over-
capacity has been reduced and the profitability of the remaining vessels
has improved (Zhang et al., 2018b). The size of vessel quotas is re-
stricted and differs among vessel groups, with bottom trawlers holding
the largest quotas for cod and haddock.
In addition to vessel quotas, the fishery is strictly regulated by a
mixture of measures, such as time and location closures and various
technical regulations relating to vessel size and fishing gear, some of
which are described in more detail below. The Norwegian Food Safety
Authority enforces technical legislation regarding how to handle and
store fish on board the vessels in order to preserve the quality of the
fish.
2.2. The ex-vessel market
The ex-vessel sale of wild-caught fish in Norway is legally protected
by the Raw Fish Act and is organized by sales organizations that have
the exclusive right to coordinate the primary sale of fish. This includes
the right to set minimum prices to secure the fishers a price that reflects
the market prices and to avoid powerful buyers using their bargaining
power to set prices that are too low for small independent fishers
(Holm, 1995).
Several different sales organizations exist, covering different fish
species or geographical regions according to where the fish is landed.
The on-board frozen fish included in this study is sold through the
largest sales organization, the Norwegian Fishermen’s Sales
Organization, which records all transactions in the market and has
provided the data for this study.
Fishers selling their frozen Atlantic cod and haddock through the
Norwegian Fishermen’s Sales Organization can choose an online auc-
tion organized by the sales organization or direct sales (for a more
detailed description of the ex-vessel market for groundfish, see Helstad
et al. (2015) and Pettersen et al., 2018). There are two large vertically
integrated companies that own both vessels (bottom trawlers) and
plants for primary processing (that is, gutting and packing whole fish or
salting or drying the fish for various export markets) and secondary
processing (that is, making consumer-ready products), with in-house
sales and marketing activity. There are also a few vertically integrated
companies, in which fishing companies have acquired onshore pro-
cessing facilities. The vertically integrated companies have different
production and sales strategies for their vessels, with some buying their
own fish directly and others selling their fish to other buyers directly or
through the auction. Some of the fish is used in their own processing,
and some is exported. It should be noted that the share remuneration
payment system whereby the crew receives a fixed share of the rev-
enues rather than a fixed wage (see, e.g., McConnell and Price,
2006)—and the strong position of the Norwegian Seafarers’Union—-
makes it difficult for vertically integrated companies to buy the fish
directly from their own vessels at low prices. If they do, skilled crew
members may leave for jobs on other vessels or industries. Autonomous
fishing companies sell fish both directly to buyers and through the
auction. Some have developed long-term relationships with processing
companies to which fish is sold directly. These differences in the re-
lationships between buyers and sellers may lead to market imperfec-
tions, thereby influencing the prices (Gobillon et al., 2017).
Frozen Atlantic cod and haddock are sold for secondary processing
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
2
worldwide. In 2017, 69,291 tons of frozen whole Atlantic cod (headed
and gutted) were exported (source: the Norwegian Seafood Council). In
addition, some of the frozen Atlantic cod is purchased and processed by
processors in the Norwegian clipfish industry, in which cod of a large
size is in demand. In the UK, frozen Atlantic cod and haddock are
mostly thawed and filleted by wholesalers and used for different types
of chilled and frozen consumer packs for the large grocery retailers
(Sogn-Grundvåg et al., 2013,2014).
2.3. The three vessel groups
In this study, three vessel groups are in focus, that is, oceangoing
bottom trawlers and longliners as well as large coastal vessels fishing
with Danish seines. The bottom trawlers are licensed to fish with
bottom trawling only and are allocated about 30 % of the yearly
Norwegian TAC for Atlantic cod and haddock. In 2017, the trawler
group consisted of 35 vessels, a decrease from 51 vessels in 2009, in-
dicating substantial consolidation in the group. Bottom trawlers mostly
freeze their catch on board as headed and gutted fish. Freezing is
conducted to preserve the quality and to allow longer trips. In addition,
the prices for frozen headed and gutted Atlantic cod have generally
been higher than those for fresh fish (Pettersen et al., 2018). The
trawlers are not allowed to fish within 12 nautical miles of the Nor-
wegian coastline. Some of the trawlers have quotas for prawns (Pan-
dalus borealis) in addition to groundfish. It should be noted that new
technology is being developed that may improve the quality of fish,
such as on-board live storage, so that the fish can be slaughtered ef-
fectively (Digre et al., 2017), and improved trawl technology (Brinkhof
et al., 2018).
As of 2017, the group of oceangoing longliners consisted of 20
vessels, a decrease from 33 vessels in 2009, implying considerable
consolidation in the group. Longliners can use gillnets in addition to
longlines. Gillnets are used to fish for saithe, as saithe is not a typical
bottom feeder, thus making gillnets more effective than longlines. The
vessels in this group are allocated about 8% of the yearly Norwegian
TAC for Atlantic cod and haddock. The longliners are not allowed to
fish with trawls, but in 2018, they were allowed to use Danish seines to
fish their quotas for Atlantic cod, haddock and saithe. This liberal-
ization in gear use came about after pressure from the Norwegian
Fishermen’s Association, which argued to the Norwegian Ministry of
Trade, Industry and Fisheries that, due to the high costs of fuel and bait,
this vessel group should be allowed to use Danish seines on account of
the lower costs associated with this fishing method. This illustrates how
fishers try to adapt to and even change regulatory constraints (Salas and
Gaertner, 2004). The cost of changing from longline to Danish seine is,
however, substantial, as these two fishing methods require rather dif-
ferent technologies. Interestingly though, two oceangoing longliners
are currently (2020) being built and set up for both longline and Danish
seines.
The longliners fish with around 50–70,000 hooks, which are baited
automatically when setting the line. The oceangoing longliners are not
allowed to fish within 4 nautical miles of the Norwegian coast, and their
access is restricted in certain areas outside the 4-mile limit to avoid
conflicts of interest with smaller coastal vessels. Most of the oceangoing
longliners freeze their catch on board as headed and gutted fish.
The group of large coastal vessels (above 21 m in length) consisted
of 32 vessels in 2017, a decrease from 35 in 2009. These vessels can use
several types of gear, including longline, gillnets, Danish seine and
purse seine but not trawl. The fleet is allocated about 8% of the yearly
Norwegian TAC for Atlantic cod and haddock. The size of the quotas
depends on the length of the vessel. Only a few of the vessels in this
group freeze the fish on board, and these are mainly the newest and
largest vessels. Several of the vessels in this group have licenses to fish
for pelagic fish, such as herring and mackerel, in addition to Atlantic
cod, haddock and saithe. For the pelagic species, the purse seines are
used. Of the three vessel groups in focus here, this group has the
greatest flexibility in terms of the choice of fishing method. Compared
with trawlers and longliners, it also has the broadest spectrum of quotas
(species) and few restrictions in terms of fishing areas.
The three vessel groups described above apply different fishing
methods. However, the bottom trawl and Danish seine methods share
some key characteristics, such as a very large capacity to catch fish in a
single haul. It is not unusual to keep the fish in storage bins for hours
before bleeding and gutting (Olsen et al., 2014), and to simplify the
slaughtering process, bleeding is often done after suffocation in air (van
de Vis et al., 2003), which implies reduced fish quality (Botta et al.,
1986;Rotabakk et al., 2011;Olsen et al., 2013,2014). Longliners, on
the other hand, when pulling in the longline, catch only one fish at a
time. This makes it possible to bleed and process each fish immediately
after catching, which is the main reason for the higher quality provided
by longliners (Rotabakk et al., 2011). Longliners, however, are less
technically efficient than bottom trawlers and Danish seiners, implying
higher costs of fishing. This is illustrated in Fig. 1, which shows the
average costs per kilogram of fish landed for the three vessel groups.
Fig. 1 shows that the average cost of fishing is substantially higher
for longliners than for bottom trawlers and Danish seiners. The cost
differences are relatively stable from 2010 to 2015 but increase after
that in disfavor of longlining. The costs of fishing are about 30 % higher
for longliners than for bottom trawlers and Danish seiners. However,
because all vessels included fish for several fish species and because the
profitability survey of the Directorate of Fisheries (2020) does not
provide accounting data for individual species, the graphs in Fig. 1 must
be interpreted with caution. For example, some of the bottom trawlers
also fish for prawns, which generally is a less technically efficient form
of fishing than trawling for cod and haddock. This indicates that the
cost of fishing for haddock and cod is somewhat lower than what is
shown in Fig. 1. Further, several Danish seiners also have large quotas
for mackerel and herring caught with purse seine, which is a very
technically efficient gear. This indicates that the costs of fishing for cod
and haddock is somewhat higher for this vessel group than what is
shown in Fig. 1. The reduction in fishing costs for Danish seiners after
2015 is probably due to larger catches of mackerel and herring com-
pared to cod and haddock.
Despite the back-of-the-envelope nature of these calculations, it
seems reasonable to argue that longliners have the highest costs of
fishing when fishing for cod and haddock. This implies that bottom
trawlers and Danish seiners can, to a larger extent than longliners, earn
profits through swift and intense fishing strategies where lower quality
and prices are compensated for by large quantities caught at low cost.
Such fishing strategies are relevant in a multi-species fishery such as
this, with sometimes overlapping seasons and where the number of
operating days are very high, but the resultant fish quality may be
compromised (Bertheussen and Dreyer, 2019).
Fig. 1. The average costs per kilogram (NOK) of landed fish (all species) for the
three vessel groups for 2010–2018. Source: Directorate of Fisheries (2020).
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
3
Table 1 shows the average turnover, key costs relating to fishing
methods and the operating margin for the three vessel groups for the
time period covered by the study. An inspection of Table 1 shows that
the average turnover differs substantially among the three groups, with
trawlers (NOK 87 million) having a higher turnover than longliners
(NOK 46 million) and Danish seiners (NOK 16 million), reflecting
substantial differences in the size of the quotas. The table also shows
that the average operating margin was 19 %, 9 % and 11.9 % for
bottom trawlers, longliners and Danish seiners, respectively, indicating
differences in the technical efficiency and prices gained. For 2017, the
operating margin for bottom trawlers and Danish seiners was 28.4 %
and 21.6 %, respectively, much higher than the 12.3 % operating
margin gained by oceangoing longliners in 2017.
Calculations from the data used in Table 1 show that for longliners,
bottom trawlers and Danish seiners, the average cost of fuel constituted
9.5 %, 17.7 % and 15.2 % of the total costs, respectively, thereby im-
plying that the fuel costs are substantially lower for longliners than for
bottom trawlers and Danish seiners. In terms of fuel use, it has been
found that the energy efficiency of all vessel groups in Norwegian
fisheries has improved in recent years (Jafarzadeh et al., 2016). For the
period 2003–2012, the average was 0.322–0.354 and 0.265 L of fuel per
kilo of fish landed for bottom trawlers and longliners, respectively
(Jafarzadeh et al., 2016). Danish seiners are part of a group of large
coastal vessels (above 21 m in length) which also include large coastal
vessels fishing with gillnets and longlines. For this group, the average
fuel consumption per kilogram of fish landed for the period 2003–2012
was 0.058 L of fuel (Jafarzadeh et al., 2016). It should also be noted
that fuel use varies substantially with catch rates (Schau et al., 2009).
During the period studied here, the Norwegian fishing fleet paid a
substantially lower CO
2
fuel surcharge compared to other commercial
vessels (e.g., vessels involved in fish farming), namely NOK0.29 per
liter of fuel for fishing vessels versus NOK1.65 per liter for other vessels.
This subsidy represents about 20–25 % of the fuel costs in the period
studied here (for a detailed description of the Norwegian fuel tax
concession, see Isaksen et al. (2015)). As of January 1, 2020, fishing
vessels must pay the same CO
2
fuel surcharge as other vessels, thus
implying a substantial increase in fuel costs for vessels with high fuel
consumption, such as trawlers. However, a compensation scheme will
be introduced in which those vessels with the lowest fuel consumption
within each vessel group will receive the highest compensation.
The average cost of labor (salary) was higher for longliners (43.6 %
of total costs) than for bottom trawlers (37.1 % of total costs). In ad-
dition, the cost of bait, ice and packaging accounted for 7.3 % of the
total costs for longliners versus less than 1% for bottom trawlers. Thus,
even though the fuel costs are lower for longliners than for bottom
trawlers, the additional cost of labor and bait implies a cost
disadvantage for longliners. It should be noted that the costs of labor
and bait are variable costs which depend on the size of quotas, sug-
gesting that the larger the quotas, the more man hours and bait are
required. Interestingly, Table 1 shows that the number of operating
days is very high for both bottom trawlers and longliners, implying that
their catching capacity is close to fully utilized.
Figs. 2 and 3 show the quantity of frozen headed and gutted Atlantic
cod and haddock, respectively, sold by the vessels in the three vessel
groups included in the nine-year period covered by the study. The
graphs in Figs. 2 and 3 reflect large variations in the yearly Norwegian
TACs for Atlantic cod and haddock. The graphs also show that bottom
trawlers are the most important vessel group in our data in terms of sold
frozen quantity of both cod and haddock.
2.4. Data and descriptive statistics
The transaction data used in this study include sales of on-board
frozen Atlantic cod and haddock from oceangoing bottom trawlers and
longliners as well as from large coastal vessels using Danish seines. In
addition, a fourth group of smaller vessels is used in the econometric
modelling as a base category for comparisons. This group consists of
several vessels fishing with different methods, such as gillnets, traps
and pots. The total ex-vessel value of frozen Atlantic cod and haddock
for the nine-year period covered by the data was NOK 1.9 billion and
722 million, respectively, in 2017 (the average exchange rate for 2017
was NOK 1 = USD 0.1209/EUR 0.1071).
Figs. 4 and 5 show the yearly average prices by fishing methods for
Atlantic cod and haddock. Fish caught with longline fetch the highest
Table 1
a
Average yearly turnover (NOK), key costs (NOK), profitability and operating days for the three vessel groups for 2009-2017.
Source: Directorate of Fisheries (2020).
Bottom trawlers SD Longliners SD Danish seiners SD
Turnover 87,266,127 29,341,411 46,207,512 15,456,942 15,972,066 6,064,068
Salary 25,618,196 8,182,403 18,102376 7,126,327 5,634,277 1,964,444
Fuel 12,250,122 2,488,370 3,939,830 823,156 1,166,995 353,409
Bait, ice, packaging 712,906 153,395 3,021,599 755,243 101,142 96,385
Other costs 38,581,224 9.990,096 1,162,942 271,840 653,809 127,603
Total costs 68,989,286 17,667,728 26,226,748 8,303,135 7,670,408 2,407,564
Operating margin (%) 19,0 6,7 41,522,774 12,303,474 13,786,236 4,263,031
Operating days 311,3 13,0 333,4 11,2 215,6 7,9
No. vessels in sample 32,2 3,8 9,0 3,8 11,9 7,4
No. vessels in population 41,1 6,1 15,3 2,1 14,2 2,6
a
The numbers are calculated from the yearly profitability survey of Norwegian fisheries conducted by the Directorate of Fisheries (Directorate of Fisheries, 2020).
To save space, only the most relevant costs are included. We use the operating margin as a measure of profit rather than the more commonly used return on
investments to reflect better how the cost of fishing methods relates to earnings. Even though cod and haddock are the main species, bottom trawlers also catch other
groundfish species and some also have quotas for prawns, which may influence their profitability. Oceangoing longliners mainly fish for cod, haddock and saithe,
whereas several Danish seiners fish for pelagic species, such as mackerel (Scomber scombrus) and herring (Clupea harengus) in addition to groundfish.
Fig. 2. Yearly quantities (tons) of frozen headed and gutted Atlantic cod sold by
vessels in the three vessel groups.
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
4
prices, followed by bottom trawl and Danish seine. However, the gap in
prices between longline and the two other capture methods is larger for
haddock than for Atlantic cod. The figures also show that the price
differences between fishing methods were relatively stable over the
nine-year period. The econometric modelling will estimate the differ-
ences in prices between fishing methods, holding other factors such as
the fish size, seasonality, seller and buyer heterogeneity and so on
constant.
Each market transaction in the data includes information about the
species (Atlantic cod or haddock), total weight of the lot in kilograms,
fishing method, average size of the fish, vessel name, buyer name, price
and so on. Table 2 presents the descriptive statistics of the variables
included in the econometric models outlined below. The table shows
that the average price of Atlantic cod was substantially higher than that
for haddock. Note also that the standard deviations for the Atlantic cod
and haddock prices are substantial, probably reflecting seasonal and
yearly price differences due to seasonal differences in the traded
quantity and large yearly changes in the TAC for both species. In ad-
dition, the standard deviations for the daily quantity and transaction
quantity are substantial, indicating large variations in the traded
quantity. For Atlantic cod, the average fish size is 2.70 kg (headed and
gutted weight), with a substantial standard deviation indicating a large
size variation between lots. For haddock, the fish size is a dummy
variable with an average weight below 0.8 kg as the base.
Table 2 also shows two groups of dummy variables, that is, dum-
mies for quality and fishing methods. The reported mean for each
dummy variable is the number of observations (transactions) within
each category as a proportion of the total number of observations.
Quality-1 is a dummy that takes the value one for transactions of large
haddock (> 0.8 kg, headed and gutted weight) caught with longline,
which has a specific minimum price. There is no specific minimum
price for Atlantic cod caught by longline. Quality-2 for Atlantic cod and
haddock refers to fish of regular quality as opposed to fish that was
downgraded due to poor quality. Under the Raw Fish Act, buyers may
downgrade fish based on its quality and reduce the price by 40 %
compared with the minimum price. As can be seen from Table 2, there
is a larger share of Atlantic cod of regular quality than of haddock. The
table also shows that bottom trawling of Atlantic cod and haddock
accounted for 62 % and 59 %, respectively, of the transactions during
the sample period. Longliners accounted for 25 % and 31 % of the
Atlantic cod and haddock catches, respectively. Table 2 also shows that
42 % and 29 % of the Atlantic cod and haddock, respectively, was sold
by auction as opposed to by direct sales.
2.5. Model and econometric analysis
To examine the influence of fishing methods on the ex-vessel prices
of frozen Atlantic cod and haddock, this study follows the hedonic price
modelling literature which has been applied to estimate the value of
seafood attributes along the value chain, including at the initial point of
landing and port auctions (McConnell and Strand, 2000;Carroll et al.,
2001;Kristofersson and Rickertsen, 2007;Lee, 2014;Asche et al., 2015;
Gobillon et al., 2017), wholesale markets (Kristofersson and Rickertsen,
2004;Asche and Guillen, 2012) and the retail level (Roheim et al.,
2011;Asche et al., 2015).
Hedonic price modelling relies on characteristics theory, which
Fig. 3. Yearly quantities (tons) of frozen headed and gutted haddock sold by
vessels in the three vessel groups included.
Fig. 4. Yearly average prices per kilogram (NOK) by fishing methods for frozen
headed and gutted Atlantic cod.
Fig. 5. Yearly average prices per kilogram (NOK) by fishing methods for frozen
headed and gutted haddock.
Table 2
Descriptive statistics of the variables included in the econometric models.
Variable Atlantic cod Haddock
Mean Std deviation Mean Std deviation
Price (NOK) 21.3 5.8 16.8 5.6
Daily quantity (kg) 375,395 333,809 16,437 250,032
Transaction quantity (kg) 17,430 30,366 22,891 35,967
Fish size (kg) 2.70 2.42 0.50
Quality (base: downgraded)
Quality 1 0.19
Quality 2 0.94 0.75
Downgraded (base) 0.06 0.06
Fishing methods
Bottom trawl (base) 0.62 0.59
Longline 0.25 0.31
Danish seine 0.10 0.07
Other 0.03 0.02
Sales methods
Auction 0.42 0.29
Direct sale (base) 0.58 0.71
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
5
assumes that consumers derive utility directly from the quality attri-
butes inherent in a good (Lancaster, 1966;Rosen, 1974). Thus, the
actual price of a good can be considered as the sum of the implicit
prices of those attributes (Rosen, 1974). This indicates that the price of
an individual lot of Atlantic cod or haddock depends on its character-
istics, such as the average size of the fish, its quality and the fishing
method. Fish prices may, however, also be influenced by factors such as
the transaction quantity, daily sales quantity and seasonality (Kirman
and Vriend, 2001;Guillioni and Bucciarelli, 2011;Fluvià et al., 2012;
Gobillon et al., 2017;Sogn-Grundvåg et al., 2019b). In past research,
hedonic price modelling has used either the linear price or the loga-
rithmic price as the dependent variable. Our test results, based on
Vuong (1989) likelihood ratio test, show that the model with the
logarithmic price as the dependent variable fits the data better than the
model with linear price formulation. Thus, the basic model (Model A)
is:
∑
∑∑
=+ +
++ +
+++ +
++ +
=
==
p a b Daily Quantity b Transaction Quantity
b Fish Size c Quality d Longline
d Danish Seine d Other Fishing Methods e Auction
f Year w Month Residual
log( ) log( _ ) log( _ )
__2
___
iii
ioooi i
iii
oooi oooi i
01 2
31
2
,1
23 1
2
9
,2
12
,(1)
where, for either the cod or the haddock model, iindexes transactions,
and log is the natural logarithm function; P
i
is the prices, which are
deflated by the Norwegian consumer price index for food; Daily
Quantity is the daily traded quantity; Transaction Quantity is the quan-
tity of each transaction. The two main fishing methods (Longline and
Danish Seine) and other fishing methods together are dummies, with
bottom trawl as the base. The dummy variable for a particular fishing
method may also capture the impact on price of some unobservable
factors which are related to this fishing method. Auction is a dummy for
Atlantic cod sold by auction with contract sales as a base. The Year and
Month dummies are included in the model to control for any seasonality
patterns and yearly variations in prices due to changes in catches and
the TAC. The error term Residual captures any other unobserved factors,
such as seller and buyer heterogeneity, that might influence the prices
(Gobillon et al., 2017).
Regarding fish size and quality dummies, the measures are different
for cod and haddock. For the cod model, Fish Size is the average size of
the fish in kilograms. For haddock, the size is registered as either larger
or smaller than 0.8 kg. Thus, Fish Size in the haddock model is a dummy
variable, which is one for a weight larger than 0.8 kg and zero other-
wise. In the two models, Quality-2 is for set for fish with regular quality
with downgraded fish as the base. Additionally, Quality-1 is an addi-
tional quality dummy reflecting the specific minimum price set for large
haddock (> 0.8 kg) caught with longline. Thus, Quality-1 is an inter-
action term between Fish Size and Longline.
Since the fish prices are probably affected by market imperfections
caused by heterogeneity among buyers (e.g., different degrees of will-
ingness to pay for different fish quality and fish size, depending on the
markets and the customers whom they serve (Sogn-Grundvåg et al.,
2019b), sellers (e.g., differences in skipper and crew skills, which may
result in differences in fish quality) and buyer-seller relationships (e.g.,
some buyers and sellers are vertically integrated, which may influence
the prices), we modify the basic model (A) by including dummies for
sellers, buyers and seller-buyer pairs. We include dummies for the 50
largest buyers in terms of the quantities purchased during the 9-year
period (and the 50 largest sellers) to avoid multicollinearity. These
buyers and sellers accounted for more than 95 % of the total traded
quantity in the period covered by the data. For seller-buyer pairs,
dummies for 99 pairs are included in the model. In terms of the traded
quantity, these are the most significant pairs and accounted for about
36 % of the total transaction quantity over the sample period for cod
and about 49 % for haddock. This leads to the following five models:
Model A: the basic model represented by Eq. (1)
Model B: Model A with additional seller dummies
Model C: Model A with additional buyer dummies
Model D: Model A with additional seller and buyer dummies
Model E: Model D with additional dummies for buyer–seller pairs
Of the five models, Model A is nested under Model B and Model C,
which are further nested under Model D. Model E is a complete speci-
fication. F-tests are used to test Models A–D until rejection occurs.
Since the dependent variable is expressed in logarithm form, the
estimated coefficients for Daily Quantity and Transaction Quantity are
explained as elasticities, that is, percentage changes in price (b
1
%or
b
2
%) in response to a 1% change in the daily quantity or transaction
quantity. For the dummy variables, we follow Kennedy (1981) and
derive the percentage impact on prices by taking the standard deviation
into account. For example, for a significant estimation d
2
, the impact
(price premium) of fish caught by longliners relative to the base (fish
caught by bottom trawler) is calculated as:
100[exp(d
2
-0.5*V(d
2
))-1]% (2)
where the notation exp is the exponential function of V(d
2
), which is the
square of the standard error for d
2
.
2.6. Functional form and match effect
As the estimation results are probably subject to the functional form
of the models, we have relied on Vuong (1989) likelihood ratio test and
chose the specification with the logarithmic price as the dependent
variable. As a robustness check, the Box-Cox transformation, the other
most common test, is applied to test/choose the functional form.
