A Stated Preference Choice Approach to Understanding Angler Preferences for Management Options

Understanding Anglers’ Preferences for Fishing TournamentCharacteristics and Policies

Chi-Ok Oh Æ Robert B. Ditton Æ Robin Riechers

Received: 9 January 2006 / Accepted: 12 December 2006

� Springer Science+Business Media, LLC 2007

Abstract Saltwater fishing tournaments in the United

States are generally not regulated nor are there different

fishing regulations for tournament and nontournament an-

glers. Although much is known about those who participate

in fishing tournaments in terms of their fishing motivations,

attitudes, and characteristics, much less is known at the

angler population level regarding their preferences for

tournament opportunities. Using a stated preference choice

model with hypothetical scenarios to simulate participation

choices and understand preferences, study objectives were

to identify angler preferences for various types of tourna-

ment fishing ‘‘products.’’ Four tournament policy charac-

teristics were investigated: promotion of catch and release,

bait restrictions, whether a percentage of the tournament

entrance fee should go to support fishery management

activities, and whether a tournament should be a nonprofit

or profit-making venture. Three expectation attributes were

inserted: tournament size, trip cost per day, and whether a

tournament is family friendly. We sent seven different

versions of the mail questionnaire to 1,633 anglers. Of 795

returns, 648 were used for estimating conditional logit

models. Analysis indicated that a scenario with no man-

agement interventions was not most preferred. Anglers

most preferred a conservation-oriented option that intro-

duced additional management measures. Overall, scenarios

with management interventions were more favored than the

status quo situation (with no management interventions).

Although respondents showed reluctance to adopt other

management-related options, results generally indicated

they were increasingly concerned with sustainability of fish

stocks and potential conflicts between tournament and

nontournament users and preferred tournament products

that reflect these concerns.

Keywords Management preferences � Recreationalfishing � Stated preference choice model � Recreationalfishing � Tournament fishing

The number of marine recreational fishing tournaments has

increased dramatically in recent years. In Texas, as

Christian and Trimm (1986) reported, nearly one half of

the 56 saltwater tournament events documented in 1983

were 1 to 3 years old. A recent 2003 inventory by the Texas

Parks and Wildlife Department (TPWD) revealed 183

tournaments (or a 227% increase since 1983) (personal

communication, J. Leitz, 2006). Although comprehensive

data were not available nationwide on the current number

and growth of saltwater fishing tournaments, the trend is

likely much the same as in Texas.

Fishing tournaments have been and continue to be a

controversial use of saltwater fishery resources for several

reasons (Schmied 1994; Williams 1984). First, only a small

minority (14%) of saltwater anglers participate ‡1 d/y in

competitive fishing events (Anderson & Ditton 2003).

However, these anglers constitute a managerially signifi-

cant group because of their numbers (Falk and others

1989), presumed dedication to fishing (Ditton & Loomis

Chi-Ok Oh (&)

Department of Parks, Recreation, and Tourism Management,

Clemson University, Clemson, SC 29634-0735, USA

e-mail: [email protected]

R. B. Ditton

Department of Wildlife and Fisheries Sciences, Texas A & M

University, College Station, TX 77843-2258, USA

R. Riechers

Coastal Fisheries Division, Texas Parks and Wildlife

Department, 4200 Smith School Road, Austin, TX 78744, USA

123

Environ Manage (2007) 40:123–133

DOI 10.1007/s00267-006-0010-7

1985), ability to catch fish, and likelihood of greater

involvement in fishery management activities than non-

tournament anglers. In contrast to tournament participants,

most anglers do not view recreational fishing in competi-

tive terms and hence do not share the values held by

tournament anglers (Loomis & Ditton 1987). Second, some

tournament events are run as profit-making businesses that

provide financial returns to fee-paying participants but

make use of public fishery resources at no cost to event

organizers. All anglers in Texas (as in many other states)

must be licensed to fish in saltwater and hence contribute to

fishery management costs. However, tournaments are not

generally licensed or charged fees above and beyond those

borne by the individuals who purchase licenses. In states

without a saltwater fishing license, neither anglers nor

tournament events are required to help cover the costs of

fishery management. Third, because the number of tour-

naments is likely increasing, there has been pressure on

state agencies to establish permit systems for purposes of

knowing the temporal and spatial distributions of tourna-

ment events in state waters in support of their fishery

management responsibilities. Fourth, although the local

and regional economic impacts of saltwater fishing tour-

naments are well documented (Ditton and others 1999a,

1999b; Thailing and others 2001), there is increasing

concern for possible social and biological costs of tourna-

ment events. This is difficult for fisheries management

agencies because they, too, are in the ‘‘recreation and

tourism business’’ and seek to generate state and regional

economic impacts while maintaining resource quality. In

this regard, they also have responsibilities for providing

new fishing opportunities on a fair and equitable basis

(Loomis & Ditton 1993). Accordingly, fishery managers

want to know to what extent various tournament charac-

teristics and policies are preferred by tournament anglers as

well as by those who are not tournament participants but

who could participate under the right circumstances.

Along with the goals of protecting and managing fishery

resources on a sustainable basis, some management agen-

cies also seek to provide anglers with a diversity of fishing

‘‘products’’ to help increase overall angler satisfaction.

Although the rate of tournament participation (percent of

anglers who participate in tournaments), fishing motiva-

tions, attitudes, and sociodemographic characteristics are

known (Antia and others 2002; Falk and others 1989),

much less is known at the angler-population level in terms

of understanding their overall preferences for tournament

opportunities currently being provided or that could be

provided. Such a study, if undertaken, would likely use a

traditional opinion measurement research design whereby

anglers would be asked to indicate their level of support

for or opposition to various tournament characteristics

and policies in use or contemplated. Such an approach,

however, would likely yield little insight to the relative

importance of each tournament policy option. Anglers

would likely respond to all of the various options in a

socially acceptable way with no insight to the tradeoffs

they may be willing to make when considering tournament

options jointly.

Instead, a stated preference choice model (SPCM)

makes use of hypothetical scenarios to simulate participa-

tion choices and better understand preferences. This tech-

nique assumes that complex decisions involve several

factors being considered simultaneously. Based on this

rational assumption, SPCMs allow for an understanding of

the relationship of multiple factors as they contribute to

preferences or choice behavior (Louviere and others 2000;

Louviere & Timmermans 1990). Whereas SPCMs have

been used previously to understand consumer preferences

for a variety of new multiattribute recreational fishing

products and services, they have not been used to under-

stand consumer preferences for various aspects of saltwater

fishing tournaments.

