APPENDIX I

INTRODUCTION

A 1991 to 1994 fishing operation, as defined in this project, is a unique permit holder / vessel owner combination that recorded a landing(s) in the halibut or sablefish fishery over the time period. The project required the CFEC to examine 1991 to 1994 halibut and sablefish fishing operations and determine whether the operations received initial allocations of halibut or sablefish QS.

An operation is designated as a "leftout" if neither the permit holder nor the vessel owner received an initial QS allocation in a management area where the operation fished over the 1991 to 1994 time period.

The actual determination of "leftouts" was a complex process that involved merging several computerized datasets. Most merges were done by matching on various identifier variables from the datasets.

1. THE DATABASES

a. The CFEC People File

The CFEC has issued limited entry and interim-use permits since 1974, and vessel licenses since 1978. The CFEC maintains a database of information on the persons who have held the permits and licenses. The file is referred to as the CFEC People file.

The People file contains, among other things, the names, addresses, social security numbers (SSN), and dates of birth for permit holders and vessel owners. Under Alaska state law, only natural persons may hold fishing permits, but business entities may own or license vessels. For this reason, the SSN field on the CFEC People file can represent the SSN of a natural person, or it can be a special CFEC identifier number for a business entity.1 If the SSN is a special CFEC identifier, it is still a unique number and used exclusively for that particular entity from year to year.

b. CFEC Vessel License File

CFEC's vessel license file contains data on all the vessels licensed in Alaska since CFEC took over vessel licensing in 1978. Among other information, the file contains the ADFG vessel number, information on the vessel's attributes, and information on the name and address of the vessel owner. It also contains a unique identifier (SSN) for the vessel owner. This identifier can be used to link fish ticket catch records to the vessel owner.

Vessel owners on the file may be natural persons or other entities. As mentioned above, this means that the vessel owner unique identifier can be an SSN of a natural person or a CFEC identifier number for a business entity.

The vessel license file contains only a single identifier for the owner, even though the vessel may actually be jointly owned by a partnership or corporation. During the application process for the sablefish and halibut IFQ programs, RAM found that many vessels which CFEC had listed as being owned by a natural person were actually owned by a partnership, corporation, or other entity. CFEC has relied on the information provided by the person who licensed the vessel and as a result the vessel file may contain information that is not completely accurate.

c. CFEC Catch and Gross Earnings File

CFEC's catch files are enhanced versions of original data sets. Sablefish harvest data ultimately comes from fish tickets maintained by the Alaska Department of Fish and Game (ADF&G) and NMFS Weekly Production Reports (WPR). Halibut harvest data ultimately comes from fish tickets maintained by the International Pacific Halibut Commission (IPHC). CFEC adds information to these underlying files and puts all of the data into a common record layout.

CFEC adds to the catch record the name, address, the personal identifier, and other information on the interim-use permit holder. CFEC also adds a gross earnings estimate to each fish ticket item. For this study, CFEC also added the name, address, personal identifier, and other information on the vessel owner from the vessel license file.

Since mid-1990, NMFS has not required fish tickets for groundfish, provided the catch is not processed in waters controlled by the State of Alaska. Some sablefish vessels that fish in the EEZ have continued to fill out fish tickets, whereas others have not. As a result, there is no single source of data that shows the number of boats and permit holders participating in the sablefish fishery. Constructing a sablefish catch file required a merge of fish ticket and WPR records, with special care taken to avoid double-counting of catch records. This process is described in detail in sub-section d. below.

d. The Sablefish Catch File

There are large discrepancies between the sablefish catch that has been reported on fish tickets, and catch that has been reported on WPRs. There are several possibilities for the differences, which include catcher-processors that fill out WPRs, but don't fill out fish tickets, or fill out fish tickets only part of the time; or errors that have been made in filling out the WPRs and fish tickets; or keypunching data entry errors.

It was obvious that all the boats and all the harvest would not be included if one relied only on fish tickets, so WPR records were used to document the sablefish catch from catcher processors. In the case where catcher-processors had filled out fish tickets, the sablefish catch given on the WPR was substituted for the catch that had been documented on fish tickets. This was done to avoid double- counting of harvest for these vessels.

Fish tickets were used to document the harvest for all other boats besides catcher- processors.

WPR records do not contain information on the permit holder who was onboard the catcher-processor. When there were fish ticket catch records for catcher- processor, the permit holder was assumed to be the person listed on the fish tickets. Otherwise, if there were no fish tickets for a catcher-processor, the permit holder and the vessel owner were assumed to be the same person.

The table below lists the year, the total harvest pounds documented on WPRs and fish tickets, and the total harvest pounds resulting from the CFEC revised and combined sablefish catch data.

d. CFEC Census File

CFEC developed an extensive file that contains all of the known communities in Alaska. The file classifies the places as rural or urban based largely upon 1990 census information. In addition, the file indicates whether or not the place is "local" or "non-local" to various fishing areas. The Census file was used in this study to classify the resident type of vessel owners and permit holders. The rules for the rural/urban and local/non-local classifications are discussed in Appendix II.

e. The RAM Demographic File

RAM maintains a database of demographic information for persons involved in the restricted access halibut and sablefish fisheries. This file is referred to herein as the Demographic file. The version used for this report was current through January 3, 1996. Each observation in the Demographic file contains information on the name, most recent address, and the type of "person" the record represents (natural person, partnership, corporation, etc.). This file also has an identifier field called RAMSSN, which can be either the SSN of a natural person, or the employer identification number of a business entity, or a number generated by RAM, which occurs when persons refuse to provide their social security number. There is also a field called IFQ_ID, which is an identifier that links a "person" with all the types of QS they have received in the various species, area, vessel category, and quota share types combinations.

The IFQ_ID is derived from 3 sources, similar to the RAMSSN. The IFQ_ID can be either an encoded social security number, a Federal employer identification number (EIN), or a number that was assigned by RAM. RAM assigned IFQ_IDs when persons refused to provide their social security number.

The RAMSSN is usually the same number as the IFQ_ID (if the encoded IFQ_IDs are de-coded); however, sometimes the two variables are different. Sometimes the RAMSSN is blank, and sometimes it is a different number altogether.

e. The RAM Initial Allocation File

RAM created a database of initial QS allocations as of January 3, 1996. Each record on the database is identified with an IFQ_ID. There are also variables that show the amount of QS allocated and the species, area, vessel category, and QS type.

The Initial file also has the person name and RAMSSN variables from the Demographic file. As mentioned above, the RAMSSN field may contain either the SSN of a natural person, a Federal employer identification number (EIN), or a number generated by RAM. On some records, this field was blank. The name variables are the name of a contact person, and not necessarily the name of the business or person who actually received the QS.

2. MATCHING the RECORDS

The CFEC's first task was to attempt to match each record on the Initial file to the CFEC People file. After this match was completed, there would be a link between RAM's Initial allocation file and the CFEC records. The matching process was done by merging the IFQ_IDs and RAMSSNs on the Initial file to the unique identifier field (SSN) on the People file.

Completing this match to the CFEC People file involved several steps that are outlined below.

a) All encoded IFQ_IDs on the Initial file were de-coded. These de-coded IFQ_IDs provided an SSN that was merged to the CFEC People file. Most de- coded SSNs matched to the People file; however, some did not. Those that did not match were set aside for special analysis (see below).

b) All IFQ_IDs that represented a Federal employer identification number (EIN) were also merged to the CFEC People file. Some of these EINs matched to the People file; however, most did not. Those that did not match were set aside for special analysis (see below).

c) All observations that had IFQ_IDs that were generated by RAM were merged to the CFEC People file by the RAMSSN variable. Some of these RAMSSNs matched to the People file; however, most did not. Those that did not match were set aside for special analysis (see below).

d) Special analysis of non-matches:

Some records on the Initial file did not match to the CFEC People file when the match was attempted using the de-coded IFQ_ID (SSN), or Federal EIN, or RAMSSN. These records, (761 total records) were printed out and more information about the record was gathered which might help match it to the CFEC files.

The Initial file records contained the contact person name, their address, and their date of birth. It was possible the non-matching records could match to the People file by these fields if they didn't match by ID.2

Matching by name, address, and date of birth required a name-by-name search through the CFEC licensing and permit files. Sometimes, the non-matches occurred because of obvious keypunching errors. These records were fixed with computer code. However, most non-matches occurred when the IFQ_ID was a Federal EIN number of a partnership, corporation, or other business entity. Although the IFQ_ID could not match to the CFEC People file if it represented a Federal EIN, the contact person name for that record would often match, and therefore would represent a person who licensed a vessel or held a permit in Alaska.

Sometimes the contact person name on the Initial file appeared to be represented by two different persons on the CFEC licensing files; for example, a father and son who share the same name. In these cases, the address of the QS recipient was used, and sometimes their date of birth, to help make a correct match to CFEC records. When choices like this were made, the "quality" of the match was also recorded. The quality of the match varied, depending upon what records matched exactly to the CFEC files.3

After it was determined which person on the CFEC licensing files should be assigned the QS allocation, a variable was coded on the computer file that referenced how the determination had been made.

When the quality of the match was relatively weak, CFEC researched the NMFS RAM paper files to try to verify decisions. This was a time consuming process that nevertheless proved worthwhile as it usually provided accurate information and a correct assignment.

The authors also learned about the nature of partnerships from the NMFS paper files. Sometimes the partnerships and business entities were such that it would be more correct to assign the QS to more than one person on the CFEC files. When this occurred, records were added to the Initial file. This new and enhanced file was called the Initial2 file.

After all analysis and matching was completed, there were still 15 unique IFQ_IDs and 36 total records on the Initial2 file that had no available match to CFEC records. Because there was no match to CFEC records, it is not possible to know whether those QS recipients also fished during the 1991 to 1994 time period. These records were therefore excluded from the analysis in the Gap Report. The table below summarizes the non-matching records and the amount of QS they represent, by area. Note that some non-matching IFQ_IDs were issued both halibut and sablefish QS, and some IFQ_IDs can appear twice in the same species / area.

3. ASSIGNING INITIAL QS RECIPIENTS to the 1991 - 1994 CATCH RECORDS

As stated above, nearly all records on the Initial file eventually were linked to the CFEC People file, meaning that at some point in time the person or business entity who received an initial QS allocation also licensed a vessel in Alaska or held an Alaska commercial fisheries limited entry or interim-use permit.

The next task was to merge the records on the Initial2 file to the 1991-1994 catch records. If a fishing operation harvested sablefish or halibut between 1991 and 1994, but there was no link for that operation back the Initial2 file, then that operation's vessel owner and permit holder could be considered "leftouts."

An important consideration in this analysis is the definition of a "leftout." It is possible for a vessel owner or permit holder to receive initial QS, but not for the species that their operation harvested over the 1991-1994 time period. For example, a person may have been issued only halibut QS, but from 1991 to 1994 his fishing operation harvested both halibut and sablefish. It is also possible that a person received QS for the same species that was harvested over the 1991 to 1994 time period, but the QS was issued to him in a different area than where the 1991 to 1994 harvest occurred.

For this study, nine different nominal variables were added to the 1991 to 1994 catch records after the Initial2 file was merged to the catch file. These nominal variables took on a "yes" or "no" value depending upon where the QS was issued.

IPH: Permit holder received QS

IPHS: Permit holder received QS for the species harvested during 1991-1994

IPHSA: Permit holder received QS for the species and area where harvest occurred in 1991-1994

IVO: Vessel owner received QS

IVOS: Vessel owner received QS for the species harvested during 1991-1994

IVOSA: Vessel owner received QS for the species and area where harvest occurred in 1991-1994

IFD: Federal EIN received QS

IFDS: Federal EIN received QS for the species harvested during 1991-1994

IFDSA: Federal EIN received QS for the species and area where harvest occurred in 1991-1994

Note that there may be more that one permit holder / vessel owner combination in an area in a year. This report considers a fishing operation to be a unique permit holder / vessel owner combination that recorded a landing(s) during the 1991 to 1994 period.

In this report, the nominal variables were used to define a leftout operation. A leftout operation is one where neither the vessel owner nor the permit holder received an initial QS allocation in a management area(s) where the operation fished over the 1991 to 1994 period.