PGCA ss Part B

SUPPORTING STATEMENT
REGIONAL ECONOMIC DATA COLLECTION PROGRAM
FOR GULF COAST ALASKA
OMB CONTROL NO.: 0648-xxxx

B. COLLECTIONS OF INFORMATION EMPLOYING STATISTICAL METHODS
1. Describe (including a numerical estimate) the potential respondent universe and any
sampling or other respondent selection method to be used. Data on the number of entities
(e.g. establishments, State and local governmental units, households, or persons) in the
universe and the corresponding sample are to be provided in tabular form. The tabulation
must also include expected response rates for the collection as a whole. If the collection has
been conducted before, provide the actual response rate achieved.
For the vessel surveys, the overall population consists of all fishing vessels landing raw fish at a
port in Gulf Coast Region, Alaska during 2005. For that year, there were 1,846 vessels. This
population consists of three vessel classes – small, medium, and large vessel classes. The
population sizes are 831, 964, and 51, respectively for small, medium, and large vessels classes.
An unequal probability sampling (UPS) procedure is used to determine the sample sizes needed
for the analysis for small and medium vessel classes, which is described in Item #2 below and in
the aforementioned supplementary document. Since the population size of the large vessel class
is very small (51), we will mail out the surveys to all large vessel owners. 1 The population sizes
of local businesses and fish processors are 102 and 44, respectively.
The expected response rates for the vessel surveys are based on consideration of the following
factors. First, compared with a previous data collection project conducted for Southeast Alaska
(Hartman 2002), which achieved an overall response rate of about 30%, the number of questions
in the present project is much smaller and the quantity of information being asked is much
smaller. Second, in the present study, questions about sensitive information such as vessel cost
and expenditures are omitted. The previous Southeast study included these sensitive questions,
which significantly contributed to the low response rate. Third, input from select members of the
respondent populations helped guide survey design and question wording. Fourth, follow-up
telephone calls will also increase the response rate. Based on these factors, it is expected that,
overall, the response rate for mail survey of fishermen for the present project will be about 55%
which is much higher than in the Southeast study. For telephone interviews with local
businesses (including fish processors), a response rate of 65% is expected. For a more detailed
description of the methods we used, and will use, to increase the response rate, see Item #3
below.
1

There is population overlap between vessels landing fish in the Gulf coast region (current project) and Southwest
region. According to Commercial Fisheries Entry Commission data, there are a total of 295 vessels that land at both
Gulf Coast region and Southwest region. Out of these 295 vessels, 72 are in the small vessel class, 170 are in the
medium vessel class, and 48 are in the large vessel class. The remainder does not report vessel length. Thus, nearly
all of the large vessel class landing at Gulf Coast ports (48 vessels out of a total of 51 large vessels) also land at
Southwest ports. The questions in the surveys for Gulf Coast and Southwest data collection projects are the same
except for the region name in the questions asking about the residency of crew, skipper, and owners. Therefore, if
any of the 295 vessels are selected via Pareto sampling into the mailout samples for Gulf Coast project, we will send
the selected vessel owners only those questions which ask about the residency of the fishermen, excluding all the
other questions, thus reducing public burden. We will also acknowledge this sample overlap in our cover letter to
the vessel owners who land at both Gulf Coast and Southwest regions.

Vessel Class

Population
size

Small vessel
Medium vessel
Large vessel
Local businesses including
fish processors