In the models, we include dummies for the dominant buyers and
sellers to control for buyer and seller heterogeneity. In the auction, the
match effect between buyer and seller is probably more important than
the heterogeneity of individual buyers and sellers. In our sample, there
are 3559 seller-buyer pairs for Atlantic cod and 1631 for haddock. In
the models, we include the dummies for the top 99 pairs to control for
match effect; they account for about 36 % of total transaction quantity
in the sample period for cod and about 49 % for haddock. Therefore,
the other pairs are aggregated to the base group. Ignoring heterogeneity
between these pairs may lead to low explanatory power for the model
and may affect estimates of fishing methods (and other determinants).
We follow Gobillon et al. (2017) and use the Frisch-Waugh theorem to
estimate the models after controlling for the whole match effect.
3. Results
3.1. Testing models
Tables 3 and 4 present the econometric results from estimating the
hedonic price models for Atlantic cod and haddock, respectively. For
both models, the Newey-West covariance matrix is applied to correct
for heteroskedasticity and serial correlation in the error terms, from
which the robust standard error is estimated for inference. The adjusted
R
2
ranges are between 0.707 and 0.73 for the Atlantic cod models and
0.718 and 0.749 for the haddock models, indicating a high level of
overall goodness of fit for these models. Moreover, the models with
more variables have a greater adjusted R
2
than the models with fewer
variables. This indicates the raised goodness of fit due to the additional
variables, such as seller dummies, buyer dummies and dummies for
seller-buyer pairs. As shown in Table 5, the F-test results indicate that
for the two series of models, the restricted models (with fewer vari-
ables) are firmly rejected. Thus, Model E fits the data better than the
other models. Next, we focus on the estimation results of Model E and
calculate the price premiums/discounts using Eq. (2).
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
6
3.2. Estimation results of the cod model
For Atlantic cod, Model E in Table 2 shows that Daily Quantity is
significant and negative (-0.0032), which implies that a 10 % increase
in the daily transaction quantity would lead to a 0.032 % reduction in
price. On the other hand, Transaction Quantity is significant and posi-
tive, suggesting that a 10 % increase in the transaction quantity would
lead to a 0.016 % increase in price. These estimates are small and
indicate that Atlantic cod prices are hardly affected by the daily traded
quantity and the size of deals. Fish Size is measured in kilograms, and
thus the estimate of 0.0203 indicates that the price of Atlantic cod
would increase by 2.03 % if the fish size increased by 1 kg. Buyers’
willingness to pay for Atlantic cod of regular quality is reflected in the
estimate for Quality-2, which indicates a price premium of about 44.1 %
compared with downgraded Atlantic cod.
The dummies for longline and Danish seine are significant.
Table 3
Econometric results from estimating the hedonic price models for Atlantic cod.
Variable Model A Model B Model C Model D Model E Impact of dummies in Model E
Est. NW S.E. Est. NW S.E. Est. NW S.E. Est. NW S.E. Est. NW S.E.
Intercept 2.5442 *** 2.5528 *** 2.5404 *** 2.5414 *** 2.5406 ***
[0.0159] [0.0161] [0.016] [0.0161] [0.0161]
log (Daily-Quantity) −0.0027 *** −0.0026 *** −0.0034 *** −0.0032 *** −0.0032 ***
[0.0008] [0.0008] [0.0008] [0.0008] [0.0008]
log (Transaction-Quantity) 0.0008 * 0.0009 * 0.0017 *** 0.0016 *** 0.0016 ***
[0.0005] [0.0005] [0.0005] [0.0005] [0.0005]
Fish-Size 0.0255 *** 0.0255 *** 0.0204 *** 0.0203 *** 0.0203 ***
[0.0004] [0.0004] [0.0005] [0.0005] [0.0005]
Quality-2 0.3749 *** 0.3754 *** 0.3676 *** 0.3660 *** 0.3654 *** 44.1 %
[0.0083] [0.0083] [0.0086] [0.0085] [0.0085]
Longline 0.0854 *** 0.0650 *** 0.0838 *** 0.0751 *** 0.0745 *** 7.73 %
[0.0025] [0.0047] [0.003] [0.0049] [0.0049]
Danish-Seine −0.0593 *** −0.0746 *** −0.0616 *** −0.0686 *** −0.0681 *** −6.58 %
[0.0032] [0.0049] [0.0034] [0.005] [0.0051]
Other fishing methods −0.0173 *** −0.0015 −0.0407 *** −0.0099 −0.0062
[0.0055] [0.0078] [0.006] [0.0081] [0.0083]
Auction 0.0284 *** 0.0325 *** 0.0234 *** 0.0294 *** 0.0300 *** 3.05 %
[0.002] [0.0021] [0.0023] [0.0024] [0.0024]
With time dummies Yes Yes Yes Yes Yes
With seller dummies No Yes No Yes Yes
With buyer dummies No No Yes Yes Yes
With seller-buyer dummies No No No No Yes
R
2
0.7079 0.7125 0.7246 0.7292 0.7325
Adj. R
2
0.7076 0.7117 0.7238 0.7279 0.7302
Notes: ***, ** and * indicate significance at the 0.01, 0.05 and 0.10 level, respectively. NW S.E. represents Newey West standard error.
Table 4
Econometric results from estimating the hedonic price models for haddock.
Variable Model A Model B Model C Model D Model E Impact of dummies in Model E
Est. S.E. Est. S.E. Est. S.E. Est. S.E. Est. S.E.
Intercept 1.9480 *** 1.9115 *** 1.8728 *** 1.8513 *** 1.8586 ***
[0.0296] [0.0326] [0.0305] [0.0349] [0.0355]
log (Daily-Quantity) −0.0033 ** −0.0033 −0.0027 * −0.0027 * −0.0029 **
[0.0015] [0.0015] [0.0015] [0.0015] [0.0015]
log (Transaction-Quantity) 0.0078 *** 0.0082 *** 0.0054 *** 0.0055 *** 0.0058 ***
[0.0013] [0.0013] [0.0011] [0.0011] [0.0011]
Fish-Size 0.0039 0.0054 0.0039 0.0063 0.0060
[0.0042] [0.0042] [0.004] [0.004] [0.004]
Quality-1 0.6936 *** 0.6893 *** 0.5652 *** 0.5614 *** 0.5575 *** 74.6 %
[0.0167] [0.0167] [0.0197] [0.0201] [0.0201]
Quality-2 0.5528 *** 0.5521 *** 0.4289 *** 0.4286 *** 0.4269 *** 53.2 %
[0.0146] [0.0146] [0.0182] [0.0186] [0.0187]
Longline 0.1570 *** 0.1671 *** 0.1507 *** 0.1493 *** 0.1402 *** 15.0 %
[0.0059] [0.0117] [0.0075] [0.0133] [0.0144]
Danish-Seine −0.0650 *** −0.0319 ** −0.0571 *** −0.0336 ** −0.0301 ** −3.0 %
[0.0079] [0.0135] [0.009] [0.0138] [0.0146]
Other fishing methods 0.0049 0.0348 0.0175 0.0356 0.0301
[0.0251] [0.0266] [0.0224] [0.0242] [0.0243]
Auction 0.0052 0.0122 0.0296 *** 0.0360 *** 0.0354 *** 3.60 %
[0.0043] [0.0046] [0.0047] [0.0051] [0.0053]
With time dummies Yes Yes Yes Yes Yes
With seller dummies No Yes No Yes Yes
With buyer dummies No No Yes Yes Yes
With seller-buyer dummies No No No No Yes
R
2
0.7190 0.7237 0.7455 0.7491 0.7546
Adj. R
2
0.7183 0.7216 0.7436 0.7460 0.7492
Notes: ***, ** and * indicate significance at the 0.01, 0.05 and 0.10 level, respectively. NW S.E. represents Newey West standard error.
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
7
However, there is no price difference for fish caught by other fishing
methods and bottom trawl (the base). Compared with bottom trawl,
Atlantic cod from longliners obtains a price premium of 7.73 %, and
Atlantic cod caught using Danish seine is 6.58 % cheaper than the base.
The lower prices for Danish seiners than bottom trawlers (and long-
liners) are probably a reflection of the lower quality.
3.3. Estimation results of the haddock model
For haddock, Model E in Table 3 indicates that the daily quantity
and the transaction quantity have a weak effect on the price of haddock
(-0.0029 and 0.0058), which is not surprising given that there is a
global market for cod (Asche et al., 2002). Fish Size is not statistically
significant for haddock, which may be due to the use of a dichotomous
dummy variable (above or below 0.8 kg)—a rather coarse representa-
tion of the fish size. However, the price premium for large line-caught
haddock (Quality-1) sold under the specific minimum price is 74.6 %
compared with downgraded haddock. The dummy for regular quality
(Quality-2) shows that buyers pay a 53.2 % premium for regular quality
compared with downgraded haddock, holding other variables constant.
