Saturday, May 07, 2005

Now, lets have a FACTUAL discussion!

From Anonymous:

I applaud your call to start working the issues. I
have been at the Lab for more than 20 years, and my
highest management position was Group Leader. I have
worked with many managers at higher levels, including
some EB members.

The first and foremost issue I see today is that group
leaders have lost their ability to lead, because their
budget has been moved up to the division, and they
don’t have any control over it. One of the things
that Nanos did right, when he first became director,
was to review in depth what the group leader role was.
Unfortunately, he didn’t follow through. I suggest
we resurrect that research and implement it ASAP.

The second issue is the Enterprise Project. We need
to ask why such a large investment is needed, at the
expense of science and infrastructure at the Lab. In
particular, to support the previous issue, it has to
allow for individual cost codes for groups.

Third – the overhead issue. Nanos cut down on
overhead but not on the overhead work. This causes
all sorts of possibly illegal cost accounting
shenanigans. We need a realistic overhead rate,
coupled with an effort to decrease overhead work.

Fourth – choose group, division and directorate
leaders who have demonstrated their ability to lead
people. Not “project leaders”, “program managers”,
etc. I think you know who I am referring to, but lets
start avoiding name calling.

Lets also recognize that running the Lab today demands
an immense amount of administrative work, and that
“chiefs of staff”, “special assistants”, “senior
advisors”, etc., are an unfortunate consequence. LANL
needs to work with DOE to decrease the requirements
that generated this busy work, but as long as the
requirements exist, these people need to be
appreciated like everyone else. (disclaimer: I am NOT
in any of these categories).

Now, lets have a FACTUAL discussion!


As a former GL, I have the following comments. I agree with most of this but take issue with some of the details.

First Issue: I AGREE! In the past five years, as far as I could tell, all that the Division Office (and AD) did was tax us. Funds were usually brought in at the Group Level. What the Division Office often did do, is tell the GL that individuals in other groups had to be "funded" from project. This was often done even if those individuals were unsuitable for the work.

Second Issue: I AGREE! The Enterprise Project is unreasonably expensive, on the order of $150M. And, there are very serious concerns about how well, if at all, it will work.

Third Issue: I DISAGREE! Sure, there was a cosmetic reduction in the G&A rate. But this was compensated by other increases. For instance, originally, the AD offices were covered out of G&A. After the Nanos reductions, there became a tax added to the Division Overhