Marine Policy 66 (2016) 39–49

Contents lists available at ScienceDirect

Marine Policy
journal homepage: www.elsevier.com/locate/marpol

Identifying channels of economic impacts: An inter-regional structural

path analysis for Alaska fisheries
Chang K. Seung
Alaska Fisheries Science Center, National Marine Fisheries Service, National Oceanic and Atmospheric Administration, 7600 Sand Point Way NE, Seattle, WA
98115-6349, USA

art ic l e i nf o a b s t r a c t

Article history: Alaska fisheries have strong spillover effects on economies of other states (especially the state of Wa-
Received 4 November 2015 shington) due to their dependence on imports from these other states. Several studies attempt to develop
Received in revised form inter-regional or multi-regional economic impact models to investigate these spillover effects, and cal-
11 January 2016
culate the multipliers for Alaska fisheries. However, these multipliers measure only total economic im-
Accepted 12 January 2016
pacts, failing to provide fishery managers with the information on how and along what channels these
total economic impacts are generated and transmitted throughout the regions. This paper uses an inter-
Jel Classification: regional structural path analysis (IRSPA) to identify the various channels (paths) through which the
R15 economic impacts of an initial shock to a seafood sector are transmitted, amplified, and spilled over to
Q22
other regions, within an inter-regional social accounting matrix (IRSAM) framework for two US regions –
Keywords: Alaska and the rest of US (RUS).
Alaska fisheries Published by Elsevier Ltd.
Inter-regional economic impacts
Social accounting matrix
Structural path analysis

1. Introduction
Much of the labor income generated in seafood-related in-
dustries in Alaska flows out of the state because a large share of
workers are nonresidents. In 2010, about 20% of total private and
state and local government employment in Alaska was accounted
for by nonresidents. Consequently, about 14% of the total labor
income produced in private industries and state and local gov-
ernments in Alaska leaked out of the state. Outflows of labor in-
come are the largest in (i) seafood processing (65%)1; (ii) agri-
culture, forestry, fishing, and hunting (43%, mostly fishing); (iii)
mining (29%); (iv) accommodation (26%); (v) transportation and
warehousing (23%); and (vi) arts, entertainment, and recreation
(20%) sectors [1].
In addition, a large amount of capital used in Alaska industries,
including seafood, is owned by nonresidents. This means that
much of the capital income from these industries leaks to other
states. Many fishing vessels operating in waters off Alaska are
owned by non-Alaskan residents. Also, many of the goods and
services used by consumers and industries in Alaska are imported
from other states. In 2008, the total value of imports to Alaska ($16
E-mail address: Chang.Seung@noaa.gov
1
That is, 65% of labor income in this industry leaks out of the state.
billion) from non-Alaska US states accounted for about 31% of the
total value of production ($51 billion) in the state [20]. Therefore,
there are additional impacts from exogenous shocks to fisheries or
other industries in Alaska affecting those other states that are not
captured in a single-region economic impact model.
Several studies have used an inter-regional or multi-regional
economic impact model such as social accounting matrix (SAM)
model [19,20] to capture these additional impacts of Alaska fish-
eries, and calculated the inter- or multi-regional multipliers. For
example, Seung [19] calculated the economic impacts of Alaska
fisheries for three regions in the US including Alaska, the West
Coast, and the rest of US using multipliers from a three-region
SAM model. However, these multipliers measure only total eco-
nomic impacts, failing to provide fishery managers with the in-
formation on how and along what channels these total economic
impacts are generated and transmitted throughout the regions.
The present study departs from these studies, and uses an in-
ter-regional structural path analysis (IRSPA) to identify various
paths (channels) through which an initial shock to Alaska fisheries
generates inter-regional impacts within an inter-regional social
accounting matrix (IRSAM) for two regions, Alaska and the rest of
US. Conventional economic impact analysis would provide the
fishery managers only a multiplier number measuring the overall
economic impacts of a seafood industry. As such, it would not

http://dx.doi.org/10.1016/j.marpol.2016.01.015
0308-597X/Published by Elsevier Ltd.
40 C.K. Seung / Marine Policy 66 (2016) 39–49

explain how and through what paths the initial shocks are trans- Y =(I − S )−1X (2)
mitted in generating the total economic impacts.2 Unlike the
⎡y ⎤ ⎡ ⎤ ⎡x ⎤
conventional multiplier analysis, the IRSPA can reveal the me- ⎢ 1⎥ ⎢ Z11z12 ⎥ 1⎥
where Y = ⎢ ⎥, S =⎢ ⎥, and X =⎢⎢ ⎥. Here S is matrix of direct
chanism of interactions among different economic sectors in the ⎢⎣ y2 ⎥⎦ ⎢⎣ z Z ⎥⎦
21 22 ⎣ x2 ⎦

