ArticlePDF Available

Abstract and Figures

This study explores trade-offs between fish quality, fishing efficiency, costs and profitability across three different 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 substantial 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.
Content may be subject to copyright.
Contents lists available at ScienceDirect
Fisheries Research
journal homepage: www.elsevier.com/locate/shres
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-os between sh quality, shing eciency, costs and protability across three dif-
ferent vessel groups in the Norwegian groundsh shery, that is, vessels shing 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 dierences in sh 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 shing are substantially higher for
longliners than for bottom trawlers and Danish seiners, which explains the dierences in protability favoring
the more technically ecient bottom trawlers and Danish seiners. Policy implications are discussed considering
trade-os between sh quality, ex-vessel prices and vessel protability. In a highly regulated shery such as the
Norwegian groundsh shery, with individual vessel and vessel group quotas based on historical shing rights,
policy intervention is important for optimal use of limited sh stocks but is not necessarily straightforward.
1. Introduction
It is well known that dierent shing methods inuence the quality
of landed sh dierently. 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 llets due to poor draining of blood
(Botta et al., 1986;Margeirsson et al., 2007;Olsen et al., 2014). When
sh is caught in large hauls, which is often the case with bottom
trawling and Danish seine, the sh often dies before bleeding (Olsen
et al., 2014). In addition, it is not unusual for the sh to be kept in
storage bins for hours before bleeding (Olsen et al., 2014). This sim-
plies the slaughtering process (Van de Vis et al., 2003), but the sh
must be bled while it is still alive to facilitate good drainage of blood.
Dierences in sh quality caused by dierent shing gear may lead
to dierent 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 bluen tuna in Hawaii (McConnell and
Strand, 2000) and Japan (Carroll et al., 2001). Dierent shing
methods may also impact the environment in dierent ways, which
may inuence the market value of the sh. Sogn-Grundvåg et al. (2013,
2014) estimated price premiums in the UK grocery retail market in the
10.424.6 % range for line-caught chilled and frozen Atlantic cod and
haddock compared with sh caught with other shing 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 sh may relate to better product quality (Rotabakk et al., 2011)
and also to the well-documented negative environmental eects of
other shing methods, including the detrimental eect 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 inuenced the demand for line-caught sh (Sogn-
Grundvåg et al., 2013;Zhang et al., 2018a).
Vessels using dierent shing methods also dier in technical e-
ciency, inuencing the costs of shing, which may vary between vessels
and vessel groups (Guttormsen and Roll, 2011;Asche and Roll, 2018).
And, importantly, the cost of shing and sher behavior may be in-
uenced by the way sheries are managed, for instance, by restrictions
related to vessel size and gear types to limit shing eort and by sub-
sidizing input factors such as capital and fuel to support shers
https://doi.org/10.1016/j.shres.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@noma.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-obetween sh quality, prices
gained and the costs of shing caused by gear restrictions and quota
allocations between vessel groups shing with dierent gear. For ex-
ample, the most ecient shing methods and vessels may provide sh
of low quality but may still be more protable than less technically
ecient shing methods providing sh of high quality and at better
prices. This may be further complicated by the fact that shing vessels
often hold quotas for several dierent species with sometimes over-
lapping seasons which may lead to ineciency because shers may
choose to sh too much of some species relative to other species, as
found by Asche and Roll (2018) for the Norwegian trawler eet. This
indicates that shery management excessively focused on input re-
strictions and protability in the harvesting sector may result in land-
ings of sh of reduced quality, which may aect 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 FAOs request to utilize sh stocks in
a way that contributes to the nutritional, social and economic value of
wild sh 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 protability of the same
vessel groups. Covering a period of 9 years (20092017), 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 sh is
headed and gutted and frozen at sea to preserve the quality and allow
longer trips. The sh 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 shery management can be highlighted and discussed.
2. Material and methods
2.1. Fishery management
The Norwegian groundsh shery is the most valuable shery in
Norway, and by far the most important species in this shery is Atlantic
cod, followed by haddock. The shing eet participating in the
groundsh shery is diverse, ranging from small coastal vessels shing
with jig machines, gillnets and hand-baited longlines, delivering fresh
catches to local sh 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 shery management system is based on several
policy goals, including preventing overshing and maintaining sus-
tainable coastal communities (Standal and Aarset, 2008). In addition,
resource rent and eet protability are key objectives of the manage-
ment system, as in many other countries (Flåten and Heen, 2004;
Hannesson, 2013;Zhang et al., 2018b).
The groundsh shery is regulated on a single species basis with a
total allowable catch (TAC) for the main groundsh species. The yearly
TAC for Atlantic cod and haddock is set based on scientic advice from
the International Council for the Exploration of the Seas (ICES).
Although scientic recommendations are given within a range, the
quotas tend to uctuate more or less in accordance with the variation in
biomass estimates, which uctuated substantially during the period
studied.
The management system divides the large and diverse shing eet
into a number of dierent 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 sh 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 shery and transferring the quota
to the acquirers vessel. Vessel quotas have been raised several times to
further stimulate consolidation, reduce the overall shing capacity and
enhance protability. This has been successful in the sense that over-
capacity has been reduced and the protability of the remaining vessels
has improved (Zhang et al., 2018b). The size of vessel quotas is re-
stricted and diers among vessel groups, with bottom trawlers holding
the largest quotas for cod and haddock.
In addition to vessel quotas, the shery is strictly regulated by a
mixture of measures, such as time and location closures and various
technical regulations relating to vessel size and shing gear, some of
which are described in more detail below. The Norwegian Food Safety
Authority enforces technical legislation regarding how to handle and
store sh on board the vessels in order to preserve the quality of the
sh.
2.2. The ex-vessel market
The ex-vessel sale of wild-caught sh 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 sh. This includes
the right to set minimum prices to secure the shers a price that reects
the market prices and to avoid powerful buyers using their bargaining
power to set prices that are too low for small independent shers
(Holm, 1995).
Several dierent sales organizations exist, covering dierent sh
species or geographical regions according to where the sh is landed.
