Employee Ranking Systems LO17553

FRASER,Doug (doug.fraser@deetya.gov.au)
Wed, 25 Mar 1998 17:20:37 +1000

Replying to LO17523 --

I'm just about to unsubscribe, but thought, as a parting gift, it might be
useful to contribute some recent ideas on performance appraisal:

(a) a relevant reference supplied by a contributor to Govteval; (b) my
analysis of the implications of same; (c) (Word attachment) a checklist I
drew up for analysing performance appraisal systems.

[Host's Note: Sorry, I don't distribute attached files on the LO list;
I've deleted (c) from this msg. ...Rick]

(a) ----forwarded msg-----

An article in the latest Canadian Journal of Program Evaluation entitled
looks at some of the issues being posted. It is by J. Peter Venton, "A
General Theory of Delegation, Accountability and Empowerment" Vol.12 #2

Venton comments that the problem of inadequate understanding of these
terms stems from an underdeveloped literature which does not employ common

Here is how he unpacks Accountability:
Context: The problem of the possible lack of agent commitment provides
the context of accountability.
Purpose: The purpose of accountability is to build the agent's
commitment to serve the goals of the principal. An important byproduct
if that commitment is the principal's trust and confidence in the agent.
Definition: Accountability is a condition in which the agent is likely
obliged to report measurable gaps between the principal's expectations
and the results of the discharge of the agent's delegated duties, to
explain the gaps satisfactorily, and to make amends in the event of an
unsatisfactory explanation.

He comments that inherent in the definition of accountability is the
assumption tha the principal's expectations must be measurable and that
"without measures there is, by definition, no accountability".

As for the program evaluator's role in this, he notes: "Unsophisticated
principals may assume that the agent's actions and decisions are the sole
cause of all results. Satisfactory explanations depend on the agent's
ability to attribute gaps to external factors rather than factors under
the agent's control. The measurement of this attribution is a core
function of program evaluators. Without such program evaluation there are
large risks to the agent."

- --
Saxon Harding,School of Business,
Carleton University, 1125 Colonel By Drive,
Ottawa, Ont. Canada. K1S 5B6
E-mail: sharding@business.carleton.ca
Phone: 613-237-8038

This email comes from the list: govteval
-----end of forwarded msg-----


The model of accountability cited in Saxon's posting gets, in my view, to
the core of the problem.

This is the premise that the "principal" is perfectly informed on the
required outcomes, and perfectly committed to them, whereas the "agent"
is not.

I wonder how many people who have worked in service-oriented public
agencies would see this linear model as a good reflection of what really
happens. Certainly the experience of our own soon to be disbanded
Commonwealth Employment Service provides a long list of counter-examples.

For those of you who have not already made the connection, two of my
recent postings are relevant to the failure of this model:

(a) the other day I mentioned that senior agency executives and ministers
tend to operate in a closed environment of pressures, values and
incentives which exclude real outcomes; whereas the people at the counter
have a much better feel for at least a part of the need. Consequently,
most goal displacement tends, pace the public choice theorists, to occur
precisely at the level of the "principals";

(b) my concept of fuzzy logic means that governments (the higher-level
"principals") are often precluded from defining their "expectations"

My experience is that in practice, the outcomes of any political process
are negotiated ones. That is, not only the actual outcomes but even the
desired outcomes are unknowable in advance; they are defined progressively
by the inputs and interactions of different groups all the way down the
supply chain.

Somehow this still seems to happen even when an authoritarian, Taylorist
framework is imposed on the process. It's just that this imposed famework
impedes organisations from arriving efficiently at appropriate negotiated
outcomes. Whereas the goal of a post-bureaucratic organisational form, to
which all "reformers" of whatever colour feel obliged to pay lip service,
is to harness individual inputs at all levels in a way that facilitates
the process. And this in turn ties into the disciplines of organisational
learning and knowledge management, which begin from the premise that
information, knowledge and expertise are dispersed throughout the
organisation or delivery chain, and that the management problem is one of
coordination rather than subordination.

(Curiously, the purchaser-provider model, which describes exactly the same
situation as the principal-agent problem, only with a positive spin on it,
captures this latter aspect of reality much more accurately.)

>From your perspective as someone who helps to develop or implement the
accountability framework, the critical question is therefore: does it
favour the efficient and transparent negotiation of outcomes, or does it
hide and impede the process by imposing an inappropriate Father-knows-best

For a post-bureaucratic model of government or organisation doesn't
eliminate accountability. It simply redefines it in what may well be a
more demanding sense: the electorate (the shareholders) have delegated a
government (a board of idrectors) to ensure that particular classes of
outcome are most effectively negotiated on their behalf, and they are
entitled to know that the process has indeed been efficient and fair, and
that whatever goal displacement has occurred has been in the public (the
company's) interest.

The central element in this definition of accountability is
*transparency*. Hence my constant advocacy of a descriptive as opposed to
a normative model of evaluation, and of performance measurement in

(c) --FRAMEWORK for assessing PA--
[Host's Note: Deleted attached file... Rick]

Doug Fraser
DEETYA, Canberra

Learning-org -- Hosted by Rick Karash <rkarash@karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>