two regions, and serve as a complementary tool to the conven- IRSAM coefficients and (I − S )−1 is the IRSAM multiplier matrix.
tional multiplier analysis in fishery managers' decision-making. In Here (I  S)  1 is called the IRSAM multiplier matrix or matrix of
the next section (Section 2), a description of an IRSAM model for IRSAM inverse coefficients. yr is a column vector for region r
Alaska fisheries is provided. Section 3 describes the IRSPA with a consisting of the following endogenous sub-vectors:
brief review of previous structural path analysis (SPA) studies. Ar ¼vector of regional industry output
Section 4 provides a description of data used, which is followed by Qr ¼vector of regional commodity output
Section 5 where results are discussed. Conclusions follow. Vr ¼ vector of total primary factor payments
IBTr ¼ indirect business tax payments
Hr ¼vector of total household income
2. Alaska fisheries IRSAM model SGr ¼total state and local government income or revenue
Zrr for region r is:
This section describes the 2008 IRSAM model used in this ⎡ 0 Mr 0 0 0 0 ⎤
study. It relies on Seung [19] except that this section describes a ⎢ ⎥
⎢ Ur 0 0 0 Cr GDr ⎥
two-region IRSAM while Seung [19] describes a three-region SAM. ⎢ Vr 0 0 0 0 0 ⎥
Readers are referred to King [6] for a more detailed discussion of a Z rr = ⎢ ⎥
SAM, Holland and Wyeth [4] for a regional level SAM model, and ⎢ IBTr 0 0 0 0 0 ⎥
⎢ 0 0 Fr 0 0 STRr ⎥
Round [17] and Roberts [15] for the structure of an IRSAM. The ⎢ ⎥
structure of the IRSAM used in this study is similar to that in ⎣ 0 0 SFr BTSr HTXr IGTr ⎦
Round [17] and Roberts [15], and is available upon request.
where:
In the IRSAM, each region has 61 endogenous accounts; thus,
Ur ¼absorption matrix
with two regions there are a total of 122 (61
2) endogenous ac-
Vr ¼matrix of primary factor payments coefficients
counts in the IRSAM. The 61 endogenous accounts for each region
IBTr ¼ matrix of indirect business tax coefficients
include 28 industries, 26 commodities, 3 value-added accounts
Mr ¼market share matrix
(labor income, capital income, and indirect business tax), 3 house- Fr ¼matrix of factor payment to household coefficients
hold accounts (low-, medium-, and high-income households), and a SFr ¼matrix of state and local factor tax coefficients
state and local government account. The 28 industries (Table 1) BTSr ¼matrix of state and local indirect business tax
include 14 seafood industries (11 fish harvesting industries and coefficients
3 seafood processing industries) and 14 non-seafood industries. The Cr ¼matrix of household consumption coefficients
26 commodities include 11 fish species (corresponding to the 11 HTXr ¼matrix of state and local government direct household
fish harvesting industries), one processed seafood (which is an ag- tax coefficients
gregation of the commodities produced in the three seafood pro- GDr ¼matrix of state and local government demand coefficients
cessing industries), and 14 non-seafood commodities (correspond- STRr ¼ matrix of state and local government transfer
ing to the 14 non-seafood industries). Major species (commodities) coefficients
in the model are Pacific cod, pollock, sablefish, crab, halibut, and IGTr ¼ matrix of intergovernmental transfers
salmon. The three processing industries are catcher-processors, zrs is:
motherships, and shorebased processors.
⎡0 0 0 0 0 0⎤
The IRSAM has 4 exogenous accounts, which include the federal ⎢ ⎥
government, capital (savings and investment), an account to handle ⎢ 0 IMrs 0 0 0 0⎥
international trade and financial flows, and an account balancing ⎢ 0 0 LKrs 0 0 0⎥
z rs = ⎢ ⎥
between the two regions and the rest of the world (ROW). ⎢0 0 0 0 0 0⎥
The IRSAM model is represented as: ⎢0 0 0 0 0 0⎥
⎢⎣ 0 0 0 0 0 0⎥⎦
⎡ y1 ⎤ ⎡ Z11z12 ⎤ ⎡ y1 ⎤ ⎡ x1 ⎤
⎢ y ⎥=⎢ ⎥ ⎢ ⎥+
⎣ 2 ⎦ ⎣ z21Z22 ⎦ ⎣ y2 ⎦ ⎢⎣ x2 ⎥⎦ (1) where IMrs is matrix of imports from region r to s and LKrs is
matrix of leakage of factor income from region s to region r. xr is a
where yr and xr denote the column vectors of endogenous and column vector consisting of the following exogenous sub-vectors:
exogenous accounts, respectively, for region r and Zrr is a sub- ear ¼vector of exogenous demand for regional industry output
matrix containing coefficients showing the intra-regional trans- eqr ¼ vector of exogenous demand for regional commodity
actions and zrs a submatrix containing coefficients showing inter- output
regional transactions (i.e., transactions between regions r and s, evr ¼vector of exogenous factor payments
r≠s). All the coefficients in Zrr and zrs matrices are derived by di- etr ¼ exogenous indirect business tax payments
viding the elements in the columns in the IRSAM by the column ehr ¼vector of exogenous federal transfers to households
totals. egr ¼federal transfers to state and local government.
Alternatively, Eq. (1) can be expressed as following: There are three non-zero exogenous demand vectors – eqr, ehr
and egr. The elements of eqr are components of final demand for
2
This may be an overstatement because conventional economic impact ana- commodities including Federal government demand, investment
lysis conducted using input–output (IO) and SAM models can also identify the demand, and export demand. The elements of ehr include Federal
upstream (i.e., backward linkage) sectors that will be impacted by a policy shock. government transfers to households and remittances from ROW to
Compared with conventional economic impact analysis, however, a structural path
households. The components of egr include Federal government
analysis goes one step further because it can investigate the concentration,
strength, and speed of various transmission channels of economic impacts transfers to state and local government. Injections of income into a
generated. region occur through final demand components in eqr and extra-
 C.K. Seung / Marine Policy 66 (2016) 39–49
Table 1
Industry aggregation scheme for IRSAM model.
IMPLAN Sectors
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 16 (Replaced with estimated data)
Sector 71 (Replaced with estimated data)
Sector 71 (Replaced with estimated data)
Sector 71 (Replaced with estimated data)
Sectors 1–15, 17, and 18
Sector 19
Sectors 20–29
Sectors 30–32, 495, and 498
Sectors 33–45
Sectors 46–70, 72–84, 85–141, and 143–389
Sectors 142 and 396
Sector 390
Sectors 391–395, and 397–400
Sectors 401–412
Sectors 413–424
Sectors 425–436
Sectors 437–494
Sectors 496, 497, 499–506, and 507–509
a
Miscellaneous includes all the other sectors not included in the sectors above, that is, government and
non-NAICS, state and local government services, and federal government services.
regional payment components in ehr and egr. Leakages include
taxes paid to the Federal government, savings, and payments for
commodities imported from ROW.
3. Inter-regional structural path analysis (IRSPA)
Regional economists have used SPAs to identify the paths
through which the economic impacts of a certain policy for, or an
exogenous shock to, an economic sector (origin) on another sector
(destination) are generated and transmitted. SPA investigates the
concentration, strength, and speed of various transmission channels
or paths [13]. Concentration refers to the share of total economic
impacts from an exogenous shock that travel through one or more
paths that connect different endogenous sectors in a SAM. Strength
is measured by the size of the contribution of a path to the total
impacts represented by the SAM multipliers. Speed relates to the
share of the contribution of the path that connects directly from the
origin to the destination without passing through any sector more
than once. The larger the number of transactions among the sectors,
the longer time it is likely to take for the transmission of the effects
along the paths with higher lengths to materialize.
Lantner [7] first developed the SPA method, and utilized the
method within an input–output (IO) model. The first application of
the method to a SAM model is found in [2]. Within a SAM model
framework for the Republic of Korea, they used the analysis to
investigate the effects of specific production activities on other
production activities, factors of production, and different types of
households. Since then, economists have used the method for
different issues, including, for example, environmental issues (e.g.,
[24,8]), the role of different types of households in the economy of
a rural area [16], differential effects of economic growth on labor
income by gender [13]. Parra and Wodon [13] and Kasumba and
Ruhiiga [5] review the previous SPA applications in detail.
The application of the method to fisheries is rare. There is only
41
Industries in IRSAM
Pacific whiting harvesting
Atka mackerel harvesting
Flatfish and rockfish harvesting
Pacific cod harvesting
Pollock harvesting
Sablefish harvesting
Crab harvesting
Halibut harvesting
Salmon harvesting
Herring and other finfish harvesting
Other fish harvesting
Catcher-processor (processing only)
Mothership
Shorebased processor
Agriculture
Oil and gas extraction
Other mining
Utilities
Construction
Other manufacturing
Refined petroleum
Wholesale trade
Transportation
Retail
Information
Finance, insurance, real estate, renting, and leasing (FIRERL)
Services
Miscellaneousa
one study that used an SPA to investigate the economic impacts of
fisheries within a single-region SAM framework. Seung [21] used
an SPA for the fisheries of the Southeast Alaska. The present study
is the first to conduct an IRSPA for fisheries. The description of the
SPA method below is based on [2,,10,13,21]. For details, see [2].
3.1. Basic definitions
Suppose that the coefficient matrix for an IRSAM is denoted by S,
and the elements in the matrix are denoted by sji, where i and j denote
any two endogenous accounts in the IRSAM. The major difference
between an SPA for a single-region model and an IRSPA is that i and j
could be any two endogenous accounts for the two regions, r and s, in
an IRSPA; in a SPA for a single-region model, the two accounts are
defined for only the single region. The elements of the resulting IRSAM
multiplier matrix, (I S)  1, are denoted by mji. Any endogenous ac-
count in the IRSAM is considered a pole. Any two poles, i and j, are
connected by an arc starting from i and ending at j, called arc(i, j). An
element in S, sji, is the intensity of arc(i, j). A path is defined by a se-
quence of consecutive arcs. The length of a path is equal to its number
of arcs. An elementary path is a path that does not pass through any
pole more than once. A circuit is a path that starts from and ends at the
same pole. Three kinds of influence can be measured between any
two endogenous accounts in an IRSAM: direct influence, total influ-
ence, and global influence. Let (i-j)p denote the path p from i to j. See
Fig. 1 below, which shows the network of elementary paths and ad-
jacent circuits linking multiple poles.
3.2. Direct influence
The direct influence of i on j, through an elementary path, is the
change in the income (i.e., revenue) of account j (destination) in-
duced by a unitary increase in i (origin), where the incomes of all
the other accounts except those along the elementary path are as-
sumed to be constant. The direct influence from i to j is calculated
42
smk
k m
ski skm
skf sfm
i f
Fig. 1. Poles, elementary paths, and adjacent circuits (adapted from Fig. 4 in [10]).
by multiplying the intensities of all arcs forming the elementary
path. For example, the direct influence induced by a three-arc ele-
mentary path connecting poles i, k, m, and j in Fig. 1, is given by
I(Di → j) p=ski smk sjm
where k and m are the intermediate poles located along the ele-
mentary path between poles i and j.
3.3. Total influence
Along an elementary path from i to j, there may be some effects
transpiring through the circuits adjacent to the path. These circuits
will amplify the direct influence. Total influence includes both the
direct influence and the indirect effects produced through the
adjacent circuits. Fig. 1 shows how these indirect effects occur.
Consider an elementary path whose origin is i and the destination
j. First, the influence from i to k and that from m to j are given by
the direct influences, ski and sjm, respectively.
The connection between the two poles, k and m, needs further
explanation. One way to go from k to m is to travel along the ele-
mentary path, arc(k,m). The direct influence along this path (k-m)
is smk. Another way to go from k to m is to travel around (i) the
circuit between k and m and (ii) the circuit that links k and m
through f. After one round of feedback, the influence from k to m is
given by smk 2skm+sfm skf smk 2 (skm+sfm skf ). After t rounds of feedback,
the influence is given by smk [smk (skm + sfm skf )]t smk (smk + sfm skf )t . Fi-
nally, any influence has to be transmitted from m to j. Based on a
geometric series argument, total influence is derived as:
I(Ti → j) p = ski smk sjm ⎡⎣ 1−smk ( skm + sfm skf ) ⎤⎦ =I(Di → j) p Mp.
−1
Here, Mp=[1 − smk (skm + sfm skf )]−1 is called path multiplier.
Therefore, Eq. (4) implies that the total influence is equal to the
direct influence multiplied by the path multiplier.
3.4. Global influence
