Thursday, September 15, 2005

Orders from line management

Dear Doug

As a low-level LANL supervisor I just went through another grueling
month. Our top line management has ordered us to make our personnel
performance scores fit a gaussian (random) distribution regardless of
the quality of the people in their respective peer groups. Half of my
group's employees are really angry, demoralized and threatening to quit
as they are all good performers and don't deserve to have artificial
constraints placed on their reviews. I see yet another lawsuit in the

That's they way LANL staff have been "rewarded" for their contributions for as long as I can remember.

LM, take note: you could do much better than UC has done at implementing an incentive system that rewards performance.
Here in lies the problem, not so much with the directive as described, but with the people in charge of implementing the directive. This "low-level" supervisor (team leader?) doesn't even realize that a gaussian distribution is anything BUT a random distribution. This SUPERvisor goes on lament the fact that they supervise a population of people that can only be described as having grown up in Lake Woebegon, where all the women are strong, all the men are good-looking, and ALL the children are above average! Get real LANL supervisor, make the tough calls and earn your keep for a change. Real anger can lead to catharsis in ones career, improving performance in future years.
9:33 AM is clearly a LANL manager who has thoroughly been schooled in the UC way. I hope your Lockheed supervisors smoke you out quickly.
Take it from someone who has taught for 10+ years at all level of higher education. If the group (team, etc.) has <10-15 people, then insisting on a Gaussian shape is simply wrong. On the other hand, for a larger group, the stars have to be really perfectly aligned to get something other than a Gaussian. In any case, evaluations are relative to peers, and, while everybody may be an oveachiever in one's particular group, some are still better overachievers than the rest. It simply cannot be that everyone is better than the average.
But I thought LANL only had the "best and the brightest"?
Not to mention the "World's Greatest Science Serving Society"

Damn we're good.
Now, if UPTE would just earn its keep and figure a way to ensure that every employee receives a better than average raise.
Its easy to attach UPTE but; one of the key areas that SPSE (LLNL's employee organization) has done some real good in is that of pay/performance. LLNL managers, like managers, had a very poor record of paying for performance; using very corrupt practices which would seem very familiar to LANL employees. SPSE won the point in California Courts.
As a result LLNL practices are a good deal better than LANL's. Hopefully, even if UC wins the bid (shudder) Anastasio will bring in some of the LLNL practices. If so, the LANL employees will benefit from the efforts of SPSE. But, this is simply a hope. LANL management may convince him that "New Mexico is different". If so, the game goes on.
You meant, of course, "...even if the Bechtel/UC consortium wins..."
Individual performance reviews as conducted by LANL, and most other employers, are nothing but demoralizing wasters of time. By pitting each butthead against the other, they are especially destructive to team/group efforts (See Deming on that one).

The supposed link between pay and performance is tenuous at best. With 2.8% to pass out it simply makes no sense to attempt to reward people differently. (You get 2.37% because your ORC is 8.31. You get 2.52% because your ORC is 5.49.)

But that's what we'll do, and we will spend thousands of hours away from the work that the taxpayer wants us to do in order to explain and attempt to justify these differences.
Why 2.8% when Aug'04 to Aug'05 inflation was over 3.6%?...and this doesn't even include the effects of spiked energy prices that happened in the last few weeks. This place doesn't even keep up with inflation anymore.
The original poster should not take it personal. I had the same stupid-ass nonsense for many years as a GL.

I worked very hard to staff a very strong Group and I pushed these people to do very good work and they did.

But, it made no difference at all in our division. We had to reduce my recommended ORC scores every year. Each group got basically the same percentage for raises. Indeed, because I had hired in some very good people, our average salaries were above the division average and we usually got a smaller percentage so that the other groups could catch up. In the end, many of the members of the group found out that they could do better on ORC score and salary in other divisions and so they left.
There may be a union at LANL some day but it is not going to be UPTE. The leadership, past and present, are just not representative of the LANL workforce.
Tying salary increases to performance always ends up being a subjective process. Managers can always invent "objective" justifications for larger increases for their sycophants, yes-persons, and boot licks. I fail to see why HR managers bother with merit raise systems when they should find some more useful work to do.

I see the place has only become slightly worse in this regard since 20 years ago when my group leader was honest enough to explain my low raise because my lips didn't form the correct embuchure to kiss his attached wind instrument.
LANL is no longer competitive on salaries and benefits. Just look outside sometime.

After this most recent 2.8% insult, I plan to vote with my feet prior to UC's official departure.
9/15/2005 04:53:48 PM said:
"But, it made no difference at all in our division. We had to reduce my recommended ORC scores every year. Each group got basically the same percentage for raises."