This indicates that downgraded haddock is of substantially lower
quality than haddock of regular quality.
For haddock caught with the longline method, the price premium is
15.0 %, and the estimate for Danish seine fishing implies a discount of
3.0 % compared with the base. The estimate for other fishing methods
is insignificant.
3.4. Simulation analysis
For the fishing methods, the price premiums for Atlantic cod and
haddock are further compared. For longline, the price premium in
percentage terms for Atlantic cod is smaller than that for haddock (7.73
% versus 15.04 %). However, the price premiums in NOKs depend on
the respective price for the base product and may provide different
comparison results. Using the estimated coefficients, we obtained the
average fitted price for fish caught by trawl. Setting the dummy vari-
ables for the fishing method to 0 and taking the means of the numeric
variables in the models yielded the expected price for the base product,
which is about NOK 20.85 and 16.68 per kilogram for Atlantic cod and
haddock, respectively. We derived the expected price from the expected
price in the logarithm scale by taking the variance of residuals into
account, following Duan (1983). This yields price premiums for the
longline method of about NOK 1.61 and 2.51 per kilogram for Atlantic
cod and haddock, respectively. Since the estimated premiums for the
longline and Danish seine are relative to the same base, we can derive
the price premium for longline relative to Danish seine. For Atlantic
cod, longline obtains a 13.3 % premium over the Danish seine method.
For haddock, longline obtains a 18.6 % premium over Danish seine
fishing.
To illustrate the effect of the different fishing methods on the value
generated at the ex-vessel level of the value chain, we calculate the
potential gains in value if the Atlantic cod and haddock caught with
bottom trawl and Danish seine had been caught with longline, which
has the highest prices. If the quantity of fish caught using Danish seines
in the nine-year period covered by the study had been caught with
longline, the value of the fish would have increased by about NOK 50
million and NOK 31 million for Atlantic cod and haddock, respectively.
Similarly, if the quantity of fish caught with bottom trawling had been
caught with longlines, the value of the fish would have increased by
NOK 509 million for Atlantic cod and NOK 330 million for haddock.
Thus, the additional value gained if all the cod and haddock quotas had
been moved from bottom trawlers and Danish seiners to longliners is
about NOK 920 million. However, as illustrated in Fig. 1, the cost of
fishing is substantially higher for longlining than for bottom trawling
and Danish seining. A simple back-of-the-envelope calculation using the
additional costs of fishing with longline for the same extra quantity of
fish landed by trawlers and Danish seiners indicates an increase in costs
of about NOK 1637 million.
3.5. Robustness checks
We further performed the Box-Cox test by setting a grid from –2to
2. The optimized transformation parameters indicate the test results are
inconclusive. Accordingly, we estimated Models A, B, C and E for cod
and haddock. For the two sets of models, the F-test indicates that Model
E is the “best model”. The estimation results are reported in Appendixes
A and B.
For the estimated coefficients of the fishing methods (and the other
dummy variables), the logarithm functional form reports the percen-
tage price premium, while the linear functional form reports the price
premium in NOK. We further compared the estimated coefficients of
fishing methods from the two functional forms. As we discussed before,
the price premium in percentage terms is 7.73 % for Atlantic cod and
15.04 % for haddock, relative to their respective base, fish caught by
trawl. By taking the fitted price for the base, we obtained a price pre-
mium for the longline fishing method of about NOK 1.61 and 2.51 per
kilogram for Atlantic cod and haddock, respectively, which are close to
their counterparts in the model with the linear functional form (NOK
1.87 and 2.66 per kilogram).
Since Model E fits data better than the other models, we use the
Frisch-Waugh theorem to estimate this model and control for the whole
match effect. All variables except for the dummies for seller-buyer pairs
are centered with respect to their mean at the level of seller-buyer pairs.
Model E is estimated with the centered variables and without the
dummies for seller-buyer pairs. In our dataset, there are about 3.6 % of
pairs with only one transaction in the 9-year period for cod and 4.4 %
for haddock. We aggregate these seller-buyer pairs with only one
transaction as one pair to avoid zeros in the data. We then estimated the
level and logarithm functional forms. Tables 6 and 7 report the esti-
mation results for cod and haddock.
In Tables 6 and 7, besides the Newey-West standard error (NW SE),
we also reported the Driscoll and Kraay standard error (DK SE) (Driscoll
and Kraay, 1998) to control for correlation between day-to-day un-
observables. As shown in Tables 6 and 7, the two types of standard
errors generally do not affect the significance level for each variable.
For either cod or haddock, the logarithm functional form generates a
higher (adjusted) R
2
value than the linear functional form. Compared to
the original Model E, the new logarithm model has a higher R
2
value,
and the new linear model has a lower R
2
value for either cod or had-
dock. This may indicate that the logarithm functional form fits the data
better than the linear functional form. For the logarithm functional
form, the new model with the (almost) complete match effect increases
the R
2
value by 10.2 % for cod, and 12.2 % for haddock. The coefficient
of Longline in the new logarithm model is 0.0655 for cod, which is
smaller than the counterpart in Model E (0.0854); the inverse is true for
haddock (0.1639 versus 0.157), which may marginally affect the
Table 5
F-test results for the econometric models.
Model Against Degrees of freedom Statistic value p-value
Atlantic cod model:
Model B Model A 50 8.62 < 0.01
Model C Model A 50 32.4 < 0.01
Model D Model B 50 32.7 < 0.01
Model D Model C 50 9.01 < 0.01
Model E Model D 99 3.27 < 0.01
Haddock model:
Model B Model A 50 3.52 < 0.01
Model C Model A 50 21.7 < 0.01
Model D Model B 50 21.0 < 0.01
Model D Model C 50 2.95 < 0.01
Model E Model D 99 2.33 < 0.01
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
8
simulation results in the previous section.
4. Discussion
The empirical findings in this study clearly show that fishing
methods influence the price of frozen Atlantic cod and haddock after
controlling for the influence of other variables, such as the size of fish
and seasonality. The study also controls for the likely influence on price
of market imperfections caused by heterogeneity among sellers, buyers
and buyer-seller relationships (Gobillon et al., 2017). More specifically,
the results show that frozen Atlantic cod from longliners obtains price
premiums on average of 12.6 % and 15.0 % compared with that caught
by bottom trawlers and Danish seiners, respectively. For frozen had-
dock, longliners obtain an average price premium of 13.3 % over
bottom trawlers and 20.0 % over Danish seiners.
These results indicate that the quality and value of the limited
Atlantic cod and haddock resources can be enhanced by allocating
larger quotas to longliners at the expense of bottom trawlers and Danish
seiners. The better fish quality and environmental sustainability and
reputation of line-caught Atlantic cod and haddock compared with
other fishing methods also provide good opportunities for enhanced
value-adding in secondary processing and marketing. This should lead
to improved value creation from the limited Atlantic cod and haddock
stocks, much in line with the FAO’s request to take better care of the
scarce marine resources (Food and Agriculture Organization (FAO,
2018).
However, the costs of fishing are substantially higher for longliners
than for bottom trawlers and Danish seiners, which largely explains the
differences in profitability favoring the more technically efficient
bottom trawlers and Danish seiners. This also explains why longliners
through political pressure are now allowed to use Danish seines for
groundfish, including cod and haddock. The lower costs of fishing with
Danish seines are likely to improve the profitability of longliners
adopting this fishing method, but the quality of fish is likely to be re-
duced, with implications for value-adding and marketing.
Thus, policymakers face the following dilemma: should more quotas
be allocated to longliners at the expense of bottom trawlers and Danish
seiners to gain higher total value from the limited Atlantic cod and
haddock resources, or should bottom trawlers and Danish seiners, the
most technically efficient and profitable fishing methods, be favored at
the expense of the less efficient and profitable longliners? And should
longliners be allowed to fish with Danish seines?
In trying to solve these dilemmas, policymakers must consider
several complicating issues. First, the current fleet of longliners utilizes
its fishing capacity to the maximum. Thus, allocating additional quotas
to this group will imply new investments in vessels to enhance the
Table 6
Econometric results from estimating the hedonic price models for cod, with complete match effect.