Two previous studies have used an SPCM to have an-

glers indicate their preferred trips making use of various

characteristics. First, Aas and others (2000) used an SPCM

to study anglers’ responses to various trip profiles (three

harvest variables and two expectation variables to explore

group differences). They found major choice differences

between angler groups based on gear use (fly-only, general,

and nonfly anglers). Although the status quo condition was

more favorable to nonfly anglers, fly-only anglers were

more likely to support regulation changes than other

groups. The preference differences in harvest regulations

suggested that conflict and dissatisfaction could be re-

solved by spatial segregation of the angler segments (i.e.,

use of zoning). Second, Oh and others (2005) investigated

the choices that red drum (Sciaenops ocellatus) anglers

made between hypothetical fishing trips as defined by four

harvest regulations and three fishing expectation variables

(including trip costs) and ranked the various management

scenarios being considered. Not surprisingly, the status quo

management scenario (with the most conservative harvest

restrictions) was least preferred by anglers and the most

liberal harvest restriction scenario was most preferred.

When anglers’ fishing expectations were taken into

account, however, anglers likely wanted more strict man-

agement that closely resembled the status quo management

option. They were willing to relax certain management

regulations but overall were guided more by conservation

than exploitation.

Although most states have left the matter of tournament

formats to private sector providers, some management

agencies, i.e., TPWD, want to encourage more fishery-

conservation practices in fishing tournaments and event

characteristics that are preferred by a broader cross-section

124 Environ Manage (2007) 40:123–133

123

of anglers (including current participants and those not now

participating in tournament events). Management officials

wanted to know the extent to which various event char-

acteristics and policies were preferred by the angling

community. In particular, they wanted to know which

tournament policies and characteristics were most and least

preferred before any policy decision-making efforts were

undertaken. Overall, they sought to enhance the positive

aspects of tournaments, decrease the negative aspects of

tournament events, and make them attractive overall to

more anglers. The purpose of this article is to (1) under-

stand the underlying rationale for anglers making trade-offs

in multiattribute tournament trip profiles associated with

various event characteristics and policies and (2) identify

the differences in concerns and preference for management

of tournament fishing based on previous tournament par-

ticipation (i.e., tournament anglers and nontournament

anglers). In doing so, we sought to provide managers with a

means for making pragmatic decisions (Nielsen 1985) that

maximize angler satisfaction consistent with their more

traditional fishery management responsibilities.

Methods

The methods section includes some background on the

selection of model attributes and levels, SPCMs in general,

and survey design considerations. The basic steps for using

an SPCM include identifying the important management

attributes; allocating appropriate levels to each attribute;

generating choice sets with identified attributes and levels;

presenting scenarios and acquiring responses from target

samples; and analyzing the preferences and management

program with an appropriate model. In the following sec-

tions, we provide insight and details for each step in

implementing a stated preference discrete choice model.

Identification of Attributes and Levels

The first step in developing an SPCM requires the selection

of tournament angler preference attributes and respective

levels. Relevant attributes and levels were identified from

open-ended comments provided by nontournament anglers

in previous statewide angler surveys (e.g., Anderson &

Ditton 2003) as well as from previous studies of tourna-

ment anglers and their fishing behavior, preferences, and

expenditures (e.g., Ditton & Loomis 1985; Loomis &

Ditton 1987). We also met with fisheries managers to better

understand angler preferences in light of current or future

management regulations. Four tournament selection char-

acteristics were identified as relevant policy attributes to

anglers: (1) promotion of catch and release, (2) bait

restriction, (3) whether a percentage of the tournament

entrance fee should go to support fishery management

activities, and (4) whether a tournament is a nonprofit or a

profit-making venture. A detailed description of the levels

of each attribute is listed in Table 1. Two or three levels

were assigned to each attribute to describe the particular

policy options involved.

The insertion of general expectation attributes was

beneficial in that attributes beyond agency control (such as

trip cost, including tournament fees and other trip ex-

penses, and attributes related to tournament size and

structure) were likely important to anglers’ decision mak-

ing regarding their participation in fishing tournaments

(Aas and others 2000; Fedler 1998; Gillis & Ditton 2002;

Oh and others 2005). Consequently, issues such as tour-

nament size, trip cost per day, and whether or not a tour-

nament is family friendly were integrated into the study

design. Revisions of levels and attributes were made as a

result of feedback from a series of pretests with anglers in

local fishing clubs. The goal of pretesting was to determine

whether the attributes used were significantly meaningful

to respondents as well as whether the range of levels for

each attribute was sufficiently substantial to reflect all of

the policy options considered (Bennett & Adamowicz

2001)

Model

The choice experiment model (or SPCM) was originally

developed in transportation choice research (Hensher 1994;

Louviere, 1988a) and later extended and further refined

in marketing and environmental studies (Ben-Akiva &

Lerman 1985; Louviere 1988b; Louviere and others 2000).

The SPCM is derived from well-grounded random utility

theory, which indicates that individuals make choices

to maximize utility (i.e., random utility maximization)

(Louviere 2000, 2001; Louviere and others 2000).

Based on obtained preferences, utility can be estimated

using the indirect utility function, which is comprised of a

deterministic component and a random error component

(Louviere 1988b; Louviere and others 2000). A determin-

istic component can be estimated to represent the vector of

coefficients of levels and attributes to obtain the part-worth

utilities for attributes. The indirect utility function of a

representative angler on a choice of fishing trip j can be

represented as

[Uj ¼ VjðAÞ þ ej ¼ lA0b þ ej; ð1Þ

where Uj is the utility of an alternative fishing trip j, Vj is

the deterministic component of utility to be estimated, and

ej is the unobservable error component of utility; l, a

scale parameter, is normally assumed to be 1. Because it

is rationally assumed that individuals always seek to

Environ Manage (2007) 40:123–133 125

123

maximize their utility, anglers will prefer an alternative of

tournament fishing trip i over j when Ui > Uj.