The on-board frozen sh included in this study is sold through the
largest sales organization, the Norwegian Fishermens 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 Fishermens 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 groundsh, 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 sh or
salting or drying the sh 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 shing companies have acquired onshore pro-
cessing facilities. The vertically integrated companies have dierent
production and sales strategies for their vessels, with some buying their
own sh directly and others selling their sh to other buyers directly or
through the auction. Some of the sh 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 xed share of the rev-
enues rather than a xed wage (see, e.g., McConnell and Price,
2006)and the strong position of the Norwegian SeafarersUnion-
makes it dicult for vertically integrated companies to buy the sh
directly from their own vessels at low prices. If they do, skilled crew
members may leave for jobs on other vessels or industries. Autonomous
shing companies sell sh both directly to buyers and through the
auction. Some have developed long-term relationships with processing
companies to which sh is sold directly. These dierences in the re-
lationships between buyers and sellers may lead to market imperfec-
tions, thereby inuencing 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 clipsh industry, in which cod of a large
size is in demand. In the UK, frozen Atlantic cod and haddock are
mostly thawed and lleted by wholesalers and used for dierent 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 shing
with Danish seines. The bottom trawlers are licensed to sh 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 sh. 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 sh (Pettersen et al., 2018). The
trawlers are not allowed to sh within 12 nautical miles of the Nor-
wegian coastline. Some of the trawlers have quotas for prawns (Pan-
dalus borealis) in addition to groundsh. It should be noted that new
technology is being developed that may improve the quality of sh,
such as on-board live storage, so that the sh 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 sh for saithe, as saithe is not a typical
bottom feeder, thus making gillnets more eective 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
sh with trawls, but in 2018, they were allowed to use Danish seines to
sh their quotas for Atlantic cod, haddock and saithe. This liberal-
ization in gear use came about after pressure from the Norwegian
Fishermens 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 shing method. This illustrates how
shers 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 shing 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 sh with around 5070,000 hooks, which are baited
automatically when setting the line. The oceangoing longliners are not
allowed to sh within 4 nautical miles of the Norwegian coast, and their
access is restricted in certain areas outside the 4-mile limit to avoid
conicts of interest with smaller coastal vessels. Most of the oceangoing
longliners freeze their catch on board as headed and gutted sh.
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 eet 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 sh on board, and these are mainly the newest and
largest vessels. Several of the vessels in this group have licenses to sh
for pelagic sh, 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 exibility in terms of the choice of shing method. Compared
with trawlers and longliners, it also has the broadest spectrum of quotas
(species) and few restrictions in terms of shing areas.
The three vessel groups described above apply dierent shing
methods. However, the bottom trawl and Danish seine methods share
some key characteristics, such as a very large capacity to catch sh in a
single haul. It is not unusual to keep the sh in storage bins for hours
before bleeding and gutting (Olsen et al., 2014), and to simplify the
slaughtering process, bleeding is often done after suocation in air (van
de Vis et al., 2003), which implies reduced sh 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 sh at a
time. This makes it possible to bleed and process each sh immediately
after catching, which is the main reason for the higher quality provided
by longliners (Rotabakk et al., 2011). Longliners, however, are less
technically ecient than bottom trawlers and Danish seiners, implying
higher costs of shing. This is illustrated in Fig. 1, which shows the
average costs per kilogram of sh landed for the three vessel groups.
Fig. 1 shows that the average cost of shing is substantially higher
for longliners than for bottom trawlers and Danish seiners. The cost
dierences are relatively stable from 2010 to 2015 but increase after
that in disfavor of longlining. The costs of shing are about 30 % higher
for longliners than for bottom trawlers and Danish seiners. However,
because all vessels included sh for several sh species and because the
protability 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 sh for prawns, which generally is a less technically ecient form
of shing than trawling for cod and haddock. This indicates that the
cost of shing 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 ecient gear. This indicates that the costs of shing for cod
and haddock is somewhat higher for this vessel group than what is
shown in Fig. 1. The reduction in shing 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
shing when shing for cod and haddock. This implies that bottom
trawlers and Danish seiners can, to a larger extent than longliners, earn
prots through swift and intense shing strategies where lower quality
and prices are compensated for by large quantities caught at low cost.
Such shing strategies are relevant in a multi-species shery such as
this, with sometimes overlapping seasons and where the number of
operating days are very high, but the resultant sh quality may be
compromised (Bertheussen and Dreyer, 2019).
Fig. 1. The average costs per kilogram (NOK) of landed sh (all species) for the
three vessel groups for 20102018. 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 shing
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 diers 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), reecting
substantial dierences 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
dierences in the technical eciency 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 eciency of all vessel groups in Norwegian
sheries has improved in recent years (Jafarzadeh et al., 2016). For the
period 20032012, the average was 0.3220.354 and 0.265 L of fuel per
kilo of sh 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 shing with gillnets and longlines. For this group, the average
fuel consumption per kilogram of sh landed for the period 20032012
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 shing eet paid a
substantially lower CO
2
fuel surcharge compared to other commercial
vessels (e.g., vessels involved in sh farming), namely NOK0.29 per
liter of fuel for shing vessels versus NOK1.65 per liter for other vessels.
This subsidy represents about 2025 % 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, shing
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 reect 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 shing with dierent 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 shing methods for
Atlantic cod and haddock. Fish caught with longline fetch the highest
Table 1
a
Average yearly turnover (NOK), key costs (NOK), protability 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 protability survey of Norwegian sheries 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 prot rather than the more commonly used return on
investments to reect better how the cost of shing methods relates to earnings. Even though cod and haddock are the main species, bottom trawlers also catch other
groundsh species and some also have quotas for prawns, which may inuence their protability. Oceangoing longliners mainly sh for cod, haddock and saithe,
whereas several Danish seiners sh for pelagic species, such as mackerel (Scomber scombrus) and herring (Clupea harengus) in addition to groundsh.