The global influence from i to j is simply given by element mji
from the IRSAM multiplier matrix, (I  S)  1, and is the sum of the
total influences along all possible elementary paths connecting i
and j (not just the paths shown in Fig. 1):
n functions to regional industries, including fisheries. While this assumption may not
I(Gi → j) =mji = ∑ I(Ti → j) p
p=1
where n is the number of all possible elementary paths going from
i to j. The output multiplier for i, which is often calculated in
conventional economic impact analysis, is simply given by
J J
∑j = 1 I(Gi → j ) = ∑j = 1 mji where J is the total number of industries.
4. Data
In 2012, about 52% by weight of the total US commercial fishing
harvest came from the waters off Alaska [12,3]. This accounted for
C.K. Seung / Marine Policy 66 (2016) 39–49
about 4.8 billion pounds of fish and shellfish with an ex-vessel value
of about $1.9 billion. Of this, groundfish accounted for 54.2% of ex-
vessel value, salmon 22.7%, halibut 7.4%, shellfish 14.6%, and herring
1% [3]. Of the groundfish species caught in Alaska waters, pollock
has the highest commercial value. In 2012, pollock catch totaled
1.31 million metric tons, accounting for 60% of the total groundfish
sjm catch and 50% of ex-vessel value for groundfish, or $497 million [3]
The IRSAM in this study has two regions, Alaska and the rest of
US (RUS). RUS includes all non-Alaska US states. Data for seafood
j industries in IMPLAN (IMpact analysis for PLANning; [11]) are not
reliable,3 and may produce biased results. Therefore, it was ne-
cessary to use more reliable data for the seafood industries. The
Alaska fish harvesting and processing sectors were constructed
independently using data from more reliable sources. Detailed
descriptions of the data assembly for the Alaska seafood industries
are found in [22]. The Alaska seafood industry data thus estimated
(3) are combined with 2004 IMPLAN non-seafood industry data to
generate Alaska SAM. See [18] for a more detailed description of
how the data was assembled for Alaska SAM.
Data for RUS region was based on (i) the 2006 input-output (IO)
model for Pacific Coast Fisheries (IO-PAC) developed by the Northwest
Fisheries Science Center (NWFSC), (ii) 2008 NMFS landings data,4 and
(iii) 2008 IMPLAN data for non-Alaska states. IO-PAC is an IMPLAN-
based model augmented with fishery-related industry and commodity
detail generated using data collected from economic surveys of vessels
participating in West Coast fisheries [9]. Since the data years for these
data sets are different, these data were adjusted to 2008 levels using
the national GDP price deflator series5 in order to match the data
years. Trade flows among the two regions were estimated using IM-
PLAN version 3 based on 2008 data. This procedure required con-
structing additional IMPLAN models for the Alaska and RUS regions for
2008, and executing the interregional commodity trade flow estima-
tion utility in IMPLAN version 3.
Additionally, it was necessary to estimate interregional non-
commodity flows (factor income payments, transfer payments,
and financial flows) between the two regions, and with respect to
ROW. Data in the Alaska SAM regarding utilization of nonresident
labor in fishing and other Alaska industries were useful for this
purpose. However, there were no data available to indicate origin
or destination of interregional transfer payments and financial
flows. This study assumed simply that these interregional flows
are zero. This assumption may serve to understate the inter-
(4)
regional linkages to some extent. For more details on preparing
the data for the IRSAM, see [20] and [19].
5. Results
This study presents results from IRSPA6 for two origin sectors
3
IMPLAN data are widely used for regional economic analysis of sub-regions in
the US. However, when the data are used for analysis of fisheries, the data suffer
from some critical weaknesses. First, IMPLAN applies national-level production
be problematic for many regional industries, use of average production relation-
(5) ships may not accurately depict regional fish harvesting and processing technolo-
gies. Second, the employment and earnings of many crew members in the com-
mercial fishing sector are not included in the IMPLAN data because IMPLAN is
based on state unemployment insurance program data which exclude “uncovered”
employees such as self-employed and casual or part-time workers.
4
〈http://www.st.nmfs.noaa.gov/st1/commercial/landings/annual_landings.
html〉.
5
Another way of making this adjustment is to use nominal GDP ratios to in-
flate 2004 and 2006 to 2008 levels. Doing this may better adjust for both scale as
well as the price effects.
6
The SPA was implemented using a software called SimSIP SAM developed by
Juan Carlos Parra and Quentin Wodon at The World Bank. Parra and Wodon [14]
present procedures on how to implement SPA and other SAM-related analysis
 C.K. Seung / Marine Policy 66 (2016) 39–49
(accounts) in the IRSAM, Alaska pollock commodity (AK-PLCK-C, i.e.,
raw pollock) and Alaska processed seafood commodity (AK-SFOOD-
C).7 The destination sectors include industries, value added,
households, and state and local government, in both of the two
regions. Table 2 presents the IRSPA results for only the top ten
destination industries in the two regions when the origin is Alaska
pollock commodity. The top ten destination industries are de-
termined by the magnitude of the global influence. Table 3 reports
the results for top 16 destination industries in the two regions when
the origin is Alaska processed seafood commodity. Table 4 presents
the IRSPA results for the three value-added sectors for each of the
two regions while Table 5 reports the results for the three house-
hold sectors and the state and local government for each of the two
regions. Both tables (Tables 4 and 5) show results for each of two
origin sectors. All these tables (Tables 2, 3, 4, and 5) shows only
those impacts which are generated along a maximum of six arcs in
the elementary paths and whose ratio of total influence to global
influence (TG ratio)8 is larger than 3%. This impact threshold is used
in the presentation of the results because of space constraints.
Acronyms are used in the text and the tables to denote en-
dogenous sectors in the IRSAM. See Table A.1 in Appendix. In Table
A.1, the acronyms used for industries end with “-A”. Although not
shown in Table A.1, the acronyms for commodities end with “-C”.
Also, the regional notations, AK- and RU-, are used to denote the
two regions. For example, AK-PLCK-A, RU-SERV-C, and AK-MED-
HH denote, respectively, Alaska pollock industry, RUS service
commodity, and Alaska medium-income households.
Table 2 shows that the global influence of Alaska pollock com-
modity on, for example, the Alaska service industry (mji from IRSAM
multiplier matrix, destination sector) is 0.209, which implies that a
$1 million increase in demand for Alaska pollock commodity (i.e.,
raw pollock) will result in a total increase in Alaska service industry
output of $209,000. Six of the top 10 industries which AK-PLCK-C
has the largest global influence over are Alaska industries (AK-PLCK-
A, AK-OILG-A, AK-REFN-A, AK-FIRERL-A, AK-SERV-A, and AK-MIS-
CEL-A) while the other four industries are RUS industries (RU-
OMAN-A, RU-FIRERL-A, RU-SERV-A, and RU-MISCEL-A).
Results indicate that there are several elementary paths through
which the initial impacts are transmitted to the destination sector.
Direct influence of Alaska pollock commodity (raw pollock) can
occur along a three-arc elementary path. For example, the three-arc
elementary path denoted by AK-PLCK-C-AK-PLCK-A-AK-SERV-
C-AK-SERV-A states that production of Alaska pollock commodity
(catch of pollock) by Alaska pollock harvesting industry (AK-PLCK-
C-AK-PLCK-A) requires purchase by Alaska pollock harvesting in-
dustry of Alaska service commodity (AK-PLCK-A-AK-SERV-C),
which is produced by Alaska service industry (AK-SERV-C-AK-
SERV-A). In short, the three-arc elementary path represents direct
expenditures made by Alaska pollock industry, in order to produce
Alaska pollock commodity, on Alaska service commodity (which is
used as an intermediate input in Alaska pollock industry) produced
by Alaska service industry.
The direct influence for this three-arc path (0.028) is multiplied
by the path multiplier of 1.310 that captures the effects occurring
in the adjacent circuits between the two poles (AK-PLCK-C and AK-
SERV-A). The resulting number (0.036 ¼0.028
1.310) is the total
influence. A larger path multiplier means that the effects gener-
ated from the circuits between origin and destination poles are
stronger than the effects with a smaller path multiplier. 17.3% of
(footnote continued)
using the software.
7
For brevity, this section reports results for only two seafood sectors. The re-
sults for other origin sectors are available upon request.
8
TG ratio measures the relative contribution of an elementary path in gen-
erating the impacts of an origin sector on a destination sector.
43
global influence is accounted for by the total influence transpiring
along the three-arc elementary path.
The direct influence can also occur along more than three arcs
between the origin and destination sectors. For example, the path
denoted by AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-
SERV-C-AK-SERV-A states that an increase in demand for AK-
PLCK-C produced by AK-PLCK-A results in an increase in labor
income (AK-LAB) in Alaska, which is then distributed to Alaska
high-income household (AK-HIGH-HH), and finally increases the
demand (by the high-income household) for Alaska service com-
modity (AK-SERV-C) produced by Alaska service industry (AK-
SERV-A). The TG ratio for this path is 16.8%.
The sum of all TG ratios from all the elementary paths con-
necting an origin sector and a destination sector equal 100%.
However, since this study presents only the impacts produced along
a maximum of six arcs in the elementary paths and with TG ratios
larger than 3%, the sum for most of the destination sectors in the
tables does not sum to 100%. For example, the sum of the TG ratios
for Alaska service industry (destination) reported in Table 2 is equal
to only 62.0% (17.3þ16.8þ11.5þ7.3þ4.9 þ4.3) because it excludes
the impacts generated along the paths whose length is larger than
six and the impacts whose TG ratios are smaller than 3%.
In some cases, the impacts pass through industry sectors other
than the destination sector. For example, the impacts represented
by the five-arc path, AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-
REFN-A-AK-SERV-C-AK-SERV-A, are generated through an in-
termediate industry sector (AK-REFN-A). This path states that har-
vest of Alaska pollock requires use of an intermediate input (AK-
REFN-C) produced by AK-REFN-A, which in turn purchases AK-
SERV-C. In other cases, the poles which the initial impacts pass
through are value-added sectors and household sectors. When the
destination sector is AK-SERV-A, the sum of TG ratios for the ele-
mentary paths connecting the origin and the destination and pas-
sing through value-added (factor income) sectors and household
sectors (medium- and high-income households) is equal to 40.5%
(¼16.8þ11.5þ7.3þ 4.9), which accounts for a significant share of
the global influence. Note that this number (40.5%) is the sum of
only the TG ratios reported in the table. If the sum includes the TG
ratios not shown in the table, the resulting sum could be much
larger than 40.5%. This indicates the important role of these value-
added and household sectors in generating the economic impacts.
Results indicate that the RUS industry that the Alaska pollock
commodity has the largest impacts (global influence) on is the RUS
service industry (RU-SERV-A), as evidence by the largest global
influence (0.363). The next two RUS industries that the origin
sector has the largest global influence over are RUS other manu-
facturing industry (RU-OMAN-A) (0.323) and RUS FIRERL industry
(RU-FIRERL) (0.239). In the case of the RUS other manufacturing
industry, the largest percentage (25.6%) of global influence is ac-
counted for by the total influence generated along the four-arc
path represented by AK-PLCK-C-AK-PLCK-A-AK-OMAN-C-RU-
OMAN-C-RU-OMAN-A. This path states that the Alaska pollock
industry uses commodities from other manufacturing industry in
Alaska (AK-OMAN-A), but that, because the supply of the com-
modity produced in Alaska (AK-OMAN-C) cannot satisfy the de-
mand, the Alaska pollock industry imports the commodity from
RUS, which is denoted by the arc, AK-OMAN-C-RU-OMAN-C, in
the four-arc path.
The increase in outputs of some RUS industries arising from
increased Alaska pollock catch occurs either (i) through an in-
crease in exports to Alaska of RUS commodities due to an increase
in Alaska household income or (ii) through an increase in con-
sumption by RUS households of the RUS commodities due to an
increase in RUS household income. An example of (i) above is
represented by the path, AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-
HI-HH-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A. This path
 44
Table 2
Results from IRSPA for Alaska pollock commodity (Origin¼ Alaska pollock commodity).