Variable Dependent variable: log(Price) Dependent variable: Price
Est. NW S.E DK S.E. Est. NW S.E DK S.E.
log (Daily-Quantity) −0.0038 0.0008 *** 0.001 *** −0.0879 0.0145 *** 0.0145 ***
log (Transaction-Quantity) 0.0019 0.0005 *** 0.0006 *** 0.0246 0.009 *** 0.009 ***
Fish-Size 0.0208 0.0005 *** 0.0007 *** 0.4793 0.0104 *** 0.0104 ***
Quality-2 0.3599 0.0087 *** 0.0143 *** 5.8867 0.1036 *** 0.1036 ***
Longline 0.0655 0.0071 *** 0.0113 *** 1.8577 0.1209 *** 0.1209 ***
Danish-Seine −0.0716 0.0065 *** 0.0101 *** −1.1254 0.1219 *** 0.1219 ***
Other fishing methods 0.0005 0.0111 0.0178 0.3872 0.2091 0.2091 *
Auction 0.0395 0.0027 *** 0.0036 *** 0.6532 0.046 *** 0.046 ***
With time dummies Yes Yes
With seller dummies Yes Yes
With buyer dummies Yes Yes
Complete match effect Yes Yes
R
2
0.8069 0.6882
Adj. R
2
0.8052 0.6867
Notes: ***, ** and * indicate significance at the 0.01, 0.05 and 0.10 level, respectively. NW S.E. represents Newey West standard error, and DK S.E. Driscoll and Kraay
standard error.
Table 7
Econometric results from estimating the hedonic price models for haddock, with complete match effect.
Variable Dependent variable: log(Price) Dependent variable: Price
Est. NW S.E. DK S.E. Est. NW S.E DK S.E.
log (Daily-Quantity) −0.0036 0.0015 ** 0.0017 ** −0.0766 0.0195 *** 0.0222 ***
log (Transaction-Quantity) 0.0051 0.0011 *** 0.0012 *** 0.0519 0.0154 *** 0.0161 ***
Fish-Size 0.0086 0.004 ** 0.004 ** −0.016 0.0603 0.0588
Quality-1 0.5495 0.0209 *** 0.0265 *** 7.7769 0.2277 *** 0.2754 ***
Quality-2 0.4228 0.0193 *** 0.0253 *** 5.1965 0.194 *** 0.2593 ***
Longline 0.1639 0.0202 *** 0.0285 *** 3.0584 0.2412 *** 0.3216 ***
Danish-Seine −0.002 0.0174 0.0223 0.0449 0.2341 0.3107
Other fishing methods −0.0016 0.0331 0.039 0.9226 0.4636 ** 0.5562 *
Auction 0.0408 0.0057 *** 0.0072 *** 0.3873 0.0783 *** 0.105 ***
With time dummies Yes Yes
With seller dummies Yes Yes
With buyer dummies Yes Yes
Complete match effect Yes Yes
R
2
0.807 0.694
Adj. R
2
0.805 0.690
Notes: ***, ** and * indicate significance at the 0.01, 0.05 and 0.10 level, respectively. NW S.E. represents Newey West standard error, and DK S.E. Driscoll and Kraay
standard error.
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
9
fishing capacity—which opposes the longstanding policy of reducing
the fishing capacity. At the same time as new longliners will have to
carry substantial financial costs, influencing their profitability nega-
tively, the technical efficiency and revenue of bottom trawlers and
Danish seiners will be reduced. However, longliners might benefitif
they were allocated larger quotas of the valuable Atlantic cod at the
expense of the less valuable haddock, as this may lead to a more
profitable mix of quotas for this vessel group. This implies that bottom
trawlers and Danish seiners would receive smaller quotas for Atlantic
cod and larger quotas for haddock, which would reduce their revenues
at the same time as the costs of fishing would probably be similar, thus
implying reduced profits.
Reducing fuel consumption and greenhouse gas emissions is also an
important issue. As shown in Table 1, the cost of fuel varies sub-
stantially among the vessel groups. The new scheme for fuel surcharges
introduced in 2020 is likely to increase the cost of fuel, particularly for
vessels with high fuel consumption, such as bottom trawlers. This may
reduce the gap in profitability between the bottom trawlers and long-
liners.
The recent liberalization of fishing methods, i.e., letting longliners
use Danish seine to catch their quotas of Atlantic cod and other
groundfish, must also be considered given that the additional quotas for
Atlantic cod may be caught with the highly effective Danish seine, al-
beit with lower quality and sold at lower prices, as indicated above. It
should also be noted that the rather rigid fishery management system in
Norway, with vessel quotas owned by fishers, would make it very dif-
ficult for regulators to move quotas among vessel groups, as this may
lead to lawsuits with claims for compensation for lost quotas.
Policymakers should also consider recent research and development
into quality-enhancing technology for both Danish seiners and bottom
trawlers, such as live on-board storage and slaughtering (Olsen et al.,
2013;Digre et al., 2017;Sønvisen and Standal, 2019). Incentives, such
as a quota bonuses, are one way to enhance the adoption of such new
technology, which may lead to enhanced quality and higher prices for
bottom trawlers and Danish seiners alike. In order to improve fish
quality, the Norwegian Food Safety Authority should implement and
enforce stricter legislation to ensure faster on-board slaughtering and
bleeding of fish. Finally, the environmental sustainability of and mar-
keting opportunities for the different fishing methods should also be
considered.
CRediT authorship contribution statement
Geir Sogn-Grundvåg: Conceptualization, Investigation, Writing -
original draft, Writing - review & editing, Project administration,
Funding acquisition. Dengjun Zhang: Conceptualization,
Methodology, Formal analysis, Investigation, Writing - review &
editing. Bent Dreyer: Conceptualization, Writing - review & editing,
Project administration, Funding acquisition.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgment
The authors acknowledge the funding provided by the Norwegian
Fishermen’s Sales Organization and CRISP (Centre for Research-Based
Innovation in Sustainable Fish Capture and Processing Technology).
The authors also thank the Norwegian Fishermen’s Sales Organization
for providing the data and Jan Olav Punsvik and Sara Izquierdo for
patiently answering our queries regarding the data and organization of
the ex-vessel sales of groundfish. Finally, the very useful comments
from two anonymous reviewers are acknowledged.
Appendix A. Econometric results from estimating the hedonic price models for Atlantic cod, dependent variable: Price
Model A Model B Model C Model D Model E
Variable Est. NW S.E. Est. NW S.E. Est. NW S.E. Est. NW S.E. Est. NW S.E.
Intercept 13.1909 *** 13.2922 *** 13.3244 *** 13.2462 *** 13.2038 ***
[0.2522] [0.2584] [0.2482] [0.253] [0.253]
log (Daily-Quantity) −0.0651 *** −0.0619 *** −0.0798 *** −0.0758 *** −0.0757 ***
[0.0152] [0.0151] [0.0147] [0.0146] [0.0146]
log (Transaction-Quantity) 0.0024 0.0033 0.0208 ** 0.0186 ** 0.0192 **
[0.0093] [0.0095] [0.0091] [0.0092] [0.0092]
Fish-Size 0.5832 *** 0.5818 *** 0.4738 *** 0.4710 *** 0.4711 ***
[0.0089] [0.0088] [0.0103] [0.0103] [0.0103]
Quality-2 6.0748 *** 6.0854 *** 5.9177 *** 5.9127 *** 5.9004 ***
[0.0922] [0.0915] [0.0987] [0.0986] [0.0986]
Longline 1.9564 *** 1.6379 *** 1.9398 *** 1.8775 *** 1.8709 ***
[0.0468] [0.0887] [0.0539] [0.0896] [0.0908]
Danish-Seine −1.1021 *** −1.3216 *** −1.1478 −1.1794 *** −1.1361 ***
[0.059] [0.0937] [0.062] [0.0946] [0.0966]
Other fishing methods −0.2070 ** 0.1353 0.0000 −0.6869 *** −0.0243 0.0366
[0.103] [0.1493] [0.1149] [0.1563] [0.1613]
Auction 0.4623 *** 0.5468 *** 0.3308 *** 0.4480 *** 0.4564 ***
[0.0369] [0.0381] [0.0403] [0.0413] [0.042]
With time dummies Yes Yes Yes Yes Yes
With seller dummies No Yes No Yes Yes
With buyer dummies No No Yes Yes Yes
With seller-buyer dummies No No No No Yes
R
2
0.7332 0.7374 0.7527 0.7568 0.7602
Adj. R
2
0.7330 0.7366 0.7520 0.7557 0.7582
Notes: ***, ** and * indicate significance at the 0.01, 0.05 and 0.10 level, respectively. NW S.E. represents Newey West standard error.