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 gures also show that the price
dierences between shing methods were relatively stable over the
nine-year period. The econometric modelling will estimate the dier-
ences in prices between shing methods, holding other factors such as
the sh 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,
shing method, average size of the sh, 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 reecting seasonal and
yearly price dierences due to seasonal dierences 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 sh size is 2.70 kg (headed and
gutted weight), with a substantial standard deviation indicating a large
size variation between lots. For haddock, the sh 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 shing 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 specic minimum price. There is no specic minimum
price for Atlantic cod caught by longline. Quality-2 for Atlantic cod and
haddock refers to sh of regular quality as opposed to sh that was
downgraded due to poor quality. Under the Raw Fish Act, buyers may
downgrade sh 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 inuence of shing 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 shing methods for frozen
headed and gutted Atlantic cod.
Fig. 5. Yearly average prices per kilogram (NOK) by shing 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 sh, its quality and the shing
method. Fish prices may, however, also be inuenced 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 ts 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
deated 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 shing methods (Longline and
Danish Seine) and other shing methods together are dummies, with
bottom trawl as the base. The dummy variable for a particular shing
method may also capture the impact on price of some unobservable
factors which are related to this shing 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 inuence the prices
(Gobillon et al., 2017).
Regarding sh size and quality dummies, the measures are dierent
for cod and haddock. For the cod model, Fish Size is the average size of
the sh 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 sh with regular quality
with downgraded sh as the base. Additionally, Quality-1 is an addi-
tional quality dummy reecting the specic 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 sh prices are probably aected by market imperfections
caused by heterogeneity among buyers (e.g., dierent degrees of will-
ingness to pay for dierent sh quality and sh size, depending on the
markets and the customers whom they serve (Sogn-Grundvåg et al.,
2019b), sellers (e.g., dierences in skipper and crew skills, which may
result in dierences in sh quality) and buyer-seller relationships (e.g.,
some buyers and sellers are vertically integrated, which may inuence
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 signicant 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 ve 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 buyerseller pairs
Of the ve models, Model A is nested under Model B and Model C,
which are further nested under Model D. Model E is a complete speci-
cation. F-tests are used to test Models AD until rejection occurs.
Since the dependent variable is expressed in logarithm form, the
estimated coecients 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 signicant estimation d
2
, the impact
(price premium) of sh caught by longliners relative to the base (sh
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 eect
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 specication 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 eect 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 eect; 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 aect estimates of shing 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 eect.
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 t 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 t 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 rmly rejected. Thus, Model E ts 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
signicant 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 signicant 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 aected 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 sh size increased by 1 kg. Buyers
willingness to pay for Atlantic cod of regular quality is reected 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 signicant.
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 shing 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 signicance 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 shing 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 signicance 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 dierence for sh caught by other shing
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 reection 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 eect 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
signicant 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 sh size. However, the price premium for large line-caught
haddock (Quality-1) sold under the specic 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 shing implies a discount of
3.0 % compared with the base. The estimate for other shing methods
is insignicant.
3.4. Simulation analysis
For the shing 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 dierent
comparison results. Using the estimated coecients, we obtained the
average tted price for sh caught by trawl. Setting the dummy vari-
ables for the shing 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
shing.
To illustrate the eect of the dierent shing 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 sh caught using Danish seines
in the nine-year period covered by the study had been caught with
longline, the value of the sh would have increased by about NOK 50
million and NOK 31 million for Atlantic cod and haddock, respectively.
Similarly, if the quantity of sh caught with bottom trawling had been
caught with longlines, the value of the sh 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
shing is substantially higher for longlining than for bottom trawling
and Danish seining. A simple back-of-the-envelope calculation using the
additional costs of shing with longline for the same extra quantity of
sh 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 coecients of the shing 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 coecients of
shing 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, sh caught by
trawl. By taking the tted price for the base, we obtained a price pre-
mium for the longline shing 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 ts data better than the other models, we use the
Frisch-Waugh theorem to estimate this model and control for the whole
match eect. 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 aect the signicance 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 ts the data
better than the linear functional form. For the logarithm functional
form, the new model with the (almost) complete match eect increases
the R
2
value by 10.2 % for cod, and 12.2 % for haddock. The coecient
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 aect 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 ndings in this study clearly show that shing
methods inuence the price of frozen Atlantic cod and haddock after
controlling for the inuence of other variables, such as the size of sh
and seasonality. The study also controls for the likely inuence on price
of market imperfections caused by heterogeneity among sellers, buyers
and buyer-seller relationships (Gobillon et al., 2017). More specically,
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 sh quality and environmental sustainability and
reputation of line-caught Atlantic cod and haddock compared with
other shing 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 FAOs request to take better care of the
scarce marine resources (Food and Agriculture Organization (FAO,
2018).
However, the costs of shing are substantially higher for longliners
than for bottom trawlers and Danish seiners, which largely explains the
dierences in protability favoring the more technically ecient
bottom trawlers and Danish seiners. This also explains why longliners
through political pressure are now allowed to use Danish seines for
groundsh, including cod and haddock. The lower costs of shing with
Danish seines are likely to improve the protability of longliners
adopting this shing method, but the quality of sh 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 ecient and protable shing methods, be favored at
the expense of the less ecient and protable longliners? And should
longliners be allowed to sh with Danish seines?
In trying to solve these dilemmas, policymakers must consider
several complicating issues. First, the current eet of longliners utilizes
its shing 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 eect.
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 shing 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 eect Yes Yes
R
2
0.8069 0.6882
Adj. R
2
0.8052 0.6867
Notes: ***, ** and * indicate signicance 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 eect.
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 shing 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 eect Yes Yes
R
2
0.807 0.694
Adj. R
2
0.805 0.690
Notes: ***, ** and * indicate signicance 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
shing capacitywhich opposes the longstanding policy of reducing
the shing capacity. At the same time as new longliners will have to
carry substantial nancial costs, inuencing their protability nega-
tively, the technical eciency and revenue of bottom trawlers and
Danish seiners will be reduced. However, longliners might benetif
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
protable 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 shing would probably be similar, thus
implying reduced prots.
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 protability between the bottom trawlers and long-
liners.
The recent liberalization of shing methods, i.e., letting longliners
use Danish seine to catch their quotas of Atlantic cod and other
groundsh, must also be considered given that the additional quotas for
Atlantic cod may be caught with the highly eective Danish seine, al-
beit with lower quality and sold at lower prices, as indicated above. It
should also be noted that the rather rigid shery management system in
Norway, with vessel quotas owned by shers, would make it very dif-
cult 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 sh
quality, the Norwegian Food Safety Authority should implement and
enforce stricter legislation to ensure faster on-board slaughtering and
bleeding of sh. Finally, the environmental sustainability of and mar-
keting opportunities for the dierent shing 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 nancial
interests or personal relationships that could have appeared to inu-
ence the work reported in this paper.
Acknowledgment
The authors acknowledge the funding provided by the Norwegian
Fishermens Sales Organization and CRISP (Centre for Research-Based
Innovation in Sustainable Fish Capture and Processing Technology).
The authors also thank the Norwegian Fishermens 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 groundsh. 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 shing 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 signicance 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 shing 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 signicance 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 shing method, gear and origin: the
Spanish hake market. Mar. Policy 36, 365369. https://doi.org/10.1016/j.marpol.
2011.07.005.
Asche, F., Roll, K., 2018. Economic ineciency in a revenue setting: the Norwegian
whitesh shery. Appl. Econ. 50, 61126127. 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 dierent levels in the value chain: a note. J. Agric. Econ. 53,
101107. 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 sh size:
prices and ne-scale product attributes in Norwegian sheries. ICES J. Mar. Sci. 72,
733740. 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 sheries. Environ. Resour. Econ. Rev. 75, 137149. 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, 113120. 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-certied cod in Sweden. J. Agric. Econ. 66, 690704.
https://doi.org/10.1111/1477-9552.12106.
Botta, J.R., Squires, B.E., Johnson, J., 1986. Eect of bleeding/gutting procedures on the
sensory quality of fresh raw Atlantic cod (Gadus morhua). Can. Inst. Food Sci.
Technol. J. 19, 186190. 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
bluen tuna and implications for management. Agribus 17, 243254. 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 aeglenus)
caught by trawl: sh behaviour, stress and llet quality. Fish. Res. 189, 4254.
https://doi.org/10.1016/j.shres.2017.01.004.
Directorate of Fisheries, 2020. Protability Survey of the Norwegian Fishing Fleet (in
Norwegian). (Accessed 20 March 2020). https://www.skeridir.no/Yrkesske/Tall-
og-analyse/Loennsomhet.
Driscoll, J.C., Kraay, A.C., 1998. Consistent covariance matrix estimation with spatially
dependent panel data. Rev. Econ. Stat. 80, 549560. https://doi.org/10.1162/
003465398557825.
Duan, N., 1983. Smearing estimate: a nonparametric retransformation method. J. Am.
Stat. 78, 605610. https://doi.org/10.1080/01621459.1983.10478017.
Flåten, O., Heen, K., 2004. Fishing vessel protability and local economic link ob-
ligationsthe case of Norwegian trawlers. Mar. Policy 28, 451457. 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 sh markets organized as Dutch auctions: evidence from a
wholesale sh market in Southern Europe. Fish. Res. 127, 1825. https://doi.org/10.
1016/j.shres.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., Wol, F.C., Guillotreau, P., 2017. The eect of buyers and sellers on sh
market prices. Eur. Rev. Agric. Econ. 44, 149176. https://doi.org/10.1093/erae/
jbw006.
Guillioni, G., Bucciarelli, E., 2011. Agentsability to manage information in centralized
markets: comparing two wholesale sh markets. J. Econ. Behav. Org. 80, 2033.
https://doi.org/10.1016/j.jebo.2011.01.012.
Guttormsen, A.G., Roll, K., 2011. Technical eciency in a heterogeneous shery: the case
of Norwegian groundsh sheries. Mar. Resource Econ. 26, 293307. 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, 157177. https://doi.org/10.1086/
679972.
Hannesson, R., 2013. Norways experience with ITQs. Mar. Policy 37, 264269. 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 sh in North Norway (19972003). Mar. Resource
Econ. 20, 305322. 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, 398422. https://www.jstor.org/stable/
2393791.
Homans, F.R., Wilen, J.E., 2005. Markets and rent dissipation in regulated open access
sheries. J. Environ. Econ. Manag. 49, 381404. 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
sheries in Norway. Mar. Policy 52, 8592. https://doi.org/10.1016/j.marpol.2014.
10.028.
Jafarzadeh, S., Ellingsen, H., Annondsen, S.A., 2016. Energy eciency of Norwegian
sheries from 20032012. J. Clean. Prod. 112, 36163630. 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, 459502. https://doi.org/10.1016/
S0165-1889(00)00033-6.
Kristofersson, D., Rickertsen, K., 2004. Ecient estimation of hedonic inverse input de-
mand systems. Am. J. Agric. Econ. 86, 11271137. 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, 387412. 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),
132157. https://doi.org/10.1086/259131.
Larkin, S., Sylvia, G., 1999. Intrinsic sh characteristics and intraseasonal production
eciency: a management-level bioeconomic analysis of a commercial shery. Am. J.
Agric. Econ. 81, 2943. https://doi.org/10.2307/1244448.
Lee, M.-Y., 2014. Hedonic pricing of Atlantic cod: eects of size, freshness, and gear. Mar.
Resource Econ. 29, 259277. https://doi.org/10.1086/677769.
Margeirsson, S., Jonsson, G.R., Arason, S., Thorkelsson, G., 2007. Inuencing factors on
yield, gaping, bruises and nematodes in cod (Gadus morhua) llets. J. Food Eng. 80,
503508. https://doi.org/10.1016/j.jfoodeng.2006.05.032.
McConnell, K.E., Price, M., 2006. The lay system in commercial sheries: origin and
implications. J. Environ. Econ. Manag. 51, 295307. https://doi.org/10.1016/j.jeem.
2005.09.004.
McConnell, K.E., Strand, I.E., 2000. Hedonic prices for sh: tuna prices in Hawaii. Am. J.
Agric. Econ. 82, 133144. 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 esh quality. Fish. Res. 147, 446453. https://doi.org/10.1016/j.shres.
2013.03.009.
Olsen, S.H., Joensen, S., Tobiassen, T., Heia, K., Akse, L., Nilsen, H., 2014. Quality con-
sequences of bleeding sh after capture. Fish. Res. 153, 103107. https://doi.org/10.
1016/j.shres.2014.01.011.
Pettersen, I.K., Hestvik Brækkan, E., Myrland, Ø., 2018. Are Norwegian shermen selling
in the same market? J. Commod. Mark 12, 918. 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 oor. Nature 489, 286290. 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, 655668.
https://doi.org/10.1111/j.1477-9552.2011.00299.
Rosen, S., 1974. Hedonic prices and implicit markets: product dierentiation in pure
competition. J. Political Econ. 82, 3455. 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, 4451. https://doi.org/10.1016/j.shres.2011.08.009.
Salas, S., Gaertner, D., 2004. The behavioural dynamics of shers: management im-
plications. Fish Fish. 5, 153167. 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 sheries. J. Clean. Prod. 17, 325334. 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, 118124.
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 sh in UK supermarkets. Mar.
Policy 38, 4144. https://doi.org/10.1016/j.marpol.2012.05.017.
Sogn-Grundvåg, G., Larsen, T.A., Young, J.A., 2014. Product dierentiation with credence
attributes and private labels: the case of whitesh in UK supermarkets. J. Agric. Econ.
65, 368382. 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 whitesh in UK grocery retailing. Food Pol. 88, 110. https://
doi.org/10.1016/j.foodpol.2019.101750.
Sogn-Grundvåg, G., Zhang, D., Iversen, A., 2019b. Large buyers at a sh auction: the case
of the Norwegian pelagic auction. Mar. Policy 104, 232238. 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, 663668. 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 sh possible for
industry? Aquac. Res. 34, 211220. 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, 307333. 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 eet capacity and protability.
Mar. Policy 88, 116121. https://doi.org/10.1016/j.marpol.2017.11.017.
G. Sogn-Grundvåg, et al. Fisheries Research 230 (2020) 105672
12
... These vessel groups apply different fishing gear, but trawlers and seiners have in common a very large capacity to catch fish in a single haul. This reduces fishing costs but may compromise fish quality (Bertheussen and Dreyer, 2019;Rotabakk et al., 2011;Sogn-Grundvåg et al., 2020). Longliners, on the other hand, take onboard the fish one by one. ...
... Longliners, on the other hand, take onboard the fish one by one. This makes it possible to bleed and process each fish immediately after catching, which is the main reason for the higher fish quality generally provided by longliners (Rotabakk et al., 2011;Sogn-Grundvåg et al., 2020), leading to higher market prices both at the ex-vessel (Sogn-Grundvåg et al., 2020) and retail (Sogn-Grundvåg et al., 2013 levels of the value chain. ...
... Longliners, on the other hand, take onboard the fish one by one. This makes it possible to bleed and process each fish immediately after catching, which is the main reason for the higher fish quality generally provided by longliners (Rotabakk et al., 2011;Sogn-Grundvåg et al., 2020), leading to higher market prices both at the ex-vessel (Sogn-Grundvåg et al., 2020) and retail (Sogn-Grundvåg et al., 2013 levels of the value chain. ...
Article
Full-text available
This study addresses how commodities that go unsold at auction perform when they are subsequently sold directly or at auction. Given that the share of unsold commodities is up to 50% in some markets, this is an important yet neglected topic. With a unique dataset including approximately 40,000 frozen Atlantic cod transactions, with detailed information about each lot, including if it was previously withdrawn from the auction, we apply hedonic price modeling to investigate how these lots are priced. The results indicate that when previously withdrawn lots are sold directly, they achieve lower prices than comparable lots. In the auction market, on the other hand, previously unsold lots achieve slightly higher prices than comparable lots.
... The frozen cod is mainly landed by large oceangoing trawlers, longliners, and Danish seiners (Sogn-Grundvåg et al., 2020). These vessel groups apply different fishing gear, but trawlers and Danish seiners have in common a very large capacity to catch fish in a single haul. ...
... These vessel groups apply different fishing gear, but trawlers and Danish seiners have in common a very large capacity to catch fish in a single haul. This reduces fishing costs but may compromise fish quality (Bertheussen and Dreyer, 2019;Rotabakk et al., 2011;Sogn-Grundvåg et al., 2020). Longliners, on the other hand, catch only one fish at a time, making it possible to bleed and process each fish immediately after catching. ...
... Longliners, on the other hand, catch only one fish at a time, making it possible to bleed and process each fish immediately after catching. This is the main reason for the higher fish quality generally provided by longliners (Rotabakk et al., 2011;Sogn-Grundvåg et al., 2020), leading to higher market prices (Sogn-Grundvåg et al., 2014, 2013. ...
Article
Full-text available
This study addresses how commodities that go unsold at auction perform when they are resold directly or in a new auction. Given that the share of unsold lots is up to 50% in some markets, this is an important yet neglected topic. We focus on the “burning-effect” hypothesis that buyers may consider other buyers’ opinions – signaled by their lack of willingness to pay for a commodity – while valuing a previously unsold commodity. This implies that commodities that go unsold in auctions are believed to be of lower quality than otherwise identical commodities. With a unique dataset including almost 40,000 frozen Atlantic cod transactions, with detailed information about each lot, including if it was previously unsold at auction, we apply hedonic price modeling to investigate how these lots are priced. The results indicate that when lots are resold directly, where 67% of the unsold lots are traded, they achieve lower prices than otherwise comparable lots, implying support for a burning effect. In the auction market, on the other hand, previously unsold lots achieve higher prices than otherwise comparable lots.
... The frozen fish may also be stored in onshore cold storage plants, if for instance the fisher anticipates higher future prices. However, storage costs are incurred and fish quality decline over time, influencing prices negatively [33]. It should also be noted that since 2015 each vessel group have been allowed to transfer up to 10 % of their quotas for cod and haddock from one year to the next. ...