Destination Global influence Elementary paths Direct influence Path multiplier Total influence Total/Global (%)

AK-PLCK-A 0.983 AK-PLCK-C-AK-PLCK-A 0.983 1.000 0.983 100.0

AK-OILG-A 0.091 AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A-AK-OILG-C-AK-OILG-A 0.043 1.451 0.063 68.6
AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-OILG-A 0.008 1.449 0.012 13.1
AK-REFN-A 0.125 AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A 0.093 1.186 0.110 88.5
AK-FIRERL-A 0.124 AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-AK-FIRERL-A 0.049 1.221 0.060 48.3
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-FIRERL-C-AK-FIRERL-A 0.005 1.583 0.009 6.9

C.K. Seung / Marine Policy 66 (2016) 39–49

AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-MED-HH-AK-FIRERL-C-AK-FIRERL-A 0.005 1.581 0.009 6.9
AK-SERV-A 0.209 AK-PLCK-C-AK-PLCK-A-AK-SERV-C-AK-SERV-A 0.028 1.310 0.036 17.3
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-SERV-C-AK-SERV-A 0.023 1.547 0.035 16.8
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-MED-HH-AK-SERV-C-AK-SERV-A 0.016 1.548 0.024 11.5
AK-PLCK-C-AK-PLCK-A-AK-CAP-AK-HI-HH-AK-SERV-C-AK-SERV-A 0.010 1.473 0.015 7.3
AK-PLCK-C-AK-PLCK-A-AK-CAP-AK-MED-HH-AK-SERV-C-AK-SERV-A 0.007 1.451 0.010 4.9
AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A-AK-SERV-C-AK-SERV-A 0.006 1.547 0.009 4.3
AK-MISCEL-A 0.105 AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-MISCEL-C-AK-MISCEL-A 0.012 1.382 0.017 16.2
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-MED-HH-AK-MISCEL-C-AK-MISCEL-A 0.008 1.382 0.011 10.4
AK-PLCK-C-AK-PLCK-A-AK-CAP-AK-HI-HH-AK-MISCEL-C-AK-MISCEL-A 0.006 1.319 0.007 7.1
AK-PLCK-C-AK-PLCK-A-AK-CAP-AK-MED-HH-AK-MISCEL-C-AK-MISCEL-A 0.004 1.301 0.005 4.5
RU-OMAN-A 0.323 AK-PLCK-C-AK-PLCK-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A 0.042 1.950 0.083 25.6
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A 0.006 2.592 0.016 5.1
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-MED-HH-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A 0.005 2.589 0.014 4.4
RU-FIRERL-A 0.239 AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A 0.028 1.874 0.053 22.2
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A 0.003 2.430 0.008 3.2
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-MED-HH-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A 0.003 2.426 0.008 3.2
RU-SERV-A 0.363 AK-PLCK-C-AK-PLCK-A-AK-SERV-C-RU-SERV-C-RU-SERV-A 0.009 2.359 0.022 6.1
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-SERV-C-RU-SERV-C-RU-SERV-A 0.008 2.785 0.022 6.0
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-MED-HH-AK-SERV-C-RU-SERV-C-RU-SERV-A 0.005 2.788 0.015 4.1
AK-PLCK-C-AK-PLCK-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A-RU-SERV-C-RU-SERV-A 0.004 3.170 0.012 3.3
RU-MISCEL-A 0.119 AK-PLCK-C-AK-PLCK-A-AK-LAB-RU-LAB-RU-HI-HH-RU-MISCEL-C-RU-MISCEL-A 0.002 2.648 0.004 3.4

Note: the acronyms for industries used in this table end with “-A” and the acronyms for commodities end with “-C”. IBT¼ indirect business tax; LOW-HH¼ low-income households; MED-HH¼ medium-income households; HI-
HH¼ high-income households; SLG ¼state and local government.
Table 3
Results from IRSPA for Alaska processed seafood commodity (Origin ¼Alaska processed seafood commodity).

Destination Global influence Elementary paths Direct influence Path multiplier Total influence Total/Global (%)

AK-PLCK-A 0.106 AK-SFOOD-C-AK-CPP-A-AK-PLCK-C-AK-PLCK-A 0.050 1.001 0.050 47.3

AK-SFOOD-C-AK-SHOREP-A-AK-PLCK-C-AK-PLCK-A 0.046 1.001 0.046 43.8
AK-SFOOD-C-AK-MS-A-AK-PLCK-C-AK-PLCK-A 0.008 1.001 0.008 8.0
AK-SFOOD-C-AK-PLCK-A 0.001 1.001 0.001 0.9
AK-SALM-A 0.087 AK-SFOOD-C-AK-SHOREP-A-AK-SALM-C-AK-SALM-A 0.078 1.001 0.078 89.2
AK-SFOOD-C-AK-MS-A-AK-SALM-C-AK-SALM-A 0.007 1.001 0.007 7.8
AK-SFOOD-C-AK-SALM-A 0.001 1.001 0.001 1.5
AK-CPP-A 0.312 AK-SFOOD-C-AK-CPP-A 0.311 1.001 0.312 100.0
AK-MS-A 0.079 AK-SFOOD-C-AK-MS-A 0.079 1.001 0.079 100.0
AK-SHOREP-A 0.604 AK-SFOOD-C-AK-SHOREP-A 0.603 1.001 0.604 100.0
AK-OILG-A 0.071 AK-SFOOD-C-AK-CPP-A-AK-REFN-C-AK-REFN-A-AK-OILG-C-AK-OILG-A 0.011 1.452 0.016 22.7
AK-SFOOD-C-AK-MS-A-AK-REFN-C-AK-REFN-A-AK-OILG-C-AK-OILG-A 0.002 1.452 0.004 5.0
AK-SFOOD-C-AK-SHOREP-A-AK-REFN-C-AK-REFN-A-AK-OILG-C-AK-OILG-A 0.002 1.452 0.003 4.6
AK-SFOOD-C-AK-CPP-A-AK-REFN-C-AK-OILG-A 0.002 1.450 0.003 4.3
AK-REFN-A 0.091 AK-SFOOD-C-AK-CPP-A-AK-REFN-C-AK-REFN-A 0.024 1.187 0.028 31.0
AK-SFOOD-C-AK-MS-A-AK-REFN-C-AK-REFN-A 0.005 1.187 0.006 6.8
AK-SFOOD-C-AK-SHOREP-A-AK-REFN-C-AK-REFN-A 0.005 1.187 0.006 6.3

C.K. Seung / Marine Policy 66 (2016) 39–49

AK-SFOOD-C-AK-CPP-A-AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A 0.005 1.187 0.006 6.2
AK-SFOOD-C-AK-SHOREP-A-AK-Halibut-C-AK-Halibut-A-AK-REFN-C-AK-REFN-A 0.004 1.187 0.005 5.9
AK-SFOOD-C-AK-SHOREP-A-AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A 0.004 1.187 0.005 5.8
AK-SFOOD-C-AK-SHOREP-A-AK-SALM-C-AK-SALM-A-AK-REFN-C-AK-REFN-A 0.002 1.187 0.003 3.2
AK-SFOOD-C-AK-CPP-A-AK-PacCod-C-AK-PacCod-A-AK-REFN-C-AK-REFN-A 0.002 1.188 0.003 3.1
AK-RETAIL-A 0.097 AK-SFOOD-C-AK-SHOREP-A-AK-RETAIL-C-AK-RETAIL-A 0.018 1.075 0.019 19.7
AK-SFOOD-C-AK-CPP-A-AK-RETAIL-C-AK-RETAIL-A 0.012 1.075 0.013 13.7
AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-HI-HH-AK-RETAIL-C-AK-RETAIL-A 0.004 1.375 0.005 5.3
AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-MED-HH-AK-RETAIL-C-AK-RETAIL-A 0.003 1.372 0.004 4.5
AK-FIRERL-A 0.138 AK-SFOOD-C-AK-CPP-A-AK-FIRERL-C-AK-FIRERL-A 0.018 1.222 0.022 16.2
AK-SFOOD-C-AK-SHOREP-A-AK-FIRERL-C-AK-FIRERL-A 0.013 1.222 0.016 11.6
AK-SFOOD-C-AK-SHOREP-A-AK-SALM-C-AK-SALM-A-AK-FIRERL-C-AK-FIRERL-A 0.005 1.222 0.006 4.1
AK-SFOOD-C-AK-MS-A-AK-FIRERL-C-AK-FIRERL-A 0.004 1.222 0.005 3.5
AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-HI-HH-AK-FIRERL-C-AK-FIRERL-A 0.003 1.583 0.004 3.1
AK-SFOOD-C-AK-SHOREP-A-AK-Halibut-C-AK-Halibut-A-AK-FIRERL-C-AK-FIRERL-A 0.003 1.222 0.004 3.1
AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-MED-HH-AK-FIRERL-C-AK-FIRERL-A 0.003 1.581 0.004 3.1
AK-SERV-A 0.248 AK-SFOOD-C-AK-SHOREP-A-AK-SERV-C-AK-SERV-A 0.020 1.311 0.026 10.5
AK-SFOOD-C-AK-CPP-A-AK-SERV-C-AK-SERV-A 0.014 1.311 0.018 7.2
AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-HI-HH-AK-SERV-C-AK-SERV-A 0.011 1.547 0.018 7.1
AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-MED-HH-AK-SERV-C-AK-SERV-A 0.008 1.549 0.012 4.8
AK-SFOOD-C-AK-CPP-A-AK-LAB-AK-HI-HH-AK-SERV-C-AK-SERV-A 0.005 1.547 0.008 3.4
AK-MISCEL-A 0.114 AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-HI-HH-AK-MISCEL-C-AK-MISCEL-A 0.006 1.382 0.009 7.5
AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-MED-HH-AK-MISCEL-C-AK-MISCEL-A 0.004 1.382 0.005 4.8
AK-SFOOD-C-AK-CPP-A-AK-LAB-AK-HI-HH-AK-MISCEL-C-AK-MISCEL-A 0.003 1.382 0.004 3.6
RU-OMAN-A 0.361 AK-SFOOD-C-AK-SHOREP-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A 0.025 1.951 0.048 13.3
AK-SFOOD-C-AK-CPP-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A 0.009 1.951 0.017 4.8
RU-WHOLE-A 0.091 AK-SFOOD-C-AK-CPP-A-AK-WHOLE-C-RU-WHOLE-C-RU-WHOLE-A 0.006 1.171 0.007 7.4
AK-SFOOD-C-AK-SHOREP-A-AK-WHOLE-C-RU-WHOLE-C-RU-WHOLE-A 0.005 1.171 0.006 7.0
AK-SFOOD-C-AK-SHOREP-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A-RU-WHOLE-C-RU-WHOLE-A 0.001 2.141 0.003 3.3
RU-RETAIL-A 0.080 AK-SFOOD-C-AK-SHOREP-A-AK-RETAIL-C-RU-RETAIL-C-RU-RETAIL-A 0.003 1.236 0.004 4.6
AK-SFOOD-C-AK-CPP-A-AK-RETAIL-C-RU-RETAIL-C-RU-RETAIL-A 0.002 1.236 0.003 3.2
RU-FIRERL-A 0.265 AK-SFOOD-C-AK-CPP-A-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A 0.011 1.876 0.020 7.5
AK-SFOOD-C-AK-SHOREP-A-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A 0.008 1.876 0.014 5.3
RU-SERV-A 0.409 AK-SFOOD-C-AK-SHOREP-A-AK-SERV-C-RU-SERV-C-RU-SERV-A 0.007 2.360 0.016 3.9

Note: the acronyms for industries used in this table end with “-A” and the acronyms for commodities end with “-C”. IBT¼ indirect business tax; LOW-HH¼ low-income households; MED-HH¼ medium-income households; HI-
HH¼ high-income households; SLG ¼state and local government.

45
 46
Table 4
Results for value added.

