--On Wednesday, 12 January, 2005 08:22 -0800 EKR <ekr@xxxxxxxx>
> Sorry to be difficult, but no.
> I'd like people to explain why they think that the BCP should
> impose a bias towards outsourcing as opposed towards doing
> things in the most efficient way possible.
Personally, I've never been convinced that outsourcing is going
to be generally optimal. It isn't clear that it saves money in
the general case and may create additive management (including
accounting and administrative) cost overheads -- our (the
IASA's) costs to manage the oursourcing organization (*see
footnote) and that organization's costs to manage the people
doing the work.
In addition, for the key standards secretariat functions (I
refuse to use "Clerk" here because I think it prejudges what
those functions should be in our environment), I think we may
discover that we need to have those functions in-house, not
because of costs, but because we may want to be in a position to
have the IAD (or Exec Dir) apply very specific close supervision
of individuals and, in the worst case, single someone out and
say "you are not doing your job, correct it or you are out of
here". With an intermediate organization, you generally can't
do that without firing the organization itself (although you can
hold a lot of nice meetings trying to get the organization to
correct the problem) and, if you can do it, you risk a
"statutory employee" situation, which is usually the worst of
all possible worlds.
So we probably agree: my preference would be to drop all of the
"outsourcing bias" language, retain the notion of zero-basing
everything on an annual basis, expect the IAOC to be public
about the reasons for any employee slots or contracts and why
they are the best (optimal, which may go beyond "efficient")
ways to accomplish those functions, and stop there.
I suggested the text that Harald adapted because my impression
at the time was that a decision had already been made to play
out the "one employee" doctrine as a primary principle and that
we weren't going to get unstuck from that mentality, even if
there were a few more employees than that. But perhaps this is
than and that is now and we should reopen this.
*footnote: I think the document may be confusing (I hope not
confused, and I trust it isn't deliberate) as to whether all of
these pointers to "outsourcing" imply
-- hire an organization, with its own staff, etc., to
perform a particular function, or
-- hire a collection of individuals, but on a putative
independent contractor basis, rather than employees
I have generally assumed that we are talking about the first
model because it does transfer some group management functions
out of the organization, with the IAD needing on "only" make
sure the organization is doing its job and because it ought to
be possible to write contracts in which the organization gives
us some protection against people dropping out or being subject
to truck fade without our having to assume all of the marginal
or contingency expenses. It can work especially well when the
nature of the role, reflected in the outsourcing agreement, is
(as with the RFC Editor today) "we reach agreement about the
job, go off and do it and determine details more or less as you
see fit, and we will review things a few times a year".
The second looks nice on paper, and _may_ permit faster
terminations if there are problems. But it usually turns out to
be a negative experience when someone is doing a job that
requires regular interactions, calibration, and review/ signoff
of individual decisions. But it rarely saves money and may
increase aggravation costs unless those conditions are met and
provides us zero protection against truck fade since the IAD
isn't even in a position to organize cross-training, etc.,
without an extremely high risk of having the contractors treated
as employees for tax, benefit, and working conditions purposes.
If we are going to leave "outsourcing" language in the BCP, it
might be good to explain what is intended, or at least to add
to the list of expectations of the IAOC that they will document
and justify those decisions.
Ietf mailing list