Because the indirect utility function is represented as

Equation 1, the above-mentioned equation is expressed as

ViðAÞ þ ei[VjðAÞ þ ej orViðAÞ � VjðAÞ[ej � ei ð2Þ

Because utility can not be observed directly, the probability

of choice results should be used, and the probability of

choosing alternative i over j is

Pðiji 2 MÞ ¼ PðViðAÞ � VjðAÞ[ej � ejÞ; ð3Þ

where M is all choice sets considered in the study.

Assuming the error terms of (ej–ei) are independently and

identically distributed and Gumbel distributed, the proba-

bility specification can result in the conditional logit (CL)

model with the following equation (Ben-Akiva & Lerman

1985; McFadden 1974)

Pðiji 2 MÞ ¼ exp ðViÞP

j¼M

exp ðVjÞ; ð4Þ

where M is all choice sets considered in the study. The

distributional assumptions for this model require the sat-

isfaction of the independence of the irrelevant alternatives

(IIA) property. The property of IIA states that ‘‘for a

specific individual, the ratio of the choice probabilities of

any two alternatives is entirely unaffected by the system-

atic utilities of any other alternatives’’ (Ben-Akiva &

Lerman 1985: p.108). Some of the reasons for violations of

IIA property involve the inclusion of close substitutes in

choice sets and heterogeneous preferences among respon-

dents (Morrison and others 1999).

Once the model has been estimated, willingness-to-pay

(WTP) values can be used to evaluate the effectiveness of

various proposals on the basis of changes in attributes that

reflect proposed scenario policies. WTP values can be

measured using the following equation:

1

btrip cos tðV0 � V1Þ; ð5Þ

where V0 indicates the utility acquired from the initial

condition of a fishing trip, and V1 is the utility from the

new scenario with altered levels of attributes (Hanemann

1984). Because the coefficient-of-trip cost is equivalent to

the marginal utility of income (Kaoru 1995), the coeffi-

cient-of-trip cost in this study was used as an alternative. In

addition, marginal values between a coefficient of a non-

marketed attribute (bi), and the coefficient-of-trip cost can

be calculated withbi

btrip cos t, leading to marginal WTP values

or implicit prices for an increase in a nonmarketed attribute

(Bennett & Adamowicz 2001; Hanley and others 2001). A

comparison of the implicit prices of attributes is important

in that there are further policy implications by examining

different components of alternative resource allocations

(Bennett & Adamowicz 2001).

Table 1 Proposed attributes and levels used for the choice experiments

Attribute Description Level

Selection Catch and release Enforcing catch and release restrictions for a

tournament

1. Catch and release behavior promoted

2. Catch and release behavior not promoted

Bait restriction Nature of bait allowed 1. No bait restriction

2. Artificial bait only

Tournament entrance

fee

Part of the angler entrance fee should go to TPWD to

support costs of fishery management

1. None of the tournament fee to go to TPWD

2. 10% of the tournament fee to go to TPWD

3. 20% of the tournament fee to go to TPWD

Tournament type Type of tournament held by different organizations 1. Tournament held as profit-making business

2. Tournament held by non-profit organization

General Trip cost/day Trip cost that an angler spends for a fishing trip per day

(including gas and other trip expenses)

1. $120 (approximately current travel cost per

day)

2. $150

3. $180

Family events The number of events provided for spouse and children 1. No family events

2. Some family events

3. Lots of family events

Tournament size The approximate number of participants in a

tournament

1. 100

2. 200

3. 300

126 Environ Manage (2007) 40:123–133

123

Survey Design

From the listing of 1.4 million resident anglers who pur-

chased a Texas fishing license in fiscal year 2001, a random

sample of 10,000 license holders was selected and an initial

mail survey conducted in spring 2002 (Anderson & Ditton

2003). Because most licensed saltwater anglers (86%) did

not participate in saltwater fishing tournaments, we con-

ducted a follow-up survey of those anglers who fished in

saltwater ‡1 day in the previous 12 months (N = 1633). In

summer 2004, a mail questionnaire using a choice exper-

imental design was sent to these anglers, and they were

asked about their tournament fishing trip preferences and

whether or not they had participated in fishing tournaments

during the previous year.

A fractional factorial design with main effects and two-

way interaction effects was used to generate a manageable

number of 56 choice sets. Because of a concern that

respondents always choose choice set i over choice set j, all

dominant choice sets were deliberately eliminated in the

design stage. Choice sets were then divided into seven

blocks of eight paired choice sets that were uncorrelated

(Bennett & Adamowicz 2001; Hanley and others 1998).

This was considered an effective way to decrease the

number of trip choice sets any one respondent might face.

Thus, seven different versions of the survey questionnaire

were used, each version having eight choice sets. Although

respondents may have a cognitive burden when required to

repeat a number of tasks (Bennett & Blamey 2001), eight

choice sets have been used previously and appears to be

adequate (Bates and others 2002). An example of a choice

profile is provided in Fig. 1. To simulate real market choice

behavior, each choice set included the option to not take

either trip (Bennett & Adamowicz 2001).

Results

Of the 1,633 saltwater anglers, 795 responded, yielding a

raw response rate of 48.7%. A modified Dillman Total

Design Survey Method (Dillman 1978) was used, with

three questionnaire mailings as necessary and a thank you/

reminder card sent after the first mailing. After deleting

undeliverable materials, the effective response rate was

53.0%. Although this response rate was lower than the

generally recommended level of 60%, Babbie (2001)

suggested that a 50% response rate is adequate for ana-

lytical purposes. The relatively lower response rate was

a priori expected given that most anglers had lacked tour-

nament participation, and thus tournament issues, in the

previous year and their management were less salient or

not salient at all.

Angler data collected in the 2001 statewide survey

(Anderson & Ditton 2003) were compared to check dif-

ferences between angler respondents and nonrespondents

in the 2004 tournament survey. Across sociodemographic

and general fishing behavior variables, various univariate

statistical tests used indicated that tournament survey

respondents were older, had higher household incomes, and

reported greater skill levels than nonrespondents. No sig-

nificant differences were detected between the two groups

for other important variables (e.g., total fishing days, total

cost spent for a fishing trip). Nevertheless, approximately

37% of the respondents indicated that they had participated

in saltwater fishing tournaments previously compared with

14% in the statewide angler population, who indicated that

had they participated in saltwater fishing tournaments

(Anderson & Ditton 2003). Accordingly, it is notable that

anglers with previous fishing tournament experience were

actually more likely to respond to this survey. Caution

should be exercised in generalizing study findings to the

population of saltwater anglers.

Of the 795 respondents, 147 were deleted because of

missing values in various social, economic, and fishing

behavior variables. Therefore, the final data set included

648 total responses, which yielded 5128 choice sets for

analysis after deleting choice sets with incomplete answers

from respondents.

Most (87%) respondents were male, and most (68%)

were ‡40 years of age or older (mean 48). The median

household income category of respondents was $60,000 to

$69,000, which was slightly higher than that reported in a

previous statewide angler survey (Anderson & Ditton

2003). Although most (71%) respondents indicated that

they were equally or more skilled, when they were asked to

compare their fishing ability with that of other anglers,

fewer than one third (29%) reported that they were less

skilled. When asked to compare fishing with their other

outdoor recreation activities, most anglers (71%) indicated

that fishing was their most or second most important

outdoor activity. Approximately 64% of the respon-

dents reported that they had fished in saltwater while a

tournament was in progress and that they had not been a

Suppose that you could only choose from the two tournament trips below. Which would you prefer?

TRIP A ATTRIBUTES TRIP B

Catch and release not promoted

CATCH & RELEASE Catch and release

promoted

Artificial bait only BAIT RESTRICTIONS Artificial bait only

10% of tournament fee to go to TPWD

ENTRANCE FEE None of tournament fee to

go to TPWD

Tournament held by profit-making business

TOURNAMENT TYPE Tournament held by profit-

making business

$150 TRIPCOST / DAY $150

No family events FAMILY EVENTS Some family events

100 TOURNAMENT SIZE 200

I prefer…(check one box below) TRIP A I WOULD NOT TAKE

EITHER TRIPTRIP B

Fig. 1 Example of a choice set for tournament fishing participation

Environ Manage (2007) 40:123–133 127

123

participant. When asked to indicate if tournaments nega-

tively affected the quality of their fishing, >74% of anglers

indicated no negative impacts.

CL Results

The parameter estimates of CL models are listed in

Table 2. To test the violation of the IIA property, the

likelihood ratio test between the CL (restricted) model and

the nested logit (unrestricted) model was conducted as well

as a Hausman-type test developed by Hausman and McF-

adden (1984). Although the latter test did not provide a

clear indication of results depending on which base choice

was excluded, the former test, in general, indicated a failure

to reject the null at p <.05 (Greene 2000). Consequently,

estimation results were made using the CL model and

random-parameter logit models, but only estimates of the

former with group segments are reported here because of

its methodologic simplicity. Although the random-param-

eter logit model, (which does not exhibit a concern for the

IIA property) takes into account the heterogeneous pref-

erences of angler clientele, a segmentation approach using

the CL model sufficiently reflects this concern for different

preference structures within the participant groups. This

approach also alleviates the IIA problem by relaxing the

assumption of common preferences (Oh & Ditton 2006).

The parameter estimates of the CL models are listed in

Table 2. An alternative-specific constant (ASC) was ad-

ded to represent the value of a tournament trip with

everything else held constant (Bennett & Adamowicz

2001; Boxall & MacNab 2000). The positive value for

ASC indicated that anglers were favorable toward tour-

nament participation in terms of current tournament pol-

icy characteristics (i.e., generally a laissez-faire policy

currently). Regardless of whether or not they had partic-

ipated in tournament fishing previously, it appears that

anglers were in favor of tournament fishing events and,

consequently, had some interest in tournament fishing

participation. Using the data, three different models were

estimated including a pooled model for all anglers and

two segmented group models. Effects codes were used for

the qualitative attributes of catch and release, bait

restriction, tournament type, and availability of family

events. For instance, the attribute of ‘‘having family

events at a tournament’’ with three levels was coded with

two effects-coded variables (i.e., event 1 for some family

events and event 2 for lots of family events) and the

impact of the attribute is represented by b1 and b1,respectively. Although the coefficient of the third level

was not directly estimated, it was calculated by the neg-

ative sum of the other two levels (i.e., �ðb1 þ b2Þ). Theuse of effects codes is advantageous in that the omitted

Table 2 CL model results

Model factors All Anglers Implicit prices Tournament anglers Implicit prices Non-tournament anglers Implicit prices

coefficient ($) coefficient ($) coefficient ($)

ASC catch and release 1.252a (0.162) 1.958a (0.266) 0.854a (0.205)

Promoted 0.269a (0.022) 33.4 0.243a (0.036) 26.6 0.286a (0.028) 38.0

Not promoted –0.269a (0.022) –33.4 –0.243a (0.036) –26.6 –0.286a (0.028) –38.0

Bait restriction

Artificial bait only –0.163a (0.023) –20.2 –0.171a (0.037) –18.7 –0.162a (0.030) –21.5

No bait restriction 0.163a (0.023) 20.2 0.171a (0.037) 18.7 0.162a (0.030) 21.5

Tournament fee 0.019a (0.003) 2.4 0.001 (0.004) 0.1# 0.031a (0.003) 4.1

Tournament type

Profit-making business –0.345a (0.028) –42.8 –0.337a (0.045) –36.8 –0.355a (0.036) –47.2

Non-profit organization 0.345a (0.028) 42.8 0.337a (0.045) 36.8 0.355a (0.036) 47.2

Trip Cost/ Day –0.008 (0.001) –0.009a (0.002) –0.007a (0.001)

Family Events

No events –0.385a –47.7 –0.364a –39.8 –0.404a –53.7

Some events 0.130a (0.030) 13.1 0.097a (0.049) 10.6 0.153a (0.039) 20.3

Lots of events 0.255a (0.033) 36.6 0.267a (0.053) 29.2 0.252a (0.042) 33.4

Tournament size –0.002a (0.000) –0.2 –0.002a (0.000) –0.2 –0.002a (0.000) –0.2

Log likelihood –5320.8 –1950.0 –3331.7

McFadden q2 0.0539 0.0546 0.0618

Note. The alternative specific constant was coded 1 for trip A and trip B in the choice sets and 0 for no tripa indicates statistical significant at the q = 0.05 level. SEs are in parentheses# indicates that the implicit value is not significant at the 0.05 level

128 Environ Manage (2007) 40:123–133

123

levels of attributes are uncorrelated with the intercept of

the model (Holmes & Adamowicz 2003; Louviere and

others 2000).

A goodness-of-fit measure of the pooled model (i.e.,

McFadden’s q2) was 0.054, which could be considered a

relatively low explanatory power. Although this measure is

analogous to the R2 in a conventional regression model

(Greene, 2000), there are no general guidelines for an

acceptable level of McFadden’s q2 (Ben-Akiva & Lerman

1985). All effects of the primary attributes were statisti-

cally significant (p < 0.05) (Table 2). Most attributes had

the expected signs except for promotion of catch-and-re-

lease restrictions (catch and release) as well as the one

designating part of the tournament entrance to the agency

to help defray saltwater fishery management costs (tour-

nament fee). Contrary to initial expectations, anglers pre-

ferred the option of having catch and release promoted in

tournaments compared with the option of no catch and

release in tournaments. Similarly, the positive coefficient

for tournament fee indicated that anglers were in favor of a

percentage of each tournament fee being available to the

agency for fishery management. In both cases, these

expectations were based on current practice among Texas

saltwater tournaments. Also, because most saltwater an-

glers (i.e., 63% in our sample or 86% in the previous state-

wide survey; Anderson & Ditton 2003) had not participated

in tournament events in the previous 12 months, these

unexpected signs may also reflect an increased interest in

or preference for the introduction of new types of fishing

tournaments. The options of artificial bait only (bait

restriction) and tournaments being held as profit-making

businesses (tournament type) were less preferred

(decreasing the choice probability for tournament partici-

pation) compared with the alternative of no bait restriction

and tournaments being held only by nonprofit organiza-

tions, respectively. Likewise, although the number of

family events held during tournaments (family events) was

likely to increase the choice probability for tournament

fishing participation, a strong preference was revealed for

having fewer participants in tournaments (tournament

size). Finally, the negative coefficient of travel cost (trip

cost per day) implies that anglers with higher expenditures

were less likely to participate in tournament fishing, which

coincides with consumer-demand theory.

However, the patterns of angler preferences for the

management options proposed differed between groups for

each specified model (i.e., tournament anglers and non-

tournament anglers). To test whether the parameters were

identical across the segmented models, a likelihood ratio

test of parameter equality (i.e., Chow test) was used. The

results (v2 = 78.25, p £ 0.01) indicated that the null

hypothesis of parameter indifference between two groups

was rejected. Thus, each segmented model is highly likely

to have a heterogeneous structure of underlying indirect

utility functions (Hanley and others 2006).

Furthermore, a series of Wald tests was performed for

the purpose of assessing individual parameter equality

across the two groups. The results listed in Table 3 show

that equality restriction of the parameters in ASC and

tournament fee was rejected. Consequently, nontournament

anglers showed strong support for the idea that a higher

percentage of each tournament fee should be made avail-

able to the agency to manage fishery resources. Tourna-

ment anglers, however, were much less interested in this

management option as revealed by the nonsignificant

coefficient for tournament fee. In addition, ASC and catch

and release indicated that tournament anglers were more

enthusiastic about taking tournament fishing trips and less

supportive of the restrictions for enforcing catch and re-

lease for tournaments than nontournament anglers. How-

ever, parameter inequality of catch and release was not

statistically supported. In general, these results indicated

that tournament anglers were likely more interested in

maintaining the current laissez-faire management condi-

tions than nontournament anglers.

Assessing Management Options

The SPCM provides an understanding of whether anglers

will be better or worse off depending on changes in feasible

tournament options. First of all, estimates of implicit values

provide insights to the extent of angler preference for each

management option (Table 2). Using the marginal rate of

substitution between the attribute in question and trip cost

as indicated, but with all other attributes remaining the

same (i.e., ceteris paribus), anglers were willing to pay $67

in support of the option ‘‘catch-and-release restrictions

promoted’’ compared with the option of ‘‘catch-and-re-

lease restrictions not promoted’’ and $86 in support of

‘‘tournaments held only by nonprofit organizations’’

Table 3 Wald test for attribute equality

Attribute v2 Significance level

ASC 10.79a £0.01Catch and release 0.89 0.35

Bait restriction 0.04 0.85

Tournament Fee 30.83a £0.01Tournament type 0.10 0.75

Trip cost/ day 0.69 0.41

Family events

Some events 0.78 0.38

Lots of events 0.05 0.82

Tournament size 0.83 0.36

a Indicates statistical significance at the q = 0.05 level

Environ Manage (2007) 40:123–133 129

123

compared with the option of ‘‘tournaments held by profit-

making business.’’ Furthermore, ignoring the presence of

other effects, anglers were willing to pay approximately $2

to have 10 fewer tournament participants in their event

(Table 2). The same interpretation can be applied to the

estimates of implicit values for each segmentation group.

With all other attributes remaining the same (i.e., ceteris

paribus), tournament anglers were willing to pay $74 in

support of the option ‘‘tournament held only by nonprofit

organization’’ compared with the option of ‘‘tournament

held as profit-making business’’ and $53 in support of the

option ‘‘catch-and-release restrictions promoted’’ com-

pared with the option of ‘‘catch-and-release restrictions not

promoted.’’ However, nontournament anglers were willing

to pay $20 (i.e., implicit value of $94) and $23 (i.e., im-

plicit value of $76) more to acquire the same options of

tournament type and catch and release, respectively, than

their tournament angler counterparts (Table 2).

To fully investigate changes in the level of each attri-

bute, 30 potential management scenarios were evaluated

using the CL model. (A complete set is available upon

request from the first author.) Five management scenarios

for saltwater tournament fishing trips were extracted and

presented for illustrative purposes with predicted proba-

bilities and overall WTP values using Equations 4 and 5

(Table 4). The larger the predicted probabilities and WTP

values, the more likely anglers were to prefer that partic-

ular management regime. The proposed scenarios were

selected as follows: First, three major scenarios were se-

lected as being particularly relevant to decision making

based on discussions with fisheries managers. Scenario 1 in

Table 4 was the base option (status quo with no manage-

ment interventions) for saltwater fishing tournaments in

terms of policy selection factors; scenario 3 was considered

a conservation-oriented option that introduced some addi-

tional management measures; and scenario 5 in Table 4

was considered a conservation-plus option with the most

restrictive management measures possible. Two additional

scenarios were added to Table 4 to provide additional

management insights. The levels of expectation attributes

(i.e., trip cost per day and tournament size) were also

permitted to vary depending on the modification of the

management attributes. Because of the lack of information

about the current extent of family events, this attribute was

constrained with no changes (held constant) through all

scenarios. Also, it was expected that trip cost per day

would increase if a higher percentage of the tournament fee

went to the fishery management agency and, accordingly,

that fewer tournament anglers would participate. This

scenario consideration was useful in that it showed that

anglers’ fishing trips would be affected by their expecta-

tions of how policy changes may affect fishing quality (Oh

and others 2005).