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
10
Appendix B. Econometric results from estimating the hedonic price models for haddock, dependent variable: Price
Model A Model B Model C Model D Model E
Variable Est. NW S.E. Est. NW S.E. Est. NW S.E. Est. NW S.E. Est. NW S.E.
Intercept 6.2271 *** 5.5668 *** 5.5522 *** 5.0342 *** 5.1874 ***
[0.3303] [0.354] [0.3367] [0.3727] [0.3726]
log (Daily-Quantity) −0.0735 *** −0.0727 −0.0709 *** −0.0679 *** −0.0688 ***
[0.0196] [0.0199] [0.0191] [0.0194] [0.0194]
log (Transaction-Quantity) 0.0927 *** 0.0904 *** 0.0637 *** 0.0596 *** 0.0621 ***
[0.0165] [0.0168] [0.0152] [0.0155] [0.0154]
Fish-Size −0.1578 ** −0.0992 * −0.1424 ** −0.0821 −0.0723
[0.062] [0.0614] [0.0611] [0.0606] [0.0601]
Quality-1 9.6969 *** 9.6094 *** 8.061 *** 8.0146 *** 7.9016 ***
[0.1966] [0.1967] [0.2196] [0.2223] [0.2198]
Quality-2 6.7945 *** 6.7977 *** 5.2489 *** 5.2939 *** 5.2528 ***
[0.1525] [0.1545] [0.186] [0.1913] [0.1874]
Longline 2.6369 *** 2.9710 *** 2.5938 *** 2.7467 *** 2.6585 ***
[0.0935] [0.1676] [0.1094] [0.1818] [0.1912]
Danish-Seine −0.9763 *** −0.5188 *** −0.8507 *** −0.4780 ** −0.3885 **
[0.1127] [0.1997] [0.1245] [0.1979] [0.2063]
Other fishing methods 0.9401 ** 1.5587 *** 1.0421 *** 1.5095 *** 1.4226 ***
[0.379] [0.3981] [0.3425] [0.3658] [0.3658]
Auction −0.0409 0.0437 0.2653 *** 0.3323 *** 0.3046 ***
[0.0664] [0.0683] [0.0702] [0.0722] [0.0743]
With time dummies Yes Yes Yes Yes Yes
With seller dummies No Yes No Yes Yes
With buyer dummies No No Yes Yes Yes
With seller-buyer dummies No No No No Yes
R
2
0.7453 0.7510 0.7634 0.7682 0.7746
Adj. R
2
0.7446 0.7492 0.7616 0.7654 0.7696
Notes: ***, ** and * indicate significance at the 0.01, 0.05 and 0.10 level, respectively. NW S.E. represents Newey West standard error.
References
Asche, F., Guillen, J., 2012. The importance of fishing method, gear and origin: the
Spanish hake market. Mar. Policy 36, 365–369. https://doi.org/10.1016/j.marpol.
2011.07.005.
Asche, F., Roll, K., 2018. Economic inefficiency in a revenue setting: the Norwegian
whitefish fishery. Appl. Econ. 50, 6112–6127. https://doi.org/10.1080/00036846.
2018.1489502.
Asche, F., Flaaten, O., Isaksen, J.R., Vassdal, T., 2002. Derived demand and relationships
between prices at different levels in the value chain: a note. J. Agric. Econ. 53,
101–107. https://doi.org/10.1111/j.1477-9552.2002.tb00008.x.
Asche, F., Chen, Y., Smith, M.D., 2015. Economic incentives to target species and fish size:
prices and fine-scale product attributes in Norwegian fisheries. ICES J. Mar. Sci. 72,
733–740. https://doi.org/10.1093/icesjms/fsu208.
Asche, F., Cojocaru, A.L., Pincinato, R.B.M., Roll, K.H., 2020. Production risk in the
Norwegian fisheries. Environ. Resour. Econ. Rev. 75, 137–149. https://doi.org/10.
1007/s10640-019-00391-2.
Bertheussen, B.A., Dreyer, B., 2019. Is the Norwegian cod industry locked into a value-
destructive volume logic? Mar. Policy 103, 113–120. https://doi.org/10.1016/j.
marpol.2019.02.023.
Blomquist, J., Bartolino, V., Waldo, S., 2015. Price premiums for providing eco-labelled
seafood: evidence from MSC-certified cod in Sweden. J. Agric. Econ. 66, 690–704.
https://doi.org/10.1111/1477-9552.12106.
Botta, J.R., Squires, B.E., Johnson, J., 1986. Effect of bleeding/gutting procedures on the
sensory quality of fresh raw Atlantic cod (Gadus morhua). Can. Inst. Food Sci.
Technol. J. 19, 186–190. https://doi.org/10.1016/S0315-5463(86)71629-5.
Brinkhof, J., Olsen, S.H., Ingólfur, O.A., Larsen, R.B., 2018. Sequential codend improves
quality of trawl-caught cod. PLoS One 13 (10), e0204328. https://doi.org/10.1371/
journal.pone.0204328.
Carroll, M., Anderson, J.L., Martínez-Garmendia, J., 2001. Pricing of U.S. North Atlantic
bluefin tuna and implications for management. Agribus 17, 243–254. https://doi.
org/10.1002/agr.1014.
Digre, H., Rosten, C., Erikson, U., Mathiassen, J.R., Aursand, I.G., 2017. The on-board live
storage of Atlantic cod (Gadus morhua) and haddock (Melanogrammus aeglefinus)
caught by trawl: fish behaviour, stress and fillet quality. Fish. Res. 189, 42–54.
https://doi.org/10.1016/j.fishres.2017.01.004.
Directorate of Fisheries, 2020. Profitability Survey of the Norwegian Fishing Fleet (in
Norwegian). (Accessed 20 March 2020). https://www.fiskeridir.no/Yrkesfiske/Tall-
og-analyse/Loennsomhet.
Driscoll, J.C., Kraay, A.C., 1998. Consistent covariance matrix estimation with spatially
dependent panel data. Rev. Econ. Stat. 80, 549–560. https://doi.org/10.1162/
003465398557825.
Duan, N., 1983. Smearing estimate: a nonparametric retransformation method. J. Am.
Stat. 78, 605–610. https://doi.org/10.1080/01621459.1983.10478017.
Flåten, O., Heen, K., 2004. Fishing vessel profitability and local economic link ob-
ligations–the case of Norwegian trawlers. Mar. Policy 28, 451–457. https://doi.org/
10.1016/j.marpol.2003.10.019.
Fluvià, M., Garriga, A., Rigall-I-Torrent, R., Rodríguez-Carámbula, E., Saló, A., 2012.
Buyer and seller in fish markets organized as Dutch auctions: evidence from a
wholesale fish market in Southern Europe. Fish. Res. 127, 18–25. https://doi.org/10.
1016/j.fishres.2012.04.010.
Food and Agriculture Organization (FAO), 2018. The State of World Fisheries and
Aquaculture 2018. Meeting the Sustainable Development Goals. SOFIA report. Food
and Agriculture Organization of the United Nations, Rome (Accessed 20 March
2020). http://www.fao.org/3/I9540EN/i9540en.pdf.
Gobillon, L., Wolff, F.C., Guillotreau, P., 2017. The effect of buyers and sellers on fish
market prices. Eur. Rev. Agric. Econ. 44, 149–176. https://doi.org/10.1093/erae/
jbw006.
Guillioni, G., Bucciarelli, E., 2011. Agents’ability to manage information in centralized
markets: comparing two wholesale fish markets. J. Econ. Behav. Org. 80, 20–33.
https://doi.org/10.1016/j.jebo.2011.01.012.
Guttormsen, A.G., Roll, K., 2011. Technical efficiency in a heterogeneous fishery: the case
of Norwegian groundfish fisheries. Mar. Resource Econ. 26, 293–307. https://doi.
org/10.5950/0738-1360-26.4.293.
Hammarlund, C., 2015. The big, the bad, and the average: hedonic prices and inverse
demand for Baltic cod. Mar. Resource Econ. 30, 157–177. https://doi.org/10.1086/
679972.
Hannesson, R., 2013. Norway’s experience with ITQs. Mar. Policy 37, 264–269. https://
doi.org/10.1016/j.marpol.2012.05.008.