831
964
51
146

Mail or
phone
interview
sample size
359
327
51
146

Expected
number of
respondents

Expected
response
rate

197
180
28
95

55%
55%
55%
65%

2. Describe the procedures for the collection, including: the statistical methodology for
stratification and sample selection; the estimation procedure; the degree of accuracy
needed for the purpose described in the justification; any unusual problems requiring
specialized sampling procedures; and any use of periodic (less frequent than annual) data
collection cycles to reduce burden.
Since the majority of gross revenue within each harvesting sector comes from a small number of
vessels, a simple random sampling (SRS) of vessels would only include a small portion of the
total ex-vessel value, and therefore, would be misleading. As a result, for the present project an
unequal probability sampling (UPS) method without replacement is used that accounts for this
unequal harvest in each target population. The objective of implementing the sampling task is to
estimate the employment and labor income information for each of three disaggregated
harvesting sectors using as an auxiliary variable the ex-vessel revenues provided by Commercial
Fisheries Entry Commission (CFEC) earnings data. Since each sector will be used as a separate
economic sector in IMPLAN model, we face three separate problems for three different sectors
in sampling. For small and medium vessel sectors, we use a UPS without replacement method to
identify sampling units. Since the population size of the large vessel class is small (51), we will
mail out the surveys to all large vessel owners.
3. Describe the methods used to maximize response rates and to deal with non-response.
The accuracy and reliability of the information collected must be shown to be adequate for
the intended uses. For collections based on sampling, a special justification must be
provided if they will not yield "reliable" data that can be generalized to the universe
studied.
(a) Maximizing Response Rates
Previous applications of voluntary commercial fishing surveys in Alaska (e.g., Hartman 2002)
were hampered by low response rates that principally resulted from the use of long and
complicated survey instruments. Commercial fishermen are frequently asked, and often
required, to participate in surveys from numerous organizations including NOAA, Alaska
Department of Fish and Game (ADF&G), and universities. As a result, commercial fishermen
are less likely to complete voluntary surveys that are lengthy, poorly-designed, and do not clearly
involve issues that are important to them. In this data collection, significant efforts were made to
ensure the survey instruments were short in length, contained well-designed questions, and
clearly conveyed the importance of the data collection to issues that are important to commercial
fishermen.

The mail surveys are short (6 to 7 questions depending upon the survey version, all of which
span eight pages) and avoid many sensitive questions compared with many previously-fielded
commercial fishing surveys. The set of questions was limited to only those that are essential for
achieving the objectives of the project as outlined in Part A, Item #1 above. There is only a
fraction of the number of questions asked compared with the Southeast Alaska commercial
fishing survey discussed earlier, which achieved an overall response rate of about 30%. In the
mail surveys, numerous questions on vessel expenditures that are often included in surveys of
commercial fishermen are omitted here to avoid the added complexity and likely sensitivity of
asking for this type of information from respondents. 2
The telephone scripts for use in interviews with local businesses and fish processors were
developed with similar goals in mind. Specifically, each phone script was constructed to include
only the most essential questions to ensure the telephone interviews were short in length to
minimize the time burden on respondents.
Pretesting activities that included a small focus group and several interviews with fishermen and
fish processors (totaling less than 10 individuals) were used to evaluate the content and
presentation of the survey materials, as well as to ensure input by the fishing community.
Feedback from these pretesting activities aided in non-trivial ways to the development of the
survey questions. For instance, considerable effort was made to ensure that the survey
instrument reflected considerations for the record-keeping systems kept by fishermen and used
common terms and wording used by fishermen. Participants in pretesting activities also
indicated that previous voluntary surveys often did not provide adequate assurances that the
information being requested would be handled confidentially, which often deterred them from
responding. To ensure respondents that the data they share will be kept confidential, a detailed
confidentiality statement is presented on the first page of the mail survey and mentioned upfront
in the telephone interviews. A similar statement is made in the cover letter accompanying the
mail survey.
Another reason believed to have caused low response rates in previous survey efforts is the
disinterest among respondents toward the survey purpose. Surveys that collect information that
will clearly benefit or interest respondents are more likely to be completed. The importance and
benefits of this data collection project to the respondents (fishermen, local businesses, and fish
processors) will be emphasized in the advance letter, cover letter, mail survey, and telephone
interviews. In these letters and phone interviews, the investigators clearly state that with the help
of the respondents, the important role of the respondents’ fishing and business activities in the
regional economy can be better identified and that the information they provide will be used to
enhance the fishery management practices of NMFS, and, thereby, to increase the long-run
economic benefits to the fishermen and local businesses. Making a clear link between the
survey, their participation, and the fishery and regional economy is expected to help increase the
response rate relative to previous studies.
In addition to the above steps taken to maximize response rates, the survey instruments (mail and
telephone) were subjected to significant review by several researchers with expertise on Alaska
fisheries and economic surveys to ensure the quality of the materials.
In addition to high-quality survey instruments, the set of survey protocols to be followed in
2