Results for all anglers were somewhat surprising. Sce-

nario 1 (status quo, no management interventions), which

was a priori expected to be most preferred, was not most

preferred (a predicted probability of 9.7%). Instead, anglers

most preferred scenario 3 (the conservation-oriented op-

tion), with a predicted probability of 31.5% and a WTP of

$146. This scenario included some management interven-

tions, such as promotion of catch-and-release restrictions

and that 10% of each angler’s tournament fee should go to

TPWD to help support fishery management costs. Fur-

thermore, scenario 5 (the conservation-plus option) was

also highly preferred, with a predicted probability of 25.8%

and a WTP of $121 compared with scenario 1.

A different pattern of angler preferences for the pro-

posed management scenarios was shown for each seg-

mentation group based on the magnitude of WTP values

and predicted probabilities. Although each group was

likely to be similarly supportive of more conservation-

oriented scenarios (i.e., scenarios 3, 4, and 5), in general,

nontournament anglers were more supportive of the stricter

management regime with the introduction of diverse

management measures. For example, although tournament

anglers were willing to pay $53 more than their actual

fishing expenditures (i.e., a predicted probability of

21.1%), nontournament anglers were willing to pay $171

more (i.e., a predicated probability of 28.8%) for man-

agement measures under scenario 5. Likewise, although

Table 4 The predicted probabilities and WTP of selected scenarios with constraints on expectation attributes

Scenarios Catch and

release

Bait

restriction

Fee

(%)

Type Trip

Cost

No. of

events

Size All anglers Tournament

anglers

Non-tournament

anglers

Probability

(%)

WTP

($)

Probability

(%)

WTP

($)

Probability

(%)

WTP

($)

S.1 Not promoted No restriction 0 Business 120 Some events 200 9.7 0 13.0 0 8.0 0

S.2 Not promoted No restriction 10 Business 120 Some events 200 11.7 23.9 13.1 0.5 10.9 41.2

S.3 Promoted No restriction 10 Nonprofit 150 Some events 200 31.5 146.4 31.7 97.2 31.3 181.7

S.4 Promoted Artificial only 10 Nonprofit 180 Some events 100 21.3 97.5 21.1 52.5 21.1 129.4

S.5 Promoted Artificial only 20 Nonprofit 180 Some events 100 25.8 121.4 21.1 53.0 28.8 170.7

130 Environ Manage (2007) 40:123–133

123

nontournament anglers least preferred scenario 1 (i.e.,

current restrictive management conditions), with a pre-

dicted probability of 8.0%, tournament anglers were very

much in favor of this management regime, with a predicted

probability of 13.0%. Overall, it was noted that manage-

ment scenarios with varying degrees of management

intervention were generally more favored than the status

quo situation with no management interventions. However,

nontournament anglers were more supportive of stricter

management scenarios for regulating tournament angler

behaviors than tournament anglers, who were compara-

tively more supportive of the management scenarios

reflecting the current laissez-faire policy.

Discussion

In this article, we sought a comprehensive understanding of

anglers’ preferences (irrespective of previous involvement

in saltwater tournament fishing) for various management

attributes of tournament fishing trips and their willingness

to make trade-offs among them. Fortunately, we had a

sufficient mix of tournament anglers (n = 237) and non-

tournament anglers (n = 411) to reflect the diversity of

preferences for tournament fishing management for these

two groups. A major concern at the start of the project was

whether the mail survey would be sufficiently salient to the

latter group of anglers for them to take the time to respond.

A special effort was made in the letters sent to anglers with

the questionnaire to encourage them to participate. Anglers

in the sample were told the following: ‘‘Whether you have

participated recently, previously, or at all in tournaments,

we want to know your opinions and preferences for tour-

naments as they are held today or may be held in the future.

Everybody’s preferences are important on this subject.’’

Thus, tournaments were considered a topic of concern to

all saltwater anglers and one where all saltwater anglers’

views should be taken into account.

Preference results generally indicated a certain degree of

support for the introduction of management interventions,

such as catch-and-release restrictions and making a portion

of each angler’s tournament fee available to the manage-

ment agency, perhaps because of concerns for possible

tournament-induced negative impacts. Although respon-

dents showed reluctance to adopt other management-re-

lated options (e.g., bait restrictions and type of tournament

administration), results confirmed that anglers have con-

sidered issues dealing with the sustainability of fish stocks

and potential conflicts between tournament and nontour-

nament anglers (Aas and others 2000; Gillis & Ditton

2002; Jacob & Schreyer 1980). More restrictive manage-

ment-related options were included in scenarios 3 and 5

(Table 4), but the high predicted probabilities and WTP for

these scenarios showed anglers’ willingness to accept

stricter management interventions (Gillis & Ditton 2002;

Oh and others 2005). The support for the conservation

option (scenario 3) exceeded that for the status quo option

(scenario 1). This was also likely caused by having both

tournament and nontournament anglers in the analysis.

Because no special licenses or rules have been established

for tournament anglers, it was more appropriate to conduct

the analysis using a sample of the general saltwater angler

population because all anglers have a right to be heard on

this important issue.

Although various approaches exist for taking into ac-

count the heterogeneous preferences of angler clientele, a

segmentation approach using a CL model was used in this

study to uncover underlying segments. Random-parameter

logit modeling, which assumes that parameters are ran-

domly distributed in the population, could also have been

used to incorporate preference heterogeneity (Train 1998).

Although the random-parameter logit model was also

estimated, the segmentation approach was seen as advan-

tageous in support of management decision making and

was used here for two main reasons. First, the random-

parameter logit model is limited in explaining the under-

lying sources of heterogeneity (Boxall & Adamowicz

2002), which are commonly related to the characteristics of

individual anglers. Second, although identifying segments

is often criticized because of the discretionary nature of

researchers’ arbitrary decisions (Hunt and others 2005), the

two groups used here were rigidly set using a predeter-

mined segmentation variable (i.e., previous tournament

participation or not).

Thus, to avoid possible displacement of affected angler

types, a segmentation approach based on tournament

experience was considered advantageous for understanding

how changes in management options have a differential

influence on angler segments. Overall, the results indicated

that compared with nontournament anglers, tournament

anglers likely prefer to maintain the more current laissez-

faire policy regarding fishing tournaments. Under current

management rules and regulations in Texas, all saltwater

anglers are treated the same whether or not they participate

in tournaments. However, our study findings showed that

those anglers who currently participate in tournaments

more likely favor the status quo without many management

interventions, and the remainder favor the more conser-

vation-oriented alternatives.

Some study limitations should be kept in mind. First,

this is a hypothetical contingent model, and, accordingly, it

is not known whether people’s stated behavior will match

their actual or revealed behavior. Nevertheless, the model

is useful for providing informed hypotheses for testing and

application elsewhere and insights not otherwise possible

using traditional opinion measurement techniques. Second,

Environ Manage (2007) 40:123–133 131

123

there is a concern for strategic behavior on the part of some

anglers who wanted perhaps to perpetuate the status quo

and discourage any form of change. Other anglers may

have provided socially desirable responses and even if

positive changes are made in tournaments, may still choose

not to participate. Third, compared with the finding that

only 14% of licensed saltwater anglers participated in

tournaments (Anderson & Ditton 2003), a greater per-

centage of tournament anglers (37%) in this study sample

provided their opinions and preferences. Because of the

heterogeneous concerns and preferences for management

of tournament fishing, further analysis is necessary to avoid

developing policies biased to any one group. Fourth, the

focus of this study was on a sample of the population of

licensed anglers who fish in salt water where differences in

skill, knowledge, commitment, and behavior are masked

because of an aggregate analysis. Just as it would be useful

to understand group differences in preferences based on

previous tournament participation (i.e., tournament and

nontournament anglers), it would also be beneficial in the

future to test for differences based on anglers’ specializa-

tion level (e.g., Bryan 1977; Ditton & others 1992). This

would require a much more sophisticated study design with

inclusion of various behavioral and attitudinal questions

and a much larger sample size than we had available here.

Specialization is ‘‘a continuum of behavior from the gen-

eral to the particular reflected by equipment and skills used

in the sport and activity/setting experiences’’ (Bryan 1977,

p.175). In the future, this concept will hopefully provide a

useful means for comprehending and acting on the diver-

sity found within the population of saltwater anglers.

Finally, there is the matter of the extent to which

natural resources agencies should become involved in

matters beyond traditional natural resource protection

concerns. At a recent meeting of the Texas Parks and

Wildlife Commission (TPWD), a constituent accused the

agency of engaging in social engineering, or, in other

words, promulgating rules and regulations intended to

produce desired social outcomes or in this case, particular

recreational fishing experiences. We consider this a part

of their management responsibilities. Particularly in this

case, the mission of the TPWD is clear in its support for

traditional natural resources management and conserva-

tion but also in providing ‘‘outdoor recreation opportu-

nities for the use and enjoyment of present and future

generations’’ (excerpted from TPWD Mission Statement,

http://www.tpwd.state.tx.us/business/about/mission/)

TPWD Mission Statement). In addition to the usual

concerns for overfishing and habitat requirements, each

state’s fishery agency has an opportunity to be involved in

promulgating rules and regulations that create additional

fishing ‘‘products’’ (Driver 1985) or in developing rec-

reation opportunities for its citizens, while enhancing the

state’s tourism economy. In this case, this could involve

reconfiguring fishing tournaments to make them more

attractive to anglers as well as more conservation friendly.

Acknowledgments Funding support for this research, which came

from the TPWD Coastal Fisheries Division and the Texas Agricul-

tural Experiment Station, was much appreciated.

Literature Cited

Aas O, Haider W, Hunt L (2000) Angler responses to harvest

regulations in Engerdal, Norway: a conjoint based choice

modeling approach. North American Journal of Fisheries

Management 20:940–950

Anderson DK, Ditton RB (2003) Demographics, participation,

attitudes, and management preferences of Texas anglers, 2001

(TAMU-WFSC-HD-624). Texas A & M University, Department

of Wildlife and Fisheries Sciences, College Station, TX

Antia U, McConney P, Ditton RB (2002) The socio-economic

characteristics of tournament anglers in Barbados. Gulf and

Caribbean Fisheries Institute 53:357–366

Babbie E (2001) The practice of social research, ninth edition.

Wadsworth/Thomson Learning, Belmont, CA

Bates IJ, Carson RT, Day B, Hanemann M, Hanley N, Hett T, et al.

(2002) Economic valuation with stated preference techniques: a

manual. Edward Elgar, Northampton, MA

Ben-Akiva M, Lerman SR (1985) Discrete choice analysis: theory

and application to travel demand. MIT Press, Cambridge, MA

Bennett J, Adamowicz V (2001) Some fundamentals of environmen-

tal choice modeling. In Bennett J, Blamey R (eds.) The choice

modeling approach to environmental valuation. Edward Elgar,

Northampton, MA. pp 37–69

Bennett JW, Blamey R (2001) The strengths and weaknesses of

environmental choice modeling. In Bennett JW, Blamey R

(eds.), The choice modeling approach to environmental valua-

tion. Edward Elgar, Northampton, MA. pp 227–242

Boxall P, Adamocwicz WL (2002) Understanding heterogeneous

preferences in random utility models: a latent class approach.

Environmental and Resource Economics 23:421–446

Boxall PC, MacNab B (2000) Exploring the preferences of wildlife

recreationists for features of boreal forest management: a choice

experiment approach. Canadian Journal of Forest Research

30:1931–1941

Bryan H (1977) Leisure value systems and recreational specialization:

the case of trout fishermen. Journal of Leisure Research 9:174–

187

Christian RT, Trimm DL (1986) An inventory of Texas saltwater

fishing tournaments: spatial, temporal and participation patterns

in 1983 (Management Data Series No. 97). Texas Parks and

Wildlife Department Coastal Fisheries Branch, Austin, TX

Dillman DA (1978) Mail and telephone surveys: the total design

method. Wiley, New York, NY

Ditton RB, Anderson DK, Bohnsack BL, Sutton SG (1999a) Texas

International Fishing Tournament: participants’ characteristics,

participation in fishing, attitudes, expenditures, and economic

impacts. (TAMU-WFSC-HD- 621). Texas A & M University,

Human Dimensions of Fisheries, College Station, TX

Ditton RB, Anderson DK, Thigpen III JF, Bohnsack BL, Sutton

SG (1999b) Pirate’s Cove Big Game Tournaments: partici-

pants’ characteristics, participation in fishing, attitudes, expen-

ditures, and economic impacts. (TAMU-WFSC-HD-615).

Texas A & M University, Human Dimensions of Fisheries,

College Station, TX

132 Environ Manage (2007) 40:123–133

123

Ditton RB, Loomis DK (1985) 1983 Texas International Fishing

Tournament: an analysis of participants’ characteristics, attitudes

and expenditures. (TAMU-SG-85-202). Texas A & M University

Sea Grant Program, College Station, TX

Ditton RB, Loomis DK, Choi S (1992) Recreation specialization: re-

conceptualization from social world’s perspective. Journal of

Leisure Research 24:33–51

Driver BL (1985) Specifying what is produced by management of

wildlife by public agencies. Leisure Sciences 7:281–295

Falk JM, Graefe AR, Ditton RB (1989) Patterns of participation and

motivation among saltwater tournament anglers. Fisheries

14(4):10–16

Fedler AJ (1998) Applying human dimensions information to

recreational fisheries management in the Gulf and Caribbean.

Gulf and Caribbean Fisheries Institute 50:1075–1088

Gillis KS, Ditton RB (2002) A conjoint analysis of U.S. Atlantic

billfish fishery management alternatives. North American Jour-

nal of Fisheries Management 22:1218–1228

Greene WH (2000) Econometric analysis. Prentice Hall, Upper

Saddle River, N .J

Hanemann WM (1984) Welfare evaluations in contingent valuation

experiments with discrete responses. American Journal of

Agricultural Economics 66:332–341

Hanley N, Mourato S, Wright RE (2001) Choice modeling

approaches: a superior alternative for environmental valuation?

Journal of Economic Surveys 15(3):435–462

Hanley N, Wright R, Adamowicz V (1998) Using choice experiments

to value the environment. Environmental and Resource Eco-

nomics 11:413–428

Hanley N, Wright R, Alvarez-Farizo B (2006) Estimating the

economic value of improvements in river ecology using choice

experiments: an application to the water framework directive.

Journal of Environmental Management 78(2):183–193

Hausman J, McFadden D (1984) Specification tests for the multi-

nomial logit model. Econometrica 52:1219–1240

Hensher DA (1994) Stated preference analysis of travel choices: the

state of practice. Transportation 21:107–133

Holmes TP, Adamowicz V (2003) Attribute-based methods. In

Champ PA, Boyle KJ, Brown TC (eds.) A primer on nonmarket

valuation. Kluwer Academic, Boston, MA. pp 171–220

Hunt LM, Haider W, Bottan B (2005). Accounting for varying setting

preferences among moose hunters. Leisure Sciences 27:297–314

Jacob GR, R. Schreyer (1981) Conflict in outdoor recreation: a

theoretical perspective. Journal of Leisure Research 12:368–380

Kaoru Y (1995) Measuring marine recreation benefits of water quality

improvements by the nested random utility model. Resource and

Energy Economics 17:119–136

Loomis DK, Ditton RB (1987) Analysis of motive and participation

differences between saltwater sport and tournament fishermen.

North American Journal of Fisheries Management 7:482–487

Loomis DK, Ditton RB (1993) Distributive justice in fisheries

management. Fisheries 18:14–18

Louviere JJ (1988a) Conjoint analysis modeling of stated prefer-

ence: A review of theory, methods, recent developments and

external validity. Journal of Transport Economics and Policy

22: 93–119

Louviere JJ (1988b) Analyzing individual decision-making: metric

conjoint analysis. Sage University Series on Quantitative

Applications in the Social Sciences No. 67. Sage, Newbury

Park, CA

Louviere JJ (2000) Why stated preference discrete choice modeling is

NOT conjoint analysis (and what SPDCM is?). Memetrics White

Paper

Louviere JJ (2001) Choice experiments: An overview of concepts and

issues. In J.W. Bennett, R. Blamey (eds.) The choice modeling

approach to environmental valuation. Edward Elgar, Northamp-

ton, MA. pp 13–36

Louviere JJ, Hensher D, Swait J (2000) Stated choice methods:

analysis and application. Cambridge University Press, UK

Louviere JJ, Timmermans H (1990) Stated preference and choice

models applied to recreation research: a review. Leisure Sciences

12:9–32

McFadden D (1974) Conditional logit analysis of qualitative choice

behavior. In Zarembka P (eds.) Frontiers in econometrics.

Academic, New York, NY. pp 105–142

Morrison M, Bennett JW, Blamey R (1999) Valuing improved

wetland quality using choice modeling. Water Resources

Research 35(9):2805–2814

Nielsen LA (1985) Philosophies for managing competitive fishing.

Fisheries 10(3):5–7

Oh C, Ditton RB, Gentner B, Riechers R (2005) A stated discrete

choice approach to understanding angler preferences and trade-

offs for management options. Human Dimensions of Wildlife

10(3):173–186

Oh C, Ditton RB (2006) Using recreation specialization to understand

multi-attribute management preferences. Leisure Sciences

28:369–384

Schmied RL (1994) A history and overview of fishing tournaments

(Special Report No. 46). In Proceedings of the Workshop on

Saltwater Tournaments. Atlantic States Marine Fisheries Com-

mission, Washington, DC. pp 1–8

Thailing CE, Ditton RB, Anderson DK, Murray TJ, Kirkley JE, and J.

Lucy (2001) The 2000 Virginia Beach Red, White, and Blue

Fishing Tournament: participants’ characteristics, attitudes,

expenditures, and economic impacts (Virginia Marine Resource

Report No. 2001-9, VSG-01-88). College of William and Mary,

Virginia Institute of Marine Science. Gloucester Point, VA

Train KE (1998) Recreation demand models with taste differences

over people. Land Economics 74:230–239

Williams T (1984) Fishing to win. Audubon 86(3):82–95

Environ Manage (2007) 40:123–133 133

123