... In contrast to grocery retailing, many of these outlets were also closed during the first wave of the pandemic. We compare prices before and during the pandemic for key quality attributes such as fish size, storage time, quality grade, and fishing methods, which typically are very important in price formation in fish markets [33,37,38]. Interestingly, the vessel groups differ substantially in the quality of fish landed, with longliners providing fish of higher quality and gaining better prices compared to the more technically efficient trawlers and Danish seiners, but with substantially lower profitability [33]. ...
... We compare prices before and during the pandemic for key quality attributes such as fish size, storage time, quality grade, and fishing methods, which typically are very important in price formation in fish markets [33,37,38]. Interestingly, the vessel groups differ substantially in the quality of fish landed, with longliners providing fish of higher quality and gaining better prices compared to the more technically efficient trawlers and Danish seiners, but with substantially lower profitability [33]. ...
Article
Full-text available
Small-scale fisheries have received most of the attention in the literature investigating negative impacts of the COVID-19 pandemic on seafood production. Larger fishing vessels are often perceived to be more resilient as they are better able to alter harvest patterns in response to supply shocks than smaller, less mobile vessels. In addition, larger fishing vessels often deliver storable frozen products contributing to resiliency. The supply and demand shocks caused by the COVID-19 pandemic provides an opportunity to test this hypothesis and is investigated here on the large-scale groundfish fleet in Norway. The results indicate that during the first two whole years of the pandemic the impact on price was small, but also that there were several secondary effects showing how negative shocks in some supply chains/markets are overcome.
... This may seem a strong assumption, given that several sellers and buyers in this market are vertically integrated or have developed long-term business relationships in direct sales, which, as indicated by Gobillon, Wolff and Guillotreau (2017), may improve information efficiency and influence prices in direct sales. In addition, Helstad and colleagues did not control for quality attributes such as fishing methods and downgrading, which are known to influence cod prices (Sogn-Grundvåg, Zhang and Dreyer, 2020;. ...
... For example, fishing method, which is an important observable quality signal that influences prices in capture-based fisheries (e.g. McConnell and Strand, 2000;Lee, 2014;Sogn-Grundvåg, Zhang and Dreyer, 2020;, may conceal quality variation between catches landed with the same fishing method. This variation may be caused by variations in the size of hauls, on-board processing facilities and routines among vessels fishing with the same gear (Rotabakk et al., 2011;Olsen et al., 2014). ...
... This also means that the direct sales market performs better than the auction in terms of quality-based pricing. This is relevant because quality-based pricing is important in incentivising fishers to provide high-quality fish to the market, contributing to the optimal use of limited marine resources (Sogn-Grundvåg, Zhang and Dreyer, 2020;. The informational efficiency of the auction market may, however, be improved by providing more fine-grained quality attributes that capture more of the unobserved quality. ...
Article
Full-text available
This study examines the question of selling agricultural commodities by auction or directly. Hedonic price analysis using transaction data from the sale of frozen cod in Norway shows that buyer–seller matches explain 32.4 and 13.6 per cent of the price variation in direct sales and auctions, respectively, indicating that direct sales are more informationally efficient than auctions. Meanwhile, auctions gain a price premium of 2.6 per cent over direct sales, holding other variables constant. However, a substantial increase in the use of direct sales indicates that their information efficiency is more important to sellers than the small price premium provided by auctions.
... With insufficient information about the quality of fish in individual catches, buyers may use fishing gear as a proxy for quality in their pricing. This is evident from several studies that found price differences between types of fishing gear for cod [26][27][28][29] and other species such as tuna [30,31]. At the same time, it is well known that the quality of fish may vary substantially between vessels fishing with the same gear, which may be due to variations in fishing tactics such as soaking time for gillnets and longlines [20,32,33] and haul sizes with Danish seines [34], as well as onboard handling practices and technology, including how fast the fish are bled [35,36]. ...
... We show that improved fish quality and landing timings are also possible without majority ownership but may depend on revenue-sharing arrangements, market organization, and fishery management. Second, our study also focuses on the important trade-off between fishing efficiency and fish quality [8,27]. We show that in the coastal fishery for cod in Norway, even when using the most efficient fishing gear often associated with low-quality fish in this fishery gillnets and Danish seinesfishers may land fish of high quality and achieve high prices. ...
Article
Full-text available
In the important coastal fishery for Atlantic cod in Norway quality-based pricing is largely absent. Thus, driven by a focus on fish quantity, not quality, a high share of downgraded fish is landed. This study identified some encouraging exceptions with actors trying to align incentives and increase revenues in both fishing and processing. More specifically, we examine two cases involving quality-enhancing fishing and find indications of substantial improvements in fish quality and importantly, that this is rewarded with higher prices. In a revenue-sharing arrangement with a focus on high quality fish delivered outside of the main season, fishers obtain approximately 50 % higher prices than comparable vessels fishing with the same gear in the same period. In the other case, fishers' "skip" the Norwegian ex-vessel market and sell their high-quality cod catches at Danish auctions at significantly higher prices than comparable vessels fishing in the same area and delivering in South Norway. The cases may guide fishers and fish buyers in increasing revenues through quality-enhancing fishing but also involve important dilemmas for policymakers which are highlighted and discussed.
... Market price of fishes are influenced by various attributes, such as the quality of fish, market situation, and consumer preferences (Mitra et al., 2021;Sogn-Grundvåg et al., 2021). Although these influences are studied well by the hedonic price analysis in tuna, cod, and haddock (e. g., McConnell and Strand, 2000;Lee, 2014;Sogn-Grundvåg et al., 2020), those of coral reef fishes in tropical waters are not well focused (Nassiri et al., 2021;Mitra et al., 2021). ...
... Despite the importance of individual fish quality, few studies have focused on the quality in wild captured fishes (e.g. Lee, 2014;Shimose et al., 2018;Sogn-Grundvåg et al., 2020). This study examined what factors affect and how much they influence the market value of P. leopardus with other congeners at the fish market in the Yaeyama region. ...