Origin Destination Global influence Elementary paths Direct influence Path multiplier Total influence Total/Global (%)

AK-PLCK-C AK-LAB 0.533 AK-PLCK-C-AK-PLCK-A-AK-LAB 0.265 1.304 0.345 64.8

AK-PLCK-C-AK-PLCK-A-AK-RETAIL-C-AK-RETAIL-A-AK-LAB 0.014 1.338 0.019 3.5
AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A-AK-OILG-C-AK-OILG-A-AK-LAB 0.009 1.872 0.016 3.0
AK-PLCK-C AK-CAP 0.475 AK-PLCK-C-AK-PLCK-A-AK-CAP 0.292 1.070 0.313 65.9
AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-AK-FIRERL-A-AK-CAP 0.019 1.282 0.024 5.1
AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A-AK-OILG-C-AK-OILG-A-AK-CAP 0.013 1.546 0.020 4.2
AK-PLCK-C AK-IBT 0.047 AK-PLCK-C-AK-PLCK-A-AK-RETAIL-C-AK-RETAIL-A-AK-IBT 0.005 1.129 0.006 12.9
AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A-AK-OILG-C-AK-OILG-A-AK-IBT 0.002 1.540 0.004 8.1
AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-AK-FIRERL-A-AK-IBT 0.003 1.288 0.004 7.9

C.K. Seung / Marine Policy 66 (2016) 39–49

AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A-AK-IBT 0.003 1.260 0.003 6.7
AK-PLCK-C-AK-PLCK-A-AK-WHOLE-C-AK-WHOLE-A-AK-IBT 0.002 1.091 0.003 5.6
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-RETAIL-C-AK-RETAIL-A-AK-IBT 0.001 1.420 0.001 3.1
AK-PLCK-C RU-LAB 0.476 AK-PLCK-C-AK-PLCK-A-AK-LAB-RU-LAB 0.032 2.326 0.075 15.8
AK-PLCK-C-AK-PLCK-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A-RU-LAB 0.007 3.035 0.020 4.2
AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A-RU-LAB 0.006 2.898 0.019 3.9
AK-PLCK-C RU-CAP 0.329 AK-PLCK-C-AK-PLCK-A-AK-CAP-RU-CAP 0.027 1.311 0.035 10.6
AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A-RU-CAP 0.010 2.120 0.021 6.4
AK-PLCK-C RU-IBT 0.059 AK-PLCK-C-AK-PLCK-A-AK-WHOLE-C-RU-WHOLE-C-RU-WHOLE-A-RU-IBT 0.003 1.321 0.003 5.8
AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A-RU-IBT 0.002 2.078 0.003 5.4
AK-SFOOD-C AK-LAB 0.618 AK-SFOOD-C-AK-SHOREP-A-AK-LAB 0.132 1.304 0.173 27.9
AK-SFOOD-C-AK-CPP-A-AK-LAB 0.063 1.304 0.082 13.3
AK-SFOOD-C-AK-SHOREP-A-AK-SALM-C-AK-SALM-A-AK-LAB 0.034 1.304 0.044 7.2
AK-SFOOD-C AK-CAP 0.376 AK-SFOOD-C-AK-SHOREP-A-AK-CAP 0.047 1.071 0.050 13.3
AK-SFOOD-C-AK-CPP-A-AK-CAP 0.019 1.071 0.021 5.5
AK-SFOOD-C-AK-SHOREP-A-AK-SALM-C-AK-SALM-A-AK-CAP 0.018 1.071 0.019 5.0
AK-SFOOD-C-AK-CPP-A-AK-PLCK-C-AK-PLCK-A-AK-CAP 0.015 1.071 0.016 4.3
AK-SFOOD-C-AK-SHOREP-A-AK-Halibut-C-AK-Halibut-A-AK-CAP 0.014 1.071 0.015 4.1
AK-SFOOD-C-AK-SHOREP-A-AK-PLCK-C-AK-PLCK-A-AK-CAP 0.014 1.071 0.015 3.9
AK-SFOOD-C-AK-SHOREP-A-AK-Crab-C-AK-Crab-A-AK-CAP 0.012 1.071 0.012 3.3
AK-SFOOD-C AK-IBT 0.051 AK-SFOOD-C-AK-SHOREP-A-AK-RETAIL-C-AK-RETAIL-A-AK-IBT 0.002 1.130 0.003 5.5
AK-SFOOD-C-AK-CPP-A-AK-RETAIL-C-AK-RETAIL-A-AK-IBT 0.002 1.130 0.002 3.8
AK-SFOOD-C-AK-SHOREP-A-AK-IBT 0.002 1.067 0.002 3.2
AK-SFOOD-C RU-LAB 0.534 AK-SFOOD-C-AK-SHOREP-A-AK-LAB-RU-LAB 0.016 2.326 0.038 7.0
AK-SFOOD-C-AK-CPP-A-AK-LAB-RU-LAB 0.008 2.326 0.018 3.4

Note: the acronyms for industries used in this table end with “-A” and the acronyms for commodities end with “-C”. IBT¼ indirect business tax; LOW-HH¼ low-income households; MED-HH¼ medium-income households; HI-
HH¼ high-income households; SLG ¼state and local government.
Table 5
Results for institutional income.

Origin Destination Global influence Elementary paths Direct influence Path multiplier Total influence Total/Global (%)

AK-PLCK-C AK-LOW-HH 0.028 AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-LOW-HH 0.010 1.309 0.013 45.4

AK-PLCK-C-AK-PLCK-A-AK-CAP-AK-LOW-HH 0.004 1.087 0.005 17.2
AK-PLCK-C AK-MED-HH 0.240 AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-MED-HH 0.084 1.344 0.112 46.9
AK-PLCK-C-AK-PLCK-A-AK-CAP-AK-MED-HH 0.038 1.201 0.046 19.2
AK-PLCK-C AK-HI-HH 0.311 AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH 0.112 1.345 0.150 48.3
AK-PLCK-C-AK-PLCK-A-AK-CAP-AK-HI-HH 0.051 1.238 0.063 20.4
AK-PLCK-C AK-SLG 0.085 AK-PLCK-C-AK-PLCK-A-AK-RETAIL-C-AK-RETAIL-A-AK-IBT-AK-SLG 0.005 1.825 0.009 10.5
AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-AK-FIRERL-A-AK-IBT-AK-SLG 0.003 2.082 0.005 6.4
AK-PLCK-C-AK-PLCK-A-AK-REFN-C-AK-REFN-A-AK-IBT-AK-SLG 0.002 2.039 0.005 5.5
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-HI-HH-AK-SLG 0.002 2.222 0.004 4.7

C.K. Seung / Marine Policy 66 (2016) 39–49

AK-PLCK-C-AK-PLCK-A-AK-WHOLE-C-AK-WHOLE-A-AK-IBT-AK-SLG 0.002 1.765 0.004 4.5
AK-PLCK-C-AK-PLCK-A-AK-LAB-AK-SLG 0.002 2.164 0.004 4.2
AK-PLCK-C RU-LOW-HH 0.034 AK-PLCK-C-AK-PLCK-A-AK-LAB-RU-LAB-RU-LOW-HH 0.001 2.366 0.003 8.7
AK-PLCK-C RU-MED-HH 0.206 AK-PLCK-C-AK-PLCK-A-AK-LAB-RU-LAB-RU-MED-HH 0.012 2.419 0.028 13.6
AK-PLCK-C-AK-PLCK-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A-RU-LAB-RU-MED-HH 0.002 3.134 0.007 3.6
AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A-RU-LAB-RU-MED-HH 0.002 2.978 0.007 3.3
AK-PLCK-C RU-HI-HH 0.329 AK-PLCK-C-AK-PLCK-A-AK-LAB-RU-LAB-RU-HI-HH 0.016 2.536 0.040 12.1
AK-PLCK-C-AK-PLCK-A-AK-CAP-RU-CAP-RU-HI-HH 0.008 1.743 0.014 4.1
AK-PLCK-C-AK-PLCK-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A-RU-LAB-RU-HI-HH 0.003 3.258 0.010 3.2
AK-PLCK-C RU-SLG 0.117 AK-PLCK-C-AK-PLCK-A-AK-WHOLE-C-RU-WHOLE-C-RU-WHOLE-A-RU-IBT-RU-SLG 0.002 2.241 0.005 4.4
AK-PLCK-C-AK-PLCK-A-AK-FIRERL-C-RU-FIRERL-C-RU-FIRERL-A-RU-IBT-RU-SLG 0.001 3.499 0.005 4.0
AK-SFOOD-C AK-LOW-HH 0.030 AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-LOW-HH 0.005 1.310 0.006 21.4
AK-SFOOD-C-AK-CPP-A-AK-LAB-AK-LOW-HH 0.002 1.310 0.003 10.2
AK-SFOOD-C-AK-SHOREP-A-AK-SALM-C-AK-SALM-A-AK-LAB-AK-LOW-HH 0.001 1.310 0.002 5.5
AK-SFOOD-C AK-MED-HH 0.254 AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-MED-HH 0.042 1.344 0.056 22.1
AK-SFOOD-C-AK-CPP-A-AK-LAB-AK-MED-HH 0.020 1.344 0.027 10.5
AK-SFOOD-C-AK-SHOREP-A-AK-SALM-C-AK-SALM-A-AK-LAB-AK-MED-HH 0.011 1.344 0.014 5.7
AK-SFOOD-C AK-HI-HH 0.330 AK-SFOOD-C-AK-SHOREP-A-AK-LAB-AK-HI-HH 0.056 1.346 0.075 22.8
AK-SFOOD-C-AK-CPP-A-AK-LAB-AK-HI-HH 0.027 1.346 0.036 10.8
AK-SFOOD-C-AK-SHOREP-A-AK-SALM-C-AK-SALM-A-AK-LAB-AK-HI-HH 0.014 1.346 0.019 5.8
AK-SFOOD-C-AK-SHOREP-A-AK-CAP-AK-HI-HH 0.008 1.239 0.010 3.1
AK-SFOOD-C AK-SLG 0.093 AK-SFOOD-C-AK-SHOREP-A-AK-RETAIL-C-AK-RETAIL-A-AK-IBT-AK-SLG 0.002 1.826 0.004 4.4
AK-SFOOD-C-AK-CPP-A-AK-RETAIL-C-AK-RETAIL-A-AK-IBT-AK-SLG 0.002 1.826 0.003 3.1
AK-SFOOD-C RU-LOW-HH 0.037 AK-SFOOD-C-AK-SHOREP-A-AK-LAB-RU-LAB-RU-LOW-HH 0.001 2.367 0.001 3.9
AK-SFOOD-C RU-MED-HH 0.230 AK-SFOOD-C-AK-SHOREP-A-AK-LAB-RU-LAB-RU-MED-HH 0.006 2.419 0.014 6.1
AK-SFOOD-C RU-HI-HH 0.364 AK-SFOOD-C-AK-SHOREP-A-AK-LAB-RU-LAB-RU-HI-HH 0.008 2.536 0.020 5.5