Helstad, K., Vassdal, T., Trondsen, T., Young, A.A., 2015. Price links between auction and
direct sales of fresh and frozen fish in North Norway (1997–2003). Mar. Resource
Econ. 20, 305–322. https://doi.org/10.1086/mre.20.3.42629477.
Holm, P., 1995. The dynamics of institutionalization: transformation processes in
Norwegian Fisheries. Adm. Sci. Q. 40, 398–422. https://www.jstor.org/stable/
2393791.
Homans, F.R., Wilen, J.E., 2005. Markets and rent dissipation in regulated open access
fisheries. J. Environ. Econ. Manag. 49, 381–404. https://doi.org/10.1016/j.jeem.
2003.12.008.
Isaksen, J.R., Hermansen, Ø., Flaaten, O., 2015. Stubborn fuel tax concessions: the case of
fisheries in Norway. Mar. Policy 52, 85–92. https://doi.org/10.1016/j.marpol.2014.
10.028.
Jafarzadeh, S., Ellingsen, H., Annondsen, S.A., 2016. Energy efficiency of Norwegian
fisheries from 2003–2012. J. Clean. Prod. 112, 3616–3630. https://doi.org/10.1016/
j.jclepro.2015.06.114.
Kennedy, P., 1981. Estimation with correctly interpreted dummy variables in semiloga-
rithmic equations. Am. Econ. Rev. 71, 801. https://www.jstor.org/stable/1806207.
Kirman, A.P., Vriend, N.J., 2001. Evolving market structure: an ACE model of price
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
11
dispersion and loyalty. J. Econ. Dyn. Control 25, 459–502. https://doi.org/10.1016/
S0165-1889(00)00033-6.
Kristofersson, D., Rickertsen, K., 2004. Efficient estimation of hedonic inverse input de-
mand systems. Am. J. Agric. Econ. 86, 1127–1137. https://doi.org/10.1111/j.0002-
9092.2004.00658.x.
Kristofersson, D., Rickertsen, K., 2007. Hedonic price models for dynamic markets. Oxf.
Bull. Econ. Stat. 69, 387–412. https://doi.org/10.1111/j.1468-0084.2006.00441.x.
Lancaster, K.J., 1966. A new approach to consumer theory. J. Political Econ. 74 (2),
132–157. https://doi.org/10.1086/259131.
Larkin, S., Sylvia, G., 1999. Intrinsic fish characteristics and intraseasonal production
efficiency: a management-level bioeconomic analysis of a commercial fishery. Am. J.
Agric. Econ. 81, 29–43. https://doi.org/10.2307/1244448.
Lee, M.-Y., 2014. Hedonic pricing of Atlantic cod: effects of size, freshness, and gear. Mar.
Resource Econ. 29, 259–277. https://doi.org/10.1086/677769.
Margeirsson, S., Jonsson, G.R., Arason, S., Thorkelsson, G., 2007. Influencing factors on
yield, gaping, bruises and nematodes in cod (Gadus morhua) fillets. J. Food Eng. 80,
503–508. https://doi.org/10.1016/j.jfoodeng.2006.05.032.
McConnell, K.E., Price, M., 2006. The lay system in commercial fisheries: origin and
implications. J. Environ. Econ. Manag. 51, 295–307. https://doi.org/10.1016/j.jeem.
2005.09.004.
McConnell, K.E., Strand, I.E., 2000. Hedonic prices for fish: tuna prices in Hawaii. Am. J.
Agric. Econ. 82, 133–144. https://doi.org/10.1111/0002-9092.00011.
Olsen, S.H., Tobiassen, T., Akse, L., Evensen, T.H., Midling, K.Ø., 2013. Capture induced
stress and live storage of Atlantic cod (Gadus morhua) caught by trawl: consequences
for the flesh quality. Fish. Res. 147, 446–453. https://doi.org/10.1016/j.fishres.
2013.03.009.
Olsen, S.H., Joensen, S., Tobiassen, T., Heia, K., Akse, L., Nilsen, H., 2014. Quality con-
sequences of bleeding fish after capture. Fish. Res. 153, 103–107. https://doi.org/10.
1016/j.fishres.2014.01.011.
Pettersen, I.K., Hestvik Brækkan, E., Myrland, Ø., 2018. Are Norwegian fishermen selling
in the same market? J. Commod. Mark 12, 9–18. https://doi.org/10.1016/j.jcomm.
2017.12.008.
Puig, P., Canals, M., Company, J.B., Martin, J., Amblas, D., Lastras, G., Palanques, A.,
Calafat, A.M., 2012. Ploughing the deep sea floor. Nature 489, 286–290. https://doi.
org/10.1038/nature11410.
Roheim, C.A., Asche, F., Insignares, J., 2011. The elusive price premium for ecolabelled
products: evidence from seafood in the UK market. J. Agric. Econ. 62, 655–668.
https://doi.org/10.1111/j.1477-9552.2011.00299.
Rosen, S., 1974. Hedonic prices and implicit markets: product differentiation in pure
competition. J. Political Econ. 82, 34–55. https://doi.org/10.1086/260169.
Rotabakk, B.T., Skipnes, D., Akse, L., Birkeland, S., 2011. Quality assessment of Atlantic
cod (Gadus morhua) caught by longlining and trawling at the same time and location.
Fish. Res. 112, 44–51. https://doi.org/10.1016/j.fishres.2011.08.009.
Salas, S., Gaertner, D., 2004. The behavioural dynamics of fishers: management im-
plications. Fish Fish. 5, 153–167. https://doi.org/10.1111/j.1467-2979.2004.
00146.x.
Schau, E.M., Ellingsen, H., Endal, A., Aanondsen, S.A., 2009. Energy consumption in the
Norwegian fisheries. J. Clean. Prod. 17, 325–334. https://doi.org/10.1016/j.jclepro.
2008.08.015.
Sønvisen, S.A., Standal, D., 2019. Catch-based aquaculture in Norway-Institutional
challenges in the development of a new marine industry. Mar. Policy 104, 118–124.
https://doi.org/10.1016/j.marpol.2019.02.039.
Sogn-Grundvåg, G., Larsen, T.A., Young, J.A., 2013. The value of line-caught and other
attributes: an exploration of price premiums for chilled fish in UK supermarkets. Mar.
Policy 38, 41–44. https://doi.org/10.1016/j.marpol.2012.05.017.
Sogn-Grundvåg, G., Larsen, T.A., Young, J.A., 2014. Product differentiation with credence
attributes and private labels: the case of whitefish in UK supermarkets. J. Agric. Econ.
65, 368–382. https://doi.org/10.1111/1477-9552.12047.
Sogn-Grundvåg, G., Asche, F., Zhang, D., Young, J.A., 2019a. Eco-labels and product
longevity: the case of whitefish in UK grocery retailing. Food Pol. 88, 1–10. https://
doi.org/10.1016/j.foodpol.2019.101750.
Sogn-Grundvåg, G., Zhang, D., Iversen, A., 2019b. Large buyers at a fish auction: the case
of the Norwegian pelagic auction. Mar. Policy 104, 232–238. https://doi.org/10.
1016/j.marpol.2018.06.011.
Standal, D., Aarset, B., 2008. The IVQ regime in Norway: a stable alternative to an ITQ
regime? Mar. Policy 32, 663–668. https://doi.org/10.1016/j.marpol.2007.11.007.
Van de Vis, H., Kesting, S., Robb, D., Oehlenschlager, J., Lambooij, B., Munkner, W.,
Kuhlmann, H., Kloosterboer, K., Tejada, M., Huidobro, A., Ottera, H., Roth, B.,
Sørensen, N.K., Akse, L., Besvadba, P., 2003. Is humane slaughter of fish possible for
industry? Aquac. Res. 34, 211–220. https://doi.org/10.1046/j.1365-2109.2003.
00804.x.
Vuong, Q.H., 1989. Likelihood ratio tests for model selection and non-nested hypotheses.
Econometrica. 57, 307–333. https://doi.org/10.2307/1912557.
Zhang, D., Sogn-Grundvåg, G., Asche, F., Young, J.A., 2018a. Eco-labeling and retail
pricing strategies: the U.K. Haddock market. Sustainability 10, 1522. https://doi.org/
10.3390/su10051522.
Zhang, D., Sikveland, M., Hermansen, Ø., 2018b. Fishing fleet capacity and profitability.
Mar. Policy 88, 116–121. https://doi.org/10.1016/j.marpol.2017.11.017.
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
12