Article
Body color of the coralgroupers (Serranidae, Plectropomus) is known to vary from red to brown and dark to pale, with a vivid red color is thought to be preferred in fish markets of East Asia. This study revealed quantitative effects of the body color of coralgroupers on the market value at fish market in Okinawa, Japan. Results showed that bid price of six color grades of P. leopardus are significantly different, and normal- to dark-red individuals were sold at higher auction prices than light-brown and light-red individuals. Results also showed that unit price per kg of coralgroupers increased up to 1.5 kg body weight but slightly decreased for fish larger than 4 kg. Increase of fish supply in the market greatly reduced the bid price of the coralgroupers, and condition factor (index of fatness for individual fish) influenced the bid price of individual fish. These results indicate that an effective fisheries management strategy could be available if the optimal size and color individuals were harvested for days/weeks when a shortage of wild fish catch occurred.
Article
Full-text available
Fishermen make repeated choices with respect to when, where, and how to catch their target species. While these targeting tactics—and the factors shaping them—are known to fishers and some experts, knowledge about them is largely informal and not well utilized for management purposes. To formalize information on targeting tactics, we propose a set of methods combining model-based classification of target species with generalized linear models. We apply these methods to Norwegian coastal fishing vessels that caught Atlantic cod (Gadus morhua) as a part of their catch portfolio in 2019. The data contains nearly 32000 fishing trips by 761 vessels. Gaussian mixture models identify eight latent targeting tactics. Cod contributes significantly to three of the tactics. The Herfindahl–Hirschman Index, a measure of vessel-level diversity of tactics, shows that one quarter of the vessels had a specialized strategy (targeting cod plus at most one additional tactic). While cod is often studied as a single-species fishery, we show that cod-catching vessels can be engaged in relatively pure fisheries during some fishing trips but switch to different, often more mixed targets during other trips. We term this as “sequential mixed fisheries”. This is both a challenge and an opportunity for the fisheries management.
Article
The stone crab (Menippe mercenaria and Menippe adina) fishery is the second most valuable commercial fishery in the state of Florida, with $30 million in dockside annual landings in 2019, It is a unique fishery in that only the claws are harvested, and the crab is returned to the sea where a high portion of them survive. There is substantial variability in the ex-vessel prices paid to fishers for stone crab claws, with claws yielding from $5 to over $30 per pound. However, there is limited knowledge with respect to what influences the prices of this valuable seafood product. Hedonic models allowed us to estimate the effect that different product attributes of a harvested stone crab claw have on the price. Here, we used commercial landings data from 2009 to 2020 to investigate the impact of attributes including claw size, county landed, fishing year, month, and area fished (inshore or offshore). The premiums associated with these attributes are highly variable with the largest premiums based on claw size followed by county landed, with south Florida and the Keys having higher price premiums than northwest Florida. There was interannual variation for the month attribute, with the middle of the fishing year having the highest price (January, February) and the lowest was placed on the start and end of each fishing year (October, November, May). These results provide basic economic information and regional insights into one of the most valuable fisheries in Florida and the Gulf of Mexico, and can help inform managers and stakeholders on the balance between profitability and sustainability of this unique fishery.
Article
Full-text available
Longlining is an important fishing method targeting sharks, tuna, seer fish etc. Fishing line is tied to high carbon steel hook using stainless steel snood wire in order to avoid large carnivorous fishes from biting off the hook. Corrosion is a major problem with hooks and often the hook fails at the eye region where the line is attached to the hook. How the attachment of metallic snood wire affects the hook in terms of corrosion is communicated here. Number 7, Round bent (J shaped) hook was used in this study. Corrosion rate of snood wire-rigged hooks was evaluated by exposing them to accelerated corrosion testing condition (salt spray as per ASTM B117-03-2003). The snood wire rigged hooks recorded substantial weight loss on exposure to salt fog for 300 h compared to control hooks (without snood wire attachment). Correspondingly, snood wire-rigged hooks displayed a corrosion rate of 109.316 mpy (mills per year) against the control hooks witha rate of 60.214 mpy (p<0.05).
Article
Full-text available
Although fishing is regarded as a risky production process, limited attention has been given to the impact of input factor use on production risk. Production risk is particularly important for fisher behavior and fisheries management when input factors are restricted, since input restrictions can influence production risk in addition to output levels. This paper investigates production risk by estimating production and risk functions for the main vessel groups in the Norwegian fishing fleet. The results indicate that production risk is present and that the effect of input use on production risk varies between vessel groups. Capital has a risk-reducing effect in the ocean fleet, but are risk-increasing in coastal fisheries. Fuel use is found to be a risk-increasing input for most of the vessel groups, while labor use is risk-reducing.
Article
Full-text available
In recent years, eco-labeling has become an increasingly used tool to signal sustainable sourcing of (sea-) food. While the literature has focused on price premiums associated with the labels, it is noted in this paper that eco-labels can also contribute to profitability by reducing cost, e.g. through longer product lifespans. Hence, eco-labels can be beneficial in a supply chain even without a price premium. This study applies duration analysis to explore whether two eco-labels (the MSC label of the Marine Stewardship Council and a line-caught label) influence product longevity of whitefish products in eight different grocery retailers in the UK. The results show that MSC labeled products have a 64.7% lower risk of being withdrawn from the shelves compared to non-MSC products, while products with the line-caught label have a 32.8% lower risk of being withdrawn than products without this label. The results also indicate that the influence of the eco-labels on product longevity vary considerably between the retailers.
Article
Full-text available
Catch-based aquaculture (CBA) is an important production system in many parts of the world, especially in developing countries. In Norway CBA is based on capture and storage/farming of mature, North-east Atlantic (NEA) cod (Gadus morhua). The objective is to reduce seasonal variations and add value, by storing/farming fish to take advantage of higher prices in low seasons. Despite numerous development programs and economic incentives, the development of the CBA business has been slow. Few actors are storing fish long enough to take advantage of high price in low season. A reason for this is that CBA is caught in between two sectors: the fisheries and aquaculture, with radically different institutional frameworks, creating entry barriers and a complex regulatory framework. Moreover, the legitimacy of the CBA is in question, as CBA intervene into the resource allocation mechanism in traditional fisheries. Still, it is too early to conclude that CBA has failed, as we are dealing with an industry in the making.