Note: the acronyms for industries used in this table end with “-A” and the acronyms for commodities end with “-C”. IBT¼ indirect business tax; LOW-HH¼ low-income households; MED-HH¼ medium-income households; HI-
HH¼ high-income households; SLG ¼state and local government.

47
48 C.K. Seung / Marine Policy 66 (2016) 39–49
shows that the increased Alaska labor income (AK-LAB) from in-
creased catch of pollock in Alaska leads to an increase in income of
Alaska high-income households (AK-HI-HH), which in turn, leads
to an increase in demand for RUS FIRERL commodity. This results
in an increase in output of RUS FIRERL industry.
An example of (ii) above is represented by the path, AK-PLCK-C-
AK-PLCK-A-AK-LAB-RU-LAB-RU-HI-HH-RU-MISCEL-C-RU-
MISCEL-A. This path shows that some of the labor income from AK-
PLCK-A is distributed to Alaska labor provider (AK-LAB) and the re-
maining labor income to RUS labor provider (RU-LAB), representing
non-Alaskan resident (RUS) labor income. The increase in RUS labor
income causes the income of RUS high-income households (RU-HI-
HH) to increase. This in turn induces an increased consumption by the
RUS high-income households of RUS miscellaneous commodity (RU-
MISCEL-C) and RUS miscellaneous industry output.
A heavy dependence of the Alaska pollock industry on imports
of RUS commodities is shown by the large global influences in
Table 2. The global influences are 0.323, 0.239, and 0.363, re-
spectively, for RUS other manufacturing industry (RU-OMAN-A),
RUS FIRERL industry (RU-FIRERL-A), and RUS service industry (RU-
SERV-A). These numbers are much larger than most of the global
influence numbers for the destination industries in Alaska. For
example, the global influence number for RU-SERV-A (0.363) is
much larger than the number for AK-SERV-A (0.209).
There are many elementary paths that include household sectors
as intermediate poles between the origin and destination sectors
(Table 2). In case of the Alaska service industry as a destination sector,
for example, four of the six elementary paths reported in the table
have a household sector as an intermediate pole. This indicates that
the household sectors (especially medium- and high-income house-
hold sectors) play an important role in transmitting the economic
impacts. Low-income households do not show up in the table as an
intermediate pole in Table 2, which implies that the household sector
does not play a critical role in transmitting the impacts.
Although not reported in any table, the Alaska output multi-
plier for Alaska pollock commodity, which measures the increase
in the output of all Alaska industries combined due to a $1 in-
crease in demand for Alaska raw pollock, is calculated to be 1.882.
Similarly, the RUS output multiplier for Alaska pollock commodity,
which measures the increase in the total output of all RUS in-
dustries from a $1 increase in demand for Alaska raw pollock, is
calculated to be 1.444.9 Conventional economic impact analysis
would provide the fishery managers only the multiplier numbers
such as these two numbers (1.882 and 1.444) and an element in
the IRSAM multiplier matrix (e.g., 0.983 in the second row and the
second column in Table 2). As such, it would not explain how and
through what paths the initial shocks are transmitted in generat-
ing the total economic impacts. The information on these paths is
provided only by an SPA.
An increase in Alaska processed seafood (AK-SFOOD-C) has the
largest global influence on two fish harvesting industries (AK-
PLCK-A and AK-SALM-A), reflecting the fact that much of raw fish
used in the processing is from the two harvesting industries (Ta-
ble 3). As in the case above where the origin sector is Alaska
pollock commodity, the results in Table 3 reveal that, in order to
process seafood in Alaska (AK-SFOOD-C), the three Alaska
9 able stock assessment for the species or the impacts of an exo-
The Alaska output multiplier (1.882) is supposed to be the same as the sum of
all global influence numbers for the Alaska destination industries. However, the
sum of the global influence numbers in the second column in Table 2 is only 1.637
( ¼0.983 þ 0.091 þ0.125þ0.124þ 0.209 þ0.105). Similarly, the RUS output multi-
plier (1.444) is supposed to be the same as the sum of all global influence numbers
for the RUS destination industries. However, the table shows that the sum is only
1.044 ( ¼0.323 þ0.239 þ 0.363 þ 0.119). These discrepancies occur because the table
reports only the results for a maximum of six arcs and a maximum value of 3% for
the TG ratios.
processing industries import a large amount of non-seafood
commodities from RUS. The global influences are 0.361, 0.265, and
0.409, respectively, for the three RUS destination sectors, RUS
other manufacturing industry (RU-OMAN-A), RUS FIRERL industry
(RU-FIRERL-A), and RUS service industry (RU-SERV-A). These
numbers are much larger than any global influence numbers for
the non-seafood destination industries in Alaska. As in the case
where the origin sector is Alaska pollock commodity, the role of
low-income households is not critical in transmitting the impacts
of the shock to seafood processing as evidenced by that sector not
appearing in any elementary paths in Table 3. Another notable
result is that capital income does not play a noticeable role in
generating the impacts of seafood processing (Table 3).
Table 4 clearly indicates the important role played by non-sea-
food industries in generating regional value added. For example,
when the origin sector is the Alaska pollock commodity, the sum of
the total influences on labor income (AK-LAB; destination sector)
transpiring through three non-seafood industries (Alaska retail in-
dustry, Alaska refined oil industry, and Alaska oil and gas industry)
is about 6.5% (¼3.5%þ 3.0%) of the global influence. Similarly, the
sum of the total influences of the origin sector (AK-PLCK-C) on ca-
pital income (AK-CAP; destination sector) occurring through three
non-seafood industries (AK-FIRERL-A, AK-REFN-A, and AK-OILG-A)
makes up as large as 9.3% (¼5.1þ 4.2) of the global influence.
The global influence of the Alaska pollock commodity on RUS
value added is generated either through direct leakage of value ad-
ded from Alaska to RUS (e.g., AK-PLCK-C-AK-PLCK-A-AK-LAB-
RU-LAB) or through Alaska's imports of RUS commodities (e.g., AK-
PLCK-C-AK-PLCK-A-AK-OMAN-C-RU-OMAN-C-RU-OMAN-A-
RU-LAB). Strong inter-regional impacts of Alaska pollock commodity
on RUS value added are shown by the fact that the sum of the global
influences on RUS value added sectors (0.865¼0.476þ 0.329þ0.059)
is as much as 82% of the sum of the global influences on Alaska value
added (1.055¼0.533þ0.475þ0.047).
Strong impacts of the Alaska pollock commodity on RUS
household income are shown in Table 5. In fact, the global influ-
ences of Alaska pollock commodity on some RUS household sec-
tors are larger than the influences on Alaska households. For ex-
ample, the global influence of the Alaska pollock commodity on
RUS high-income households (0.329) is larger than its influence on
Alaska high-income households (0.311).
6. Conclusion
Previous studies (e.g., [20,19]) investigated the inter-regional
economic impacts of Alaska fisheries, calculating inter-regional
multipliers. However, these multipliers are a black box [23] if
provided without complementary information, and cannot iden-
tify the structural and behavioral mechanism that produces these
multipliers. Extending an earlier study [21], the present study
opens the black box by conducting an IRSPA within an IRSAM
framework, and provides information on how and through what
paths the economic impacts of an initial shock to a seafood sector
are transmitted, amplified, and spilled over to other regions.
Fishery managers need to know the economic impacts, occur-
ring in the fishery-dependent regions, of a potential change, for
example, in the total allowable catch for a species from unfavor-
genous reduction in harvest of the species from natural disasters.
In this case, if the fishery managers are given only conventional
multipliers for the species without some complementary in-
formation about the structural and behavioral mechanism, they
cannot place confidence on the multipliers that will be used in
their decision-making. They may hesitate to accept blindly the
multiplier numbers without such complementary information.
 C.K. Seung / Marine Policy 66 (2016) 39–49 49

