Consumer Preference for Fisheries Improvement Project: Case of Bigeye Tuna in Japan

Abstract

In recent years, demand for sustainable fisheries certification, also known as seafood ecolabeling, has grown worldwide, with retailers actively promoting ecolabeled seafood, mainly in Europe and the United States. However, the costs associated with assessment and maintenance are typically incurred before certification, and the potential benefits are uncertain, which deters many fisheries from entering the certification process. The Fishery Improvement Project (FIP) is a market-driven mechanism that allows a fishery to gain recognition for its sustainable management efforts aimed at achieving sustainable certification. Market differentiation of FIP-participating fisheries from conventional fisheries has the potential to generate additional benefits that may offset some of the certification costs. However, successful differentiation efforts require consumer awareness, willingness to pay a premium, and effective communication strategies. This study investigates consumer preferences for bigeye tuna sashimi in Japan using a discrete choice experiment to determine if Japanese consumers are willing to pay a price premium for FIP-participating fisheries. The analysis resulted in a significant price premium for FIP and domestic certification valued more than international brands. These findings suggest that FIP-participating fisheries have the potential for cost recovery, even during the certification process.

1. Introduction

In recent years, the demand for sustainable fisheries certification has increased globally. Sustainably certified fisheries apply ecolabels to their products and can benefit from certification when consumers choose their products in the market. Many studies have found that consumers in European countries and Japan actually prefer to choose ecolabeled seafood [1,2,3,4,5,6]. Some studies identify the actual benefit of the ecolabel in consumer markets [2,4,7,8,9] as well as in production areas [5,10].

The Marine Stewardship Council (MSC) issues the world’s leading fishery ecolabel scheme, and MSC-certified fisheries now produce more than 15% of the world’s wild catches [11]. However, for most fisheries, except successful large-scale fisheries, it is economically difficult to maintain certification under the present ecolabel scheme over the long term, and most fisheries rely on subsidies [12]. Additionally, a resource-oriented scheme without considering the feasibility of pursuing the scheme places a burden on the producers who should be the main actors [13].

FIP is not a certification scheme, but rather a voluntary, private governance-based project that provides fisheries considering acquiring a credible sustainable certification scheme with a plan to improve their sustainable fishing practices [14]. Not limited to the fisheries, a series of stakeholders, including retailers, distributors, non-profit organizations, processors, and seafood companies, play significant roles in FIP to communicate their efforts for sustainable practices to consumers, with the aim of generating added value in the marketplace [15]. Six fisheries in Japan have engaged in FIP, including longline fisheries targeting blue shark, swordfish, and albacore tuna; rope-down aquaculture oyster fisheries; purse seine fisheries targeting sea bass and Japanese sardine; and a barrel flowing fishery targeting octopus (fisheryprogress.org, accessed on 21 February 2024). Compared to obtaining sustainable certifications such as MSC, the cost of participating in FIP is much lower, making it a more accessible option for most fisheries. If consumers recognized and endorsed FIP as a value of seafood sustainability and it can add value in terms of price, it could encourage fisheries to consider obtaining ecolabel certification via FIP. However, there is limited literature available on FIP.

Some studies analyze the effectiveness of FIP, quantifying the extent of improvement and identifying improvement factors [14,16,17]. However, no consumer preference research has been conducted to understand whether there is a value perceived by consumers for FIP (i.e., a price premium for fish taken from FIP-participating fisheries). As such, it remains uncertain whether fisheries participating in FIP have the possibility to receive economic rewards. This study conducts a discrete choice experiment (DCE) to estimate consumer preference (willingness to pay) for seafood sourced from FIP-participating fisheries.

We target seafood in the Japanese seafood market. In Japan, the number of MSC-certified fisheries has yet to increase, probably because more than 90% of Japanese fisheries are small scale, and most of them are not likely to afford the expensive fees arising from MSC certification. For such small-scale fisheries, FIP is a good option to address sustainability [18].

We evaluate a sashimi product made from bigeye tuna in the DCE. Tuna species are not only the most common types of sushi consumed by Japanese consumers, but also Japan’s strong market power drives global fisheries to harvest tuna species and export them to Japan [19]. Bigeye tuna is one of the overharvested species among the tuna species. Seafood Watch^® has evaluated 27 bigeye tuna fisheries, out of which only two in the western central or southwest Pacific Ocean are rated as either “best choice” or “certified”, while 19 fisheries in the Indian, Pacific, and Atlantic Oceans are recommended to be “avoided” (Seafood Watch, https://www.seafoodwatch.org/recommendations/search?query=%3Afree%3Bbigeye%7Cspecies%3Aspecies%3BBigeye%20tuna, accessed on 27 April 2023). International science committees also provide statuses of bigeye tuna stocks across the oceans, and some of the bigeye species residing in the Indian Ocean need to be cautious about over-exploitation from fisheries [20]. Hence, the sustainability of bigeye tuna fisheries can be an important consideration for consumers when selecting products.

2. Materials and Methods

We employ a discrete choice experiment (DCE) using bigeye tuna products as the focus [21]. The product shown in Figure 1 is a 100-g package of medium-fatty bigeye tuna (10 slices) from the Seychelles area with Spain as the country of origin. Bigeye tuna is popular in Japan as it has a taste similar to bluefin tuna, which is considered a high-end tuna species. Moreover, it is cheaper and provides the second-largest supply after yellowfin tuna in Japan [22]. Bigeye tuna is also caught in the Pacific, Indian, and Atlantic Oceans as well as off the coast of Japan [23]. It is available both domestically and imported, or fresh and frozen, which makes it the best species for a DCE.

In the experiment, we set four attributes (price, production area, ecolabel, and product condition), with two to five levels for each attribute. We designed the DCE using “dcreate” in the STATA 16 package to optimize D-efficiency [24], which created eight questions consisting of four choices for products P, Q, R, and an opt-out option (shown in Figure 2). In a hypothetical setting without payment obligation, the participants are said to overestimate their own value for the commodities, which is called hypothetical bias and skews the true estimation of consumer preferences [25]. We gave a cheap talk which is known to alleviate this bias, when a respondent was about to enter the choice experiment [26]. We added the following cheap talk in Japanese to mitigate hypothetical bias: “In this survey, you are not obligated to actually pay money. In that case, it is said that people tend to over-evaluate the commodity. Thus, please imagine that you are actually in a supermarket buying sashimi”. The survey was conducted in January 2023, and targeted consumers aged 20 to 79, living in Japan. A web research company sampled the consumers along the national population grouped by age distribution and eight districts in Japan (Hokkaido, Tohoku, Kanto, Chubu, Kinki, Shikoku, Chugoku, and Kyushu/Okinawa).

The DCE is conducted via an Internet survey. We recruited 2900 seafood consumers in Japan who are registered users of a web survey company, taking into account Japan’s regional population size, gender, and age groups. In the survey, we first asked about the recognition of seafood ecolabels and FIP after providing brief and basic information to the participants and then asked eight questions of the DCE. We also asked several sociodemographic questions such as education level, family status, attitudes toward sustainable fisheries, expense of seafood, and personal characteristics to control for heterogeneous preferences [27].

2.1. Theoretical Model

In accordance with the literature, we analyzed data obtained using the random utility model framework [28]. Considering the bias caused by regional, economic, and cultural heterogeneous consumer preferences, we employed a mixed logit model, which would estimate unbiased parameters dealing with the heterogeneity of preference. The probability that individual i chooses alternative j from the choice set J in occurrence t can be expressed by the following indirect utility function.

$$U_{ijt}=V_{ijt}+{\epsilon }_{ijt}$$

(1)

V_ijt is the non-stochastic tastes of the population and is represented by a linear expression of the matrix as follows.

$$V_{ijt}={\mathit{\beta }}_i^{\prime }{\mathit{x}}_{ijt}$$

(2)

The parameter vector ${\mathit{\beta }}_{\mathit{t}}$ for individual i is a K × 1 vector, which follows a population distribution $g\left(\mathit{\beta }|\mathit{\theta }\right)$, where the vector $\mathit{\theta }$ is the population parameters of the distribution. The vector ${\mathit{x}}_{ijt}$ represents the content of the occasion t of the alternative j, while ${\epsilon }_{ijt}$ is an error term that follows a type I extreme value distribution.