Article
Full-text available
Wild cod is a scarce and valuable natural resource. However, cod fishing along the coast of northern Norway has largely been about fishing as much as possible with the least possible resource effort, and thereby at the lowest cost. This traditional volume logic is rooted in biology, meteorology, and small scale capture technology. The logic is further enhanced by new large scale capture technology and a raw fish market where quality differences essentially are not reflected in the price of the fish. The intention of this paper is to address the extent to which the institutional framework in the Norwegian cod fisheries encourage or moderate this volume logic. In the paper fishing gear usage and the post-harvest industries’ product mix the last decade are analyzed. The findings show that the volume logic is still at work although this result in reduced quality of the catch landed, a product mix dominated by low-end products, and limited socio-economic value creation. This institutionalized volume logic is highly resistant to change. Moreover, the paper address how a competing and more customer-oriented quality logic can help create greater export values in Norwegian cod fisheries. Finally, implications are highlighted for how institutional measures can moderate the dominant volume logic and strengthen the emerging quality logic more in line with key policy objectives for Norwegian cod fisheries.
Article
Full-text available
Trawl-caught fish are frequently associated with deteriorated catch quality. This study presents a new dual sequential codend concept with the aim of improving the quality of trawl-caught fish by minimizing the frequency and severity of catch damage. During towing, the fish are retained in an anterior codend segment with the legislated mesh size. A quality improving codend segment , is attached to the aft part of the first codend segment. Its entrance is closed during the towing phase and opened at a predefined depth during haul-back. Comparing the quality of cod (Gadus morhua L.) retained in the sequential codend with cod caught in a conventional codend, demonstrated a significant improvement in the catch quality, i.e. reduction in catch damages. Cod caught in a conventional codend had only a 3.6% probability of being without visually detectable catch damage. The probability for catching cod without catch damage was five times higher when using the dual sequential codend. Furthermore, cod caught in the sequential codend had a significantly reduced probability of incurring specific catch damage, such as gear marks, poor exsanguination, ecchymosis, and skin abrasions.
Article
Full-text available
In attempts to differentiate their store imagery, grocery retailers frequently introduce new products, which are often rich in extrinsic attributes such as claims regarding healthiness and environmental sustainability. This paper explores retailers' pricing strategies for product attributes of haddock in the United Kingdom. The results show that retailers' pricing strategies vary, in particular for extrinsic product attributes such as eco-labels and country-of-origin. The high price premium generated from the most advocated eco-labels and an emphasis upon promotion strategies appear to play a role in the success of high-end retailers, and may have implications for any heightened competitive responses by larger retail chains in the future.
Article
The main focus in the inefficiency literature is on suboptimal input use and how this causes increased costs, due to technical and allocative inefficiency. Production or cost functions are then typically used to describe the underlying technology of the firm. The possible revenue loss, due to lower than maximum production levels and suboptimal output mix, has received substantially less attention. By using a revenue function to measure inefficiency, the focus, model and estimation technique presented in this article differ from those of previous studies. A shadow revenue model is used to decompose revenue inefficiency into its technical and allocative components, in which the allocative inefficiency is due to a suboptimal output mix. The approach is illustrated using panel data of Norwegian whitefish trawlers. The results reveal large inefficiencies, with respect to output levels as well as output mix, indicating that this can be an important part of the picture when investigating economic inefficiency. To identify the determinants of revenue inefficiency, we conduct a second-step regression, in which technical and allocative inefficiency is regressed upon a set of explanatory variables. The inefficiencies are partly explained by the management system and fleet structure.
Article
This paper explores the role of large buyers at the Norwegian pelagic auction. The auction is electronic, with no product inspections and conducted on a first-price, sealed-bid basis. Hedonic price modelling was applied to auction data for mackerel for the main seasons of 2013–2015, which comprise 2447 transactions accounting for a traded volume of 581,613 t. A key finding is that the largest buyers pay lower prices than smaller buyers, all else being equal. For example, holding transaction quantity and fish size constant, the largest buyer pay about 3% less than the small buyers, implying a discount of NOK49.9 million for the three years covered by the study. This is attributed to a better understanding of the auction, including other buyers’ valuations. In addition, only the largest buyers can handle the largest catches, so there is less competition and lower prices in this market segment. Another key finding is that, holding other factors constant, the largest buyers pay a lower price for fish size than that paid by small buyers. With transaction quantity held constant, the price discount for fish size is NOK 0.146 (per 10 g) for the largest buyer and the price premium for the small buyers is NOK 0.099 (per 10 g). Findings are discussed and compared with past research focusing on fish auctions. Practical implications are also discussed.
Article
The individual transferable quotas (ITQ) system is a popular fisheries management system applied in many countries. Here, vessels are scrapped or sold out of the fishery, and their quotas are transferred to the remaining vessels, resulting in lower number of active vessels. This consequently reduces the overcapacity of fishing fleets and likely improves financial performance. Unlike previous studies, this paper quantifies the impact of capacity adjustment on vessel profitability through an econometric model. Supply and demand shifters are included in the model to isolate the influence of capacity adjustment. The case studies are the total Norwegian fishing fleet and an important vessel group – the ocean-going cod trawlers. Empirical findings generally indicate a positive impact of capacity adjustment on vessel financial performance. The observed recent decline in financial profitability of the fishing fleet may therefore be attributed to other factors rather than overcapacity.
Article
Can a perishable food commodity with strong seasonal variation in supply, supplied by producers with limited operational radius, belong to a spatially integrated market? The majority of Norwegian cod is landed during a few winter months. Moreover, most of the catch is landed fresh by a fleet consisting primarily of many smaller vessels with limited operational radius. This makes landing patterns and potentially markets regional. However, there are also landings in the remainder of the year, primarily by larger oceangoing vessels such as trawlers with a larger operational radius that may ensure market integration between the regions. Market integration can also be ensured through landings of a potential substitute; frozen cod, which is less perishable. In this paper we investigate the degree of market integration between different regions for fresh cod in Norway, and between fresh and frozen cod. We find that the markets for fresh cod are integrated but not all regions adhere to the Law of One Price. Markets for fresh and frozen cod are also integrated, but not to the extent that the Law of One Price holds. Finally, there is evidence of frozen cod leading the fresh cod prices.