The results from an IRSPA may not directly alter the fishery References
management decisions, but rather will help the fishery managers
make decisions with some confidence. [1] Alaska Department of Labor and Workforce Development, 2012. Nonresidents
Working in Alaska: 2010. 〈http://laborstats.alaska.gov/reshire/historical/NON
RES10.pdf〉.
[2] J. Defourny, E. Thorbecke, Structural Path Analysis and Multiplier Decom-
position Within a Social Accounting Matrix Framework, Econ. J. 94 (373)
OMB disclaimer (1984) 111–136.
[3] Fissel, B., Dalton, M., Felthoven, R., Garber-Yonts, B., Haynie, A., Himes-Cornell,
A., Kasperski, S., Lee, J., Lew, D., Pfeiffer, L., Seung, C., 2013. Stock Assessment
The findings and conclusions in the paper are those of the and Fishery Evaluation Report for the Groundfish Fisheries of the Gulf Of
author and do not necessarily represent the views of the National Alaska and Bering Sea/Aleutian Islands Area: Economic Status of the
Groundfish Fisheries Off Alaska, 2012. Available: 〈http:/www.afsc.noaa.gov/
Marine Fisheries Service, NOAA.
refm/docs/2013/economic.pdf〉.
[4] D. Holland, P. Wyeth, SAM multipliers: Their Decomposition, Interpretation,
and Relationship to Input–output Multipliers, Research Bulletin XB 1027,
Acknowledgment College of Agricultural and Home Economics Research Center, Washington
State University, 1993.
[5] H. Kasumba, T.M. Ruhiiga, Applications potential of structural path analysis in
The author would like to thank Jerry Leonard at Northwest urban agriculture: a literature review, J. Hum. Ecol. 45 (1) (2014) 77–88.
[6] B.B. King, What is a SAM?, in: G. Pyatt, J. Round (Eds.), Social Accounting
Fisheries Science Center for providing regional economic data for Matrices: a Basis for Planning, The World Bank, 1985, pp. 17–51.
Pacific Coast fisheries. [7] R. Lantner, Recherche sur I’Interpretation du Determinant d’une Matrice In-
put–Output, Revue d'Bconomie Politique 82 (2) (1972) 435–442.
[8] M. Lenzen, Environmentally important paths, linkages and key sectors in the
Australian economy, Struct. Change Econ. Dyn. 14 (2003) 1–34.
[9] Leonard, J., P. Watson, 2011. Description of the input–output model for Pacific
Appendix. Acronyms used in the tables Coast fisheries. U.S. Dept. of Commerce, NOAA Tech. Memo. NMFS-NWFSC-111,
Seattle, WA, p. 64.
[10] Li, S., Gao, Y., He, J., 2004. SAM-based Multiplier Analysis for China's Economy.
See Table A.1
In Proceedings of the XIIINFORUM World Conference, Marina di Ascea-Velia,
Italy, September pp. 5–11.
Table A.1 [11] Minnesota IMPLAN Group Inc., 2004. IMPLAN Pro User's Guide.
Acronyms for endogenous sectors in the IRSAM model. [12] National Marine Fisheries Service (NMFS), 2013. Fisheries of the United States,
2012. U.S. Department of Commerce, Office of Science and Technology, Fish-
Sector name Acronym eries Statistics and Economics Division, 1315 East-West Highway, Silver Spring,
MD, 20910-23282.
[13] J.C. Parra, Q. Wodon, How does growth affect labor income by gender? A
Industry
structural path analysis for Tanzania, in: J.S. Arbache, A. Kolev, E. Filipiak
Pacific whiting harvesting Pwhting-A
(Eds.), Gender Disparities in Africa's Labor Markets, The World Bank, Wa-
Atka mackerel harvesting AtkaM-A shington, DC, 2010, pp. 273–296.
Flatfish and rockfish harvesting Flat_Rock-A [14] J.C. Parra, Q. Wodon, SimSIP SAM: A Tool for the Analysis of Input–Output
Pacific cod harvesting PacCod-A Tables and Social Accounting Matrices, The World Bank, 2010.
Pollock harvesting PLCK-A [15] D. Roberts, The spatial diffusion of secondary impacts: rural-urban spillovers
Sablefish harvesting Sablefish-A in Grampian, Scotland, Land Econ. 76 (2000) 395–412.
Crab harvesting Crab-A [16] D. Roberts, The role of households in sustaining rural economies: a structural
Halibut harvesting Halibut-A path analysis, Eur. Rev. Agric. Econ. 32 (3) (2005) 393–420.
Salmon harvesting SALM-A [17] J. Round, Decomposing multipliers for economic systems involving regional
Herring and other finfish harvesting Herr_Othfin-A and world trade, Econ. J. 95 (378) (1985) 383–399.
Other fish harvesting Otherfish-A [18] C. Seung, E. Waters, Evaluating supply-side and demand-side shocks for
Catcher-processor (processing only) CPP-A fisheries: a Computable General Equilibrium (CGE) model for Alaska, Econ.
Syst. Res. 22 (1) (2010) 87–109.
Mothership MS-A
[19] C. Seung, Estimating effects of exogenous output changes: an application of
Shorebased processor SHOREP-A
Multi-regional Social Accounting Matrix (MRSAM) method to natural resource
Agriculture AGRI-A
management, Reg. Sci. Policy Pract. 6 (2) (2014) 177–193.
Oil and gas extraction OILG-A [20] C. Seung, Measuring spillover effects of shocks to Alaska economy: an inter-
Other mining OMINING-A regional Social Accounting Matrix (IRSAM) model approach, Econ. Syst. Res. 26
Utilities UTIL-A (2) (2014) 224–238.
Construction CONSTR-A [21] C. Seung, Untangling economic impacts for Alaska fisheries: a structural path
Other manufacturing OMAN-A analysis, Mar. Resour. Econ. 30 (3) (2014) 331–347.
Refined petroleum REFN-A [22] The Research Group, 2007. Estimating Economic Impacts of Alaska Fisheries
Wholesale trade WHOLE-A Using a Computable General Equilibrium Model – Data Acquisition and Re-
Transportation TRANS-A duction Task Documentation. Prepared for Alaska Fisheries Science Center,
Retail RETAIL-A Corvallis, OR.
Information INFO-A [23] Thorbecke, E., 2000. The Use of Social Accounting Matrices in Modeling. In
Proceedings of the 26th General Conference of the International Association
Finance, insurance, real estate, renting, and leasing (FIRERL) FIRERL-A
for Research in Income and Wealth, Cracow, Poland, August 27–September 2.
Services SERV-A
[24] J. Xie, An environmentally extended social accounting matrix, Environ. Resour.
Miscellaneous MISCEL-A
Econ. 16 (2000) 391–406.
Value added
Labor income LAB
Capital income CAP
Indirect business tax IBT
Institutions
Low-income households LOW-HH
Medium-income households MED-HH
High-income households HI-HH
State and local government SLG