[IPAC-List] calculating turnover

[mhammer at 295.ca]

Again, a need for further clarification.  I am a bit confused by your use
of the term "termination".  Are you using that term to refer to voluntary
departures only, or does it also include any egress of incumbents, for
example by firing or retirement.

I reiterate, the numerator and denominator that might provide the clearest
and least ambiguous metric could vary with the purpose.  If the intent is
to simply identify what sort of ongoing staffing needs there might be,
then whether the opening occurs because of retirement, firing, or other
voluntary departures, is of minimal concern.  Whatever the source of the
opening, one still needs to fill it, so turnover would be how many
incumbents we had and how many were still in place for the time interval
under consideration.  If the intent is to identify retention issues, then
the denominator should probably consist of all positions MINUS retirements
and terminations.  The gap between how many were in their position at the
start of the time period, and how many were still there at the end of that
interval, gives an indication of the extent to which voluntary departures
pose a challenge.

In other words, the purpose identifies what constitutes a a normative or
non-normative departure, "terminaton", or opening.

Mark Hammer
Ottawa

> To reiterate, dividing the number of terminations by the average headcount
> does not always give a good indication of the "% of the total number of
> people working there who have left", which is what makes it misleading for
> many users. Consider the illustrative example below of a recently created
> division/company:
> Headcount at the End of Each Month (including new hires) Terminations in
> Each Month
> January 5 0
> February 5 0
> March 5 0
> April 5 0
> May 5 0
> June 5 0
> July 4 1
> August 4 0
> September 4 0
> October 4 0
> November 4 0
> December 100 10
> The total number of terminations in the year is 11 and the average
> headcount is 12.5, which leads to a turnover rate of 88% by the typical
> calculation, but that doesn't mean that 88% of the total number of people
> working there left. It means that 88% of the *average* number of people
> who
> worked in this division left, which is not usually what the
> leader/organization actually cares about, in my limited experience.
>
> Kind regards,
> Kevin
>
> On Tue, Oct 10, 2017 at 8:27 AM, Shekerjian, Rene (CS) <
> Rene.Shekerjian at cs.ny.gov> wrote:
>
>> In general, I like the approach that is simple and allows the number to
>> be
>> greater than 100%. For example, if there are 100 positions and  200
>> people
>> leave in the course of one year, you would have 200% turnover.
>> Intuitively
>> satisfying and easy to comprehend. If 50 people left, there would be 50%
>> turnover. Both numbers paint a picture that is easy to understand, and
>> depending on the industry, may be shocking or encouraging.
>>
>> The problem I see when the approach leads to 50/100 yielding turnover of
>> 30% is that it requires mental adjustments. In one sense it is accurate,
>> but it hides the fact that 50% of the total number of people working
>> there
>> have left. They may have each been in a different position or been from
>> a
>> small number of the positions (such as when a manager is unbearable),
>> but
>> the overall effect is easily understood.
>>
>> Following Mark's line of thought, I think you would want the more
>> sophisticated approach if you were trying to compute the effect on
>> productivity, output, profits, etc. But for a general measure of churn,
>> I
>> think the simple approach is better.
>>
>> René
>>
>>
>> René Shekerjian
>> Director, Testing Services Division
>>
>> Department of Civil Service
>> Empire State Plaza, Agency Building 1, Albany, NY 12239
>>
>> (518) 402-2660 | Rene.Shekerjian at cs.ny.gov
>>
>> www.cs.ny.gov
>>