« Malvolio Production Diary: table talk | Main | Malvolio Production Diary: hack, cough, wheeze »

Turn, turn, turnover, over, over

One of Your Humble Blogger’s co-workers has announced that she will be moving on to a new job elsewhere, and I thought Man, this place has a lot of turnover. And then I thought I wonder how it compares to other places. And after that, I thought I wonder how one measures that sort of thing. I looked on the internet and found a bunch of articles that all seem dim and useless; I suspect that I could use my library skills to find better ones, but I won’t. What I will do is ask my Gentle Readers what they think about it—what concerns you about turnover in a workplace? What should the Director of the Libraries be measuring? How do you think a good analysis would weight the various factors?

There are around 20 full-time employees in the library system, plus around six part-time permanent staff, plus another thirty student workers who aren’t really relevant to the question. There are currently three open positions—that is, two vacant at present and the third that made the announcement that began this post. Fifteen percent of the staff vacant is just crazy high, isn’t it? But maybe that’s a fluke. Twenty is such a small sample size that coincidence will play a big part. And this is three coming together over the summer that seem, on the face of it, pretty random: one was a retirement after many, many years in the same position; one was here just under two years and is leaving because of a spouse’s opportunity across the country; one was here two-and-a-half years and is leaving for a better position nearby.

And here’s the thing: the job that is vacant again after less than two years is designed as a starter job; it doesn’t require an advanced degree or a lot of work experience; they are likely to hire someone right out of the undergraduate program here, and that person will stay for two years or so and move on. You can argue whether that’s smart or not (well, no, I don’t really think you can argue that it’s smart in the long-term) but the turnover in the position is not a surprise. And the fellow who is retiring after many years was in a position where you want someone to stay for many years, but then he did; you do want eventual retirement, so that’s not really dire, either. I’m not saying it’s not a bad sign to lose three in a summer, just that it doesn’t necessarily imply a mass exodus.

So how do you measure? I mean, what are you comparing against? It seems to me, in a smallish place, that you want a combination of institutional memory and new blood. You want to have a bunch of people who are happy and stay for a long time, and you want to have some turnover so that things aren’t entirely calcified. You want some people who have been there for a medium-length time and look to become the people who have been there forever, because the people who have been there forever won’t be there forever. Average length of service doesn’t seem helpful. Neither, really, does the total turnover percentage over a length of time. I mean, both are informative; one hopes that both are being tracked, but I don’t think they tell you much about the health of a department.

Would you want to know what percentage of the staff have been there for ten years or more? That seems like one potentially useful piece of information. How many hires stay more than three years? That seems like another. They don’t seem to necessarily relate to each other, though. Maybe there’s a way to measure based on expectation, that is, if the expected tenure for someone in this sort of industry/position is n years, measure for each departure how that departing employee’s tenure matched to that, giving credit for those who stay longer that that time and demerit for those who don’t stay that long?

Or is there some optimum total years-of-service aggregate number for a twenty-person staff, perhaps by weighting the first three years as 1/3 value (or whatever) and the eleventh-through-twentieth at a year and a half? Such that the ideal twenty-person staff would have, oh, 250 year-service-points, and that when the number dips below 200 (or whatever) it’s a problem? I tend to assume that it’s impossible for to have a single useful score, but it there is one, it’s presumably a weighted formula along those lines, isn’t it? Such that if that number is in the danger zone, you need to look at why you aren’t keeping your people.

Of course, the manager of such a staff will always know more than a number could tell them—knows if there were a batch of bad hires that were pushed out and replaced with good ones, leaving an experience gap that is better than having a crappy staff. Or knows if there’s just a coincidence of some kind that doesn’t need to be addressed at all. Or if everyone is complaining about crappy pay all the time, and it’s just a terrible job market, such that the organization can retain good people up until the moment the job market picks up, at which time everyone will be gone. Or whatever. The problem is that the manager of such a staff, while knowing more than a number could tell them about their staff, will know less than the number could tell them about other staffs—which means that things that look like coincidences are in fact total outliers, and perhaps need to be looked in to on the chance that it isn’t so much a coincidence as nobody willing to admit why they are all looking for jobs elsewhere. Numbers are helpful for that, although in this case I have no idea which ones.

Tolerabimus quod tolerare debemus,


Offhand, i'd keep it simple and look at the percentage of the staff that leaves each year, and compute that for several years in a row. That tells you whether this year is unusual compared to previous years (as opposed to just that this month is unusual compared to last month, which you already know), and it gives you a ballpark of what "normal" for your organization is.

Comments are closed for this entry. Usually if I close comments for an entry it's because that entry gets a disproportionate amount of spam. If you want to contact me about this entry, feel free to send me email.