According to Train (2003) [29], the probability of individual i choosing a choice profile among the choice sets can be explained as follows:

$$\mathrm{Pr}\left(s_i|x_i,{\mathit{\beta }}_{\mathit{i}}\right)=\prod _{t=1}^T\frac{e^{{\mathit{\beta }}_i^{\prime }{\mathit{x}}_{i,s_{it},t}}}{{\sum }_{j=1}^J{e}^{{\mathit{\beta }}_i^{\prime }{\mathit{X}}_{ijt}}}$$

(3)

where s_i denotes the choice sequence from s_i1 to s_iT, which is conditioned by the choice profile of x_i from x_i1 to x_iT. Integrating this for the distribution of $\mathit{\beta }$ to obtain the following equation,

$$\mathrm{Pr}\left(s_i|x_i,\mathit{\theta }\right)=\int \mathrm{Pr}\left(s_i|x_i,{\mathit{\beta }}_i\right)g\left(\mathit{\beta }|\mathit{\theta }\right)d\mathit{\beta }$$

(4)

The parameters are estimated by maximizing the following simulated log-likelihood (SLL) function, which is made computable by logarithmic transformation.

$$SLL=\sum _{i=1}^I\sum _{t=1}^{T}ln\left(\frac{1}{R}\sum _{r=1}^R\frac{e^{{\mathit{\beta }}_i^{\prime }{\mathit{x}}_{i, s_{it}, t}}}{{\sum }_{j=1}^J{e}^{{\mathit{\beta }}_i^{\prime }{\mathit{X}}_{ijt}}}\right)$$

(5)

$g\left(\mathit{\beta }|\mathit{\theta }\right)$ represents the heterogeneity of individual preferences as a distribution, and Train (2003) uses this term to account for heterogeneity to estimate true consumer preferences [29]. The “mixlogit” and “wtp” packages were used in STATA16 for the estimation of parameters with robust standard errors and marginal willingness to pay [30,31].

2.2. Empirical Specification

The variables for estimation are listed in

Table 1. Attributes and levels of the products.

Attributes	Levels
Unit price (in yen)	498, 598, 698, 798
Product origin	Foreign (base), Kagoshima, Kochi, Shizuoka, Miyagi
Ecolabel	No-label (base), MSC, MEL, FIP
Condition	Thawed (base), fresh

. The most common features of commercial sashimi products that consumers are able to recognize, except for appearance, are price, volume, place of production, and product condition (thawed or fresh). In addition to these, labels including ecolabels are applied if the products are certified by certain certification schemes. Given this, we defined four attributes (price, origin, ecolabel, and product condition) that constitute a product. Four levels (498, 598, 698, and 798 yen) for price (approximately 3.64, 4.37, 5.10, and 5.83 USD, respectively, using the exchange rate on 8 December 2022), five levels (foreign, Kagoshima, Kochi, Shizuoka, and Miyagi) for place of origin, four levels (unlabeled, MSC, MEL, and FIP) for ecolabel, and two levels (thawed and raw) for product condition were defined as the levels handled in the choice experiment.

The price levels (498, 598, 698, and 798 yen) adopted in this study closely resemble the settings from the previous study on Bluefin tuna price levels (445, 645, and 845 yen) [32], while bigeye tuna generally is cheaper than bluefin tuna. This is because we considered the fact that the consumer price index for tunas in Japan has experienced a 30% increase from 2021 to 2023 (Ministry of Internal Affairs and Communications, 2022 “Consumer Price Index”, https://www.stat.go.jp/data/cpi/, accessed on 27 April 2023).

The four prefectures with the largest production volume of bigeye tuna were selected as production areas, and foreign production was included as a basis for comparing domestic production and imports (Fisheries Agency, 2023, “2021 Fishery and Aquaculture Production Statistics”, https://www.maff.go.jp/j/tokei/kekka_gaiyou/gyogyou_seisan/gyogyou_yousyoku/r3/index.html, accessed on 27 April 2023). At the moment, the wild-caught ecolabeled seafood distributed in Japan is mostly occupied by MSC and MEL, so we selected MSC and MEL certification in addition to FIP as sustainable features. We defined two types of production conditions: frozen tuna (and thawed at the supermarket) or fresh tuna, which are important purchasing factors. The volume of each package is fixed at 100 g with 10 slices of sashimi (10 g per slice), which is an appropriate size and volume for one meal of sashimi consumption. Promotional labels such as “Fresh!” and “Best choice!” are often applied to many packages in Japan; however, we did not include these attributes in order to accurately measure the attributes of our focus. In addition, alternative specific constants (asc) were included for each of the products P, Q, and R to control the order effect of the alternatives. Since each individual has different preferences for the four attributes, we set all of the parameters including constants (asc) as random parameters.

The Conservation Alliance, which sponsors the FIP, does not recommend any specific logos or designs for the FIP because the FIP is positioned as a step toward MSC, and the labeling of FIP using a logo may undermine the value of MSC [33] Without using a logo, environmentally-conscious restaurants promote FIP products as an example, demonstrating the story of sustainable fisheries management in FIP-participating fisheries to attract environmentally conscious consumers (D&DEPARTMENT Project, www.d-department.com/item/DD_EVENT_34782.html, accessed on 8 May 2023). This raises the issue that we compare the different values of an ecolabel with a logo versus the value of the FIP without a logo. A previous study suggests that logos influence consumer preference [34]. Since the value of ecolabeled seafood is at least composed of the value of sustainability and a logo, the valuation of ecolabels with logos and the FIP without a logo is not comparable or the ecolabeled seafood is overvalued by the value of a logo. Therefore, in the choice experiment, we eliminated logos and expressed ecolabels and FIP using only letters so that sustainability values could be comparable. In order to enable consumers to make a judgment about ecolabels (and FIP), we provided participants with minimum information on ecolabels (and FIP) prior to the experiment. For the first question, we asked about awareness of the ecolabels (and FIP). We defined seafood ecolabels as “certifications by third-party organizations that verify the fishery- and aquaculture-industry’s efforts to sustainably use fishery resources and care for the environment. (Note that this does not guarantee quality or safety.)” For this question, we provided the following brief explanations with logos. The subsequent question asks about recognition of FIP with minimum information provided to participants as follows.

• MSC: The Marine Stewardship Council (MSC) has its headquarters in the United Kingdom and has been operating since 1997. It targets wild fisheries, distribution, and processing industries. In all, 496 fisheries and 5614 distribution/processing companies worldwide have obtained certification, including 12 fisheries and 313 distribution/processing firms in Japan (as of April 2022).
• MEL (Marine Eco-Label Japan): A Japanese seafood ecolabel launched in 2007, targeting the wild fisheries and aquaculture industries, 14 fisheries, 53 aquaculture firms, and 100 distribution/processing companies in Japan have acquired the label (as of April 2022).
• FIP: Fishery Improvement Project (FIP) is a cooperative effort among fishermen, companies, distributors, NGOs, and others who have not obtained a seafood ecolabel to improve the sustainability of fisheries, with the ultimate goal of obtaining an ecolabel such as MSC.

3. Results

A total of 2913 consumers responded to the survey. Of these, we excluded fraudulent responses from those who chose the left (Product P), middle (Product Q) or right (Product R) column for all questions, as well as those who made such choices that they would not purchase all of the products in the experiment. Then, we obtained 2875 valid responses.

3.1. Descriptive Statistics

Our samples of households generally reflect the median consumer in Japan (see

Table 2. Descriptive statistics of participants.

	Our Statistics			National Statistics
Variable	Mean	Std.Dev. 1 Dev.	NOBS 1	Mean 2
recognize MSC	0.10	0.30	2875
recognize MEL	0.10	0.30	2875
recognize ASC	0.07	0.26	2875
recognize FIP	0.06	0.24	2875
female	0.51	0.50	2875	0.51
age	50.76	16.04	2875	53
20 s	0.13	0.33	2875	0.14
30 s	0.15	0.36	2875	0.15
40 s	0.20	0.40	2875	0.19
50 s	0.18	0.38	2875	0.19
60 s	0.17	0.38	2875	0.16
70 s	0.18	0.38	2875	0.18
housemaker	0.17	0.38	2875
student	0.03	0.16	2875
public officer	0.05	0.21	2875
executive officer	0.01	0.12	2875
part time jobber	0.15	0.36	2875
household income	442.50	401.71	2274	552 (437)
individual income	281.18	286.99	2483	307.4
have (a) child(ren)	0.58	0.49	2875
graduate school	0.04	0.19	2875

¹ Std.Dev. and NOBS denote standard deviation and number of observations, respectively. The value in parenthesis shows the median of the statistics. ² The sampling is in line with Japanese national statistics. National statistics based on “2020 Population Estimates” by the Ministry of Internal Affairs and Communications for age and gender. “2019 Comprehensive Survey of Living Conditions” by Ministry of Health, Labour and Welfare (MHLW) for household income. “2020 Basic Survey on Wage Structure” by MHLW for individual income.

). The individual income of the respondents in our survey is lower than the national statistics, but the household income is not different from the national statistics, which means that the other member(s) of the household earns more than the respondents.

Of the total households, 17% are composed of housewives (and househusbands), and nearly 60% have children. Four percent of them have a graduate or higher education, and 10% indicate that they are aware of MSC and MEL, while the aquaculture ecolabel (ASC) is recognized by 7%, and FIP by 6%.

3.2. Discrete Choice Experiment

The results of the estimation show that all parameters were significant at a 1% level (

Table 3. Estimation Results of Mixed Logit Model.

Number of Observations = 92,000			Likelihood Ratio χ2(12) = 18,617.4
Log Likelihood = −15,858.079			Prob > Chi Square = 0.0000
Variables	Coefficients	Std. Err.	z-Value	p > z	95% Confidence Interval
Mean
price	−0.02	0.00	−47.15	0.000	−0.02	−0.02
asc1	9.49	0.28	34.22	0.000	8.95	10.04
asc2	9.84	0.27	36.77	0.000	9.32	10.37
asc3	9.44	0.25	37.82	0.000	8.95	9.93
Kagoshima	1.21	0.06	19.66	0.000	1.09	1.33
Shizuoka	1.31	0.06	23.9	0.000	1.21	1.42
Miyagi	1.14	0.06	18.78	0.000	1.02	1.26
Kochi	1.45	0.07	22.18	0.000	1.32	1.58
MSC	0.20	0.07	3.1	0.002	0.07	0.33
FIP	0.28	0.07	4.05	0.000	0.14	0.41
MEL	0.52	0.06	8.13	0.000	0.39	0.64
Fresh	1.07	0.06	18.25	0.000	0.96	1.19
S.D.
price	−0.01	0.00	−48.15	0.000	−0.01	−0.01
asc1	0.47	0.05	8.74	0.000	0.37	0.58
asc2	0.10	0.08	1.19	0.234	−0.06	0.27
asc3	−0.07	0.09	−0.81	0.420	−0.24	0.10
Kagoshima	−0.05	0.08	−0.66	0.512	−0.21	0.10
Shizuoka	0.04	0.11	0.38	0.705	−0.17	0.25
Miyagi	0.06	0.07	0.85	0.396	−0.07	0.18
Kochi	−0.13	0.07	−1.78	0.074	−0.27	0.01
MSC	0.02	0.05	0.33	0.742	−0.08	0.11
FIP	0.04	0.05	0.78	0.438	−0.06	0.14
MEL	−0.11	0.05	−2.39	0.017	−0.20	−0.02
Fresh	1.64	0.05	33.09	0.000	1.55	1.74

). All of the parameters except for price, were positive, which is consistent with the interpretation. Some of the standard deviations of the mixed logit estimation were found to be significant for price, Kochi (at the 10% level), MEL, and fresh, indicating that a high degree of variation (heterogeneity of preferences) exists among individuals for these attributes. We then calculated the marginal willingness to pay (MWTP) and confidence intervals based on the obtained estimates using Krinsky and Robb (1986), yielding Figure 3 [35]. The MWTP for each attribute is shown in the form of a box plot in Figure 3, all of which are statistically significantly different from zero, as they all do not pass through zero.

The MWTP for asc1, asc2, and asc3 were 596, 618, and 592, respectively, and when averaging these values, the sum of the base (reference) attributes of MWTP (nolabel, thawed, and foreign) for bigeye tuna was estimated to be 602 yen. In contrast, the domestic brands (Kagoshima, Shizuoka, Miyagi, and Kochi) were valued at 76, 83, 72, and 91 yen higher than the foreign product. The MSC and MEL ecolabels and FIP were evaluated at 13, 32, and 17 yen higher than those without labels. Compared to the thawed products, fresh products were evaluated 67 yen higher. Note that each MWTP in Figure 2 is different in value from each base attribute (nolabel, thawed, and foreign), and thus the value of fresh bigeye tuna with a MEL ecolabel produced in Kochi is, 602 (base) + 67 (fresh) + 32 (MEL) + 91 (Kochi) = 792 yen.

4. Discussion

The MWTP for the FIP is estimated at 17 yen (3% of the base price, 602 yen). This result suggests that consumers significantly prefer seafood under the FIP. It follows that fishermen may benefit from their sustainable efforts even before obtaining a sustainable certification by communicating to consumers from the time they participate in the FIP.

In this study, the MWTP for all sustainability features (ecolabels and FIP) is generally not highly evaluated. This is mainly because we did not provide an image of ecolabels in this choice experiment to make FIP comparable to other ecolabels. According to the fact that higher willingness-to-pay (WTP) is attributed to the power of logos [34], lower WTP for ecolabels should be plausible when we deny giving logos. While our results showed that the price premium was 2.2% for MSC and 5.3% for MEL, the price premium for bluefin tuna in the previous study was more than 30% [32]. This disparity presumably comes from the lack of visual value generated by the logo.

Among the lower MWTP for ecolabels, MSC is not highly valued in the consumer market in Japan. The chi-square test in the post-estimation showed that the probability of choosing MSC is not significantly different from that of choosing FIP at 5%, but significantly different from MEL. The previous studies comparing MSC and MEL for seafood in Japan showed that MSC has been rated higher than MEL [36]. Since consumers tend to prefer local to foreign products [37], it is a possibility that the same logic may apply to consumer’s preference for local and global certifications.

The significant variation in the standard deviation of the alternative-specific constants for asc1 (Commodity P) suggests that the mixed logit model captured the left-to-right bias of selecting the leftmost item [38]. In our design of the choice experiment, Commodities P in the leftmost column, through the choice experiment, was averaged 673 yen, which is the highest compared to the other alternative commodities, Q (middle, averaged 623 yen) and R (rightmost, averaged 648 yen). The share of actual choice made by participants was 32% for commodity P, 45% for commodity Q, and 22% for commodity R. It is understandable that the probability of being chosen is highest in the cheapest middle column (Commodity Q column). However, only the leftmost column asc’s standard deviation became significant presumably because some people tend to stick to choosing the leftmost alternative. This order effect between alternatives was eliminated by setting asc as a random parameter.

5. Conclusions

In this study, a discrete choice experiment was conducted with Japanese consumers to determine whether a price premium existed for ecolabels and FIPs for bigeye tuna sashimi products in Japan. This study is the first to demonstrate the existence of a significant positive price premium of 17 yen for FIP products, indicating that Japanese consumers are willing to pay for sustainable fisheries management similar to ecolabel-certified fisheries (13 yen for MSC, 32 yen for MEL). Although this evaluation was devalued due to a lack of logos, it is possible that effective marketing strategies, such as adding a story or mentioning the SDGs at retailers, could further increase the value added.

The study also found that domestic brands and fresh products generated a significant price premium for bigeye tuna sashimi products, while the sustainability feature was valued at less than half the value of quality-based factors. The results of this study indicate that the absence of logos played a significant role in the price premium for ecolabels compared to previous studies that evaluated ecolabels with logos.

However, it is important to note that while value-added products from FIP could provide economic incentives for fisheries to obtain sustainable certifications such as MSC, this does not necessarily contribute to the improvement of sustainable fisheries. Many of the fisheries participating in FIP are within their first two years, and there are many fisheries that have not been assessed or cannot be assessed due to lack of data [39]. Because FIP has lower barriers to entry, such as lower certification costs, some fisheries enter the market solely for the added value, not for MSC certification [40]. While most reports show that these fisheries have improved their management and sustainability performance, if a price premium is inadvertently attached, it could be exploited as an affordable profit tool, with low barriers to entry, and undermine legitimate sustainable certification. To address this concern, the Conservation Alliance urges FIP participants not to buy products with low improvement potential [15]. Therefore, it is necessary to ensure a transparent assessment system that informs consumers about the sustainability performance of fisheries and enables them to choose marine products from those working to improve their sustainability. Retailers also play an important role to intermediate the FIP fisheries and consumers. On behalf of consumers, the retailers probe FIPs and provide appropriate FIP fisheries to provide consumers with the seafood.