It has been that way every year since the system started. I was always held back because my salary was higher than others in my peer group. It didn't change because 2% of my salary was always higher than 2.x% of their salary. We all got screwed, but I was happier to be me because I was able to retire with a pension that I could live on. I wish I could say the same for them.
Speaking of demoralizing, my group hands out flat dollar amounts for each ORC category. For example, two people can receive an ORC score of 8 and get the same dollar figure even if one makes $80K and the other $120K.
08:13:35 PM

There could be a very good reason one TSM makes $120k and another $80k (long histories of excellent or lousy performance, early vs. mid or late career, etc.). If they both performed at an ORC of 8 (job content + performance)why not give them the same raise? To try to force them both to some nominal average figure may over-reward the person with a history of sub-par performance who finally woke up and started working at the expense of someone with a history of excellent performance.

That's socialism.
While I don't want to defend the silly "Gaussian," I do want to point out that it isn't easy to come up with something better, as anyone who's had to do these miserable things will agree. Has anybody here had experience with a performance appraisal system that actually works? Examples would be welcome.
All this takes me back to the reason I left management at the Lab 30 odd years ago. It seems that those who dished out the money assumed that 80 to 90 percent of your people should be fired for incompetence. The money they gave you for raises would give an adequate amount to your top performer(s) only by giving nothing to everyone else. Actually the 90% incompetence level at the time was pretty much among HR and those who decided how much money to pass out, including ERDA/DOE.
Re: 9/15/2005 08:43:09 PM

The answer is no. I took a graduate course in personnel management and tore hell out of the various schemes to match salary incentives to performance measures. The curve was my favorite target because it represents a schema that in most cases does not mirror reality. What if, I claimed, you had made excellent hiring decisions and everyone did very well this year? How do you explain to a strongly performing employee that his work was good but he somehow came in low on the curve? "Why did I come in this low on the curve?", he asks. Answer that one, preferably in a way that does not suggest to the employee that he start working on his resume for another job.

The curve is not justified simply upon the basis that "we can't find a better way" to match performance to salary increases. Ignorance is not a justification for action. At best, it is a justification for inaction, which means in for this exercise giving everyone the same raise unless you can justify giving someone a better raise simply because he happened to be hired in at a substandard salary.

My final destruction of the "book" schemas for performance and salary increase coupling is that none of it applies to quasi-academic staff as found in places like LANL. If you want to apply such structured rules to the problem, you have to define specific tokens that signal a level of contribution to a scientific enterprise. What coupons can we expect a staff member to contribute to earn a decent raise this year? I say you can't because you want your people to be working for knowledge and insight rather than counting up reports, outside publications, participation in professional activities, &c. Trying to do a performanace appraisal in conventional terms for scientific staff would be like evaluating everyone of them for tenure each year. Who wants to go through that on an annual basis. I'd use the process to first remedy salary inequities among my scientific staff first and then distribute the remainder equitably. I have never seen a bad hiring decision made at the Lab for technical staff and therefor no need to use the process to motivate a staff member to look for another job.

However, when staff members float up into management, that's another story, and at the managerial level, the conventions and rules in HR apply very well. At this level, you are looking for specific performance results and we are perfectly justified in matching salary increases to how well managers produce the accomplishments we expect of them.
For a laugh and perspective on the whole performance appraisal process visit the following page:

Nice food for thought on the current management philosophy.
Unfortunately, the UC contract mandates that performance reviews be done every year for every employee. HR has tried to make this meaningful but the very nature of appraisals and salary adjustments has always been a "lose/lose" proposition. Your poorer employees are mad because of their appraisal and lousy raise and your most valued employees are expecting a higher ORC score and a much higher raise. HR tells managers that we have to do "salary management" which, to me, meant "tie the salary to similar work by others irregardless of seniority". Thus, if two people are doing the same or similar jobs and are contributing the same amount but one makes $80K and the other makes $120K, the manager should give a higher percentage to the person making $80K. It's not perfect by any means but I've yet to hear someone convince HR of a better way although I'm sure there are.
Ever since we started doing ORC scores people have complained so I'm wondering what benefit they serve except to help identify people who should go on action tracks.
If you don't like the LANL system, check out how they do it at GE! It is described in Jack Welch's autobiography (I'm not really a fan of Jack Welch, but it's an interesting read).

Basically, every year, you get to divide your direct reports into three pools. Call them A, B, and C. Once you've worked out who goes into which pool, then the fun begins, because you lay off all the folks in the C pool.

As Welch describes it, the first 2 years, this is pretty easy. Year 3 and on it gets a little harder.

I would not want to work this way.

oh boy. I just made a fair number of mistakes and posted my comment several times. Sorry folks.

Doug, can you fix :-)

I've heard the same stories about GE. Basically, they fire about 5% of
the "bottom feeders" each year. It quickly gets rid of the riff-raff, and
keeps the rest of the staff intensely focused on their job performance. The
coming RIF which awaits us under our new masters will accomplish this
same task. Anyone wish to take bets on the euphimism they come up with to
replace the over-used RIF acronym? "Workforce Restruction" is outdated.
A new, less threatening term is required. And when they start it up, you
can be sure you will never hear our new managers use the term "layoffs".
It will all be done in a very slealthy, hush-hush manner.
I didn't know you were into HPP (High Performance Posting, Ron.


Other organizations, I believe Bell Labs, have a similar program as GE to RIF the "lowest ranking" 5 or 10% of thier employees. I don't think that this would work very well in Los Alamos where there are no other jobs. I just cannot imagine trying to explain this to a job applicant during an interview.

Just as we are unable to do anything sensible with ORC scores and salary increase pools other than edict quotas, the policy of culling the non-performers would be impossible to administer in a fair manner. There would have to be politically correct quotas for women and minorities. Just a disaster.
8:43 here. 10:15:09 says, 'The curve is not justified simply upon the basis that "we can't find a better way" to match performance to salary increases. Ignorance is not a justification for action.' I agree and didn't mean to imply that it was. Still, something better needs to be found, and all I seem to be hearing here are stories of things that don't work, rather than things that do.

I wonder how this would be received: establish a flat raise schedule lab-wide that is based on age, marketplace, etc., and NOT on merit -- and then establish a large number of things that would entitle an employee to "bonuses" on top of the base salary if successfully executed by the employee. There would have to be some serious work to ensure that the "bonuses" are (1) representative of all the skills the lab needs, (2) objectively verifiable and suitable for handing out in an unprejudiced way, and (3) realistically achievable by simply doing one's work particularly well. This might not be easy.
The whole LANL "peer group" salary evaluation and performance reviews are bogus and not based on anything concrete. For people with skills and real credentials industry pays better anyway than the DOE or its labs.
Right here in NM, Intel in Rio Rancho uses the same method as GE to turn over their employee inventory annually. Maybe the whiners on this blog that think the review system is unfair should read Andy Grove's book titled "Only the Paranoid Survive". If you work at Intel you WILL be paranoid. If you are not a kiss-up it does't matter what you do, or how much you do.

Review time is affectionately called "Ranking and Spanking" by Intel's upper management. Every Group Leader MUST rank their group across the entire ±3 sigma range where -2 sigma souls get "re-deployed" even if the GL's have as few as eight employees working for them, or even they themselves face risking "re-deployment" with the unlucky souls to compete for a dozen jobs.

There seems to be many whiners on this blog that have obviously never worked anywhere else but the lab. Frankly, you should count your blessings, suck-it-up and just hope that UC keeps the contract.

Believe me this is easy street compared many other places in the private sector.
From a post I deleted:

"Well hello all...I am not sure Doug will post this...I have been banned in the past for suggesting that ADM Nanos was, in fact, not the Anti-Christ, but lets see if I get through the censors.

Since the above-referenced poster did not leave his name, I can only guess who he is, but I will, in fact continue to delete drivel such as was posted. Since I have only banned one person from this blog I should probably conclude that Stradling has attempted to make his voice heard. Again.

Gary, if that is you posting from behind the cloak of anonymity, please don't do this again.

When I first became a team leader, I got to assign ORC scores. My then Group Leader immediately came back and said that so-and-so "has never gotten below a 9 in his entire career at LANL". Now, this person is a good performer, but I never saw anything above an 8 so I didn't know what a "9" meant (I understood, however, that "9" was once the average score in NMT). Now I'm wiser. I always make sure this person gets above a "9", and try to get everyone else a real high score -- because everyone *else* will always have their scores knocked down by my ex-GL

Salary may be related to ORC score, but the single most important factor is where you start. A PhD making < $90K at LANL probably means that his management (like mine once) screwed him and converted him at the lowest possible pay. I've been there too.

Finally, I have no problem if we get rid of those whose performance and contribution are consistently low and who have made no progress toward improvement. As much as I believe we have the "best and the brightest", it's still the "B students" in each organization that really do the work. I haven't quite figured out what will happen to the B students if we get rid of all the C students.

In all the discussion of other entities attempts to manage their personnel we don't seem to want to admit the geographic oddity that is Los Alamos. LLNL and Sandia have relatively HUGE populations from which to draw employees and to which to 'return' employees that do not work out. Most other large employers are situated in or near large population centers for obvious reasons (customers, suppliers, tansportation, etc).

Let's face it. The support organizations are disproportionately populated by 'local' talent. The tech series are also heavily populated by local people. Any restructuring that looks at cutting the numbers in either of these areas will adversely affect the local communities, mostly off the Hill. That is the way the demographics work out.

I keep remembering the comment from a rather uncomfortable group meeting a few years back when the technicians received a significantly lower raise package than the TSMs did. One Tech asked the question how the lab could justify doing that. The group leader bravely answered the truth that we all already knew.

He said something to the effect that if the lab doesn't pay close to the national average for the TSMs, they will transplant somewhere where they are compensated better. If we only pay the techs what the market analysis says that we have to what are they going to do? Most techs in our group do not want to commute to ABQ to work at Sandia or move out of the area. They have a much stronger connection to the area and will bare a lower wage to not have to move.

That is a harsh reality but reality just the same.
What ever happened to salary management? Every year it is the same, raise management.

I don't see the inequity of two workers with the same ORC score getting the same raise. If I am performing at the same level as some one with 5, 10 or 15 years more experience shouldn't I be rewarded with the same raise instead of the same percentage raise?

After eight years of outstanding ORC scores I have yet to reach the 'line' from my first year. I guess doing your job well really is its own ( only ) reward.
I haven't quite figured out what will happen to the B students if we get rid of all the C students.

Karen - I know what will happen, and so do you... The people who actually do work (the "B students") will not have to uselessly sink more time into mentoring/covering-for the "C students." A cleansing would be a good thing, and might make the people who actually do work slightly more productive.

The ORC score process is a joke, so it doesn't really matter if the "C students" are around or not.
Well Doug - I guess that about proves it - Someone gives you some interesting information about how LM racks & stacks its people and you won't post it because it points out, as many have on this site (probably more than you have allowed visibility) that the amount of whining and paranoia on your site is very high, disturbingly so to many outsiders.

I am outsider, a one time "user" of the 'traditional" products of the national labs. I am not the hated and apparently feared Stradling.

I had the privilage (yes) of serving (twice!) with Pete Nanos and never found the characteristics many of your posters discuss to be evident, but I suppose people change. Or perhaps, people have to change in the circumstances they are in and they face an organization run amuck.

Now I am in private, commercial enterprises, unrelated to defense or the nuclear weapons complex, but still fascinated by the Blog because of my past life and past knowledge of the person whose actions "started" this blog.

I find many of the things that this particular thread of posts discuss (evaluation standards) occuring in private industry.

I suspect they will also begin to occur at LANL REGARDLESS of who wins. Like it or not, if these type of very common evaluation standards are implemented at LANL you will be required to align your activities with management and they (not you!) will be the final arbiters of your success in achieving that alignment

You will be held to standards that your management puts in place. You probably will be provided an opportunity to make an input, but if you decide that your input is of the quality and tone that I see often on this blog - you will be ignored and held to the standard required. Many of you will then be "rewarded" in accordance to that standard.

Again, good luck in the coming months and years. I hope this non stop Orwellian Hate Week session brings you much satisfaction.

I suspect that many of you will be unhappy regardless of whomever takes over LANL - I suspect most of your unhappiness will be of your own making.

The only thing that I regret is that my tax dollars will be funding what appears to have become a very high tech kindergarten.

All ready missiles away - Out
I am getting more and more disturbed by the tone in this blog. The discussion of ORC scores further points out that LANL collectively needs to decide what it is, and where it fits in the grand scheme of things. I heard a saying when I came here in the late 80's that LANL wes really 15 labs connected by a common janitorial service. Taking that one step further, I think we have 15 or more self identities and associated views of the outside world. It will be intersting to see how this changes next year.

Are we a nuclear weapons lab?

Are we a national security Lab?

Are we a preimier scientific lab?

Depending on how group and division management answers those three questions, Job Content and ORC scores can come out very differently for the same set of people doing the same jobs and performance.
08:18:09 PM, my bullsh*t detector went off twice while reading your comment.

1) The LM article currently at the top of the blog is one that is harshly critcal of LM, accusing them of UC-like retaliatory behavior against an employee, and

2) I also had the "pleasure" of working with former director Nanos. He was an arrogant, abusive, egotistical prick. The only people he liked having around him were obsequious, sychophantic, yes-person pricks. Guess what that makes you?
I would like to second the comments of anonymous at 9/20/2005 05:34:17 AM
Ditto what 07:55:32 AM said.
Can I second (third? or ditto?) what anonymous at 9/20/2005 09:06:26 AM said?
Mega-dittos! And I have my doubts whether Poster 8:18 pm is really who he
claims to be. If you served (twice!) with Pete Nanos, then please tell us
where and when to validate your status as one who truly knows the man.
It was Stradling, posting anonymously.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?