Hoedads first signed up with SAIF on July 1, 1981. Premiums paid to
SAIF were based on the amount of wages paid to workers at varying
rates dependent upon work type -- office and TSE being cheap a saw
work and treeplanting expensive.

As time went on, the accident frequency of the past three years would
be recorded and the Employer would receive an experience rating. This
rating would be used as multiplying factor of the premium. For
example, if you had a good overall accident record for the past three
years, you would multiply the premium owed times, say, .80, thus
reducing your bill. The reverse would be true if you had a bad record.
This rating would be updated yearly.

At the end of a policy year (after a six-month waiting period), if
the overall record of all the employers signed up with SAIF was good,
a dividend could be declared and employers could get a rebate on the
premium paid for that year (generally between 5% to 20%).

What happened back in 1983 (and caused all this fuss) is the State
Legislature took 81 million from a reserve fund in order to balance
the state budget. This action prevented SAIF from issuing rebates and
or reducing future premiums. Hoedads had paid approximately $85,000
in premiums in the `81-'82 policy year.

In the ensuing years, SAIF needed to replenish its reserves and did so
by raising its rates and/or by eliminating dividends (in what
proportion or duration I don't know). During this same time period,
claims were rising and rates skyrocketed.

Starting in the 1982-83 policy year, Hoedads engaged in a variation
of a SAIF policy called a Retro-Plan. This was a gambling type of
plan in that it stretched the limits of what could be gained lost
during a policy year based on accident experience (above and beyond
the experience rating).

Six months after the end of a Retro policy year, the year's activity
(premiums paid, claim expenses, and reserves for future costs of
existing claims) would be evaluated and either a rebate or a bill
would be issued. This would happen every year thereafter until that
policy year was "bought-out," meaning that all claims (accidents) that
occurred in the policy year would be closed by paying sums of money to
SAIF for any potential future costs of the claim. Sometimes it was
better to buy out a claim ASAP; sometimes not. It was a calculated

Each policy year stood in this own, which could cause an accumulation
of policy years that were not closed out. For example, the `82-'83
policy year was not closed until sometime in `85 or `86, but the
`83-'84 policy year was closed in early 1985. It mostly depended on
the amount and severity of claims in that year.

>>From 1985 till 1994 Hoedads had a variety of policy types. Some
were retro-type; some years were standard -- sometimes by choice and
sometimes not. There was a period of time -- I think `89 -- when
Hoedads was thrown into the assigned-risk pool. This is where they
put bad people and charged them a lot of money. You either drown or,
if you're lucky, you resurface with a lungful of water.

The potential relevance of all this is the accounting and policy
precedents set by Hoedads in deciding overlapping and time-delayed
SAIF rebates or bills, as well as a perspective on the way the
classmates of `94 view the current situation.

The way Hoedads dealt with time-delayed SAIF bills or rebates from
1983 through 1994 was based on the timing of the transaction. If the
transaction occurred before the books were closed out for th previous
year, say in February or early March, the transaction could be
incorporated into the prior year's or current year's books, thereby
affecting the profit or loss for that respective year.

If the transaction happened too late in the year, then it would be
incorporated into the year the money was received or paid out
regardless of what year was involved. For example, if Hoedads
received a rebate (dividend) from SAIF in June of 1985 for the 1983-84
policy year. Hoedads would credit the SAIF account for 1986, which
would in turn reduce the SAIF bill for 1986, which would in turn
ncrease patronage dividends and retained earnings for 1986. The
reverse would happen if a bill came instead of a rebate.

For further example, if Hoedads received it in February of 1986, a
choice could be made to apply it to the 1985 or 1986 books. At no
time, to my knowledge, did Hoedads retroactively apply such a
transaction to the year of origin. Once a patronage dividend was
declared for a year, it was not adjusted in later year.

As an aside, this was also the case in the treatment of three-year
stewardship contracts. Guesses were made each year for the amount of
income and expenses for each year, but actual income wasn't known
until the end of the three-year contract. This could (and did) create
an uneven allocation of profits and losses within those three years of
the contract, but retroactive changes were not made even it out. I
can't remember which accountant recommended this method of dealing
with it, but Hoedads did it that way on four different stewardship
contracts from 1985 through 1994.

On the surface of it all, it looks like the historical practices of
Hoedads in dealing with time-delayed payments or bills is in conflict
with Charles Glass' reported opinion that retroactive patronage
dividends is a legal (tax) obligation.

Having reread Glass' letters of June 24 and August 21 a few times now,
I become less sure that his opinion is as black-and-white as
perceived. His opinion states that his analysis is based on certain
assumptions about Hoedads and its structure, some of which might be
incomplete or not totally accurate. I wouldn't be surprised that if
additional information and history were given to him his specific
opinions would soften or change.

It doesn't seem to be in anyone's interest to finance a fog belt by
getting more and more conflicting opinions, though, so here's a quote
from his June 24 letter:

"It is also likely that these bylaws were not written with the
current situation in mind, and therefore various
interpretations could be asserted which could result in a legal
conflict. As a practical matter, the amount of proceeds
involved does not justify the cost of resolving such conflicts
and uncertainties through an adversarial process. The only
practical approach that makes sense is to develop a
distribution plan that compromises the conflicting interests as
equitably as possible, with the largest consensus attainable,
to develop some sort of due process that gives any objecting
party the opportunity to have their objections heard at their
own expense, and then determine the distribution shares and
distribute the money as
expeditiously and cost effectively as possible."

BRUCE M. 1/20/98