Vessel expenditures will be estimated using (1) the sales data collected from telephone interviews with local
businesses and fish processors and (2) a cost engineering approach.

implementation was designed to maximize response rates. For the mail survey, a modified
Dillman (2000) approach will be employed that includes four survey contacts as follows:
•
•
•
•

An advance letter notifying the respondents a few days before they receive the survey
questionnaire. This will be the first contact with the respondent.
An initial mailing sent a few days after the advance letter. Each mailing will contain
a cover letter, personalized questionnaire, and a pre-addressed stamped return
envelope.
A postcard follow-up reminder mailed 5-7 days following the initial mailing.
A follow-up phone call to encourage response and identify individuals that have
misplaced or need another copy of the survey. If the respondent agrees, the mail
survey will be completed over the phone. 3 Up to three attempts will be made to
contact each respondent for the telephone interview. Individuals needing an additional
copy of the survey will be sent one with another cover letter and return envelope.

A strict Dillman approach is not warranted given negative input from commercial fishermen
about repeated contacts beyond the phone contact.
The result of the efforts described above are compact and high-quality survey instruments that
contain questions vessel owners, local businesses, and fish processors can answer with
minimal effort. As a result, the expected response rate for the mail survey of fishermen is
expected to exceed previous survey efforts and achieve a response rate of approximately 55%.
This response rate is much higher than that in the longer and more complicated Southeast
Alaska study (30% response rate). For the telephone interviews with local businesses
(including fish processors), a response rate of 65% is assumed based on previous experience. 4
(b) Non-response
To better understand the differences between them, comparisons will be drawn between
respondents and non-respondents with respect to several observable characteristics: (1)
geographical area of landed fish, (2) ex-vessel value, and (3) species that vessels catch. This
information is available from government data for each vessel. If significant and systematic
differences between the two groups are discovered, the population parameter estimates of
interest may be adjusted by using weights formed from these variables.
4. Describe any tests of procedures or methods to be undertaken. Tests are encouraged as
effective means to refine collections, but if ten or more test respondents are involved OMB
must give prior approval.
There are no plans to conduct a pilot survey or other tests involving more than ten respondents.

3

In this case, the ex-vessel values (by species) of the vessel will be provided to the vessel owners so that they will
not have to access their records, which should greatly simplify the question and allow them to calculate the crew and
skipper payments easily. In doing this, we will make sure that the person we will be interviewing on the phone is
the true owner of the vessel. This is because we do not want to breach the confidentiality by providing the sensitive
information to the wrong person. As is seen in the mail survey questions, however, this ex-vessel information will
not be given to the respondent in the mail survey.
4
See Section A #12, Footnote 6.

5. Provide the name and telephone number of individuals consulted on the statistical
aspects of the design, and the name of the agency unit, contractor(s), grantee(s), or other
person(s) who will actually collect and/or analyze the information for the agency.
John Slanta (Census Bureau) and Dr. Dan Lew (NMFS) assisted in the development and review
of sampling procedures for this project. Their phone numbers are (301) 763-4773 and (206) 5264252, respectively.
Several NMFS economists with experience in economic survey design and implementation
reviewed the survey materials and survey protocols, including Dr. Dan Lew, Dr. Ron Felthoven,
and Dr. Brian Garber-Yonts.
Professor Hans Geier (University of Alaska, Fairbanks) is the contractor who will conduct the
data collection project, revise the IMPLAN data, and participate in developing regional
economic models.
Dr. Chang Seung (Alaska Fisheries Science Center) will conduct the statistical analysis of the
information collected, and develop regional economic models with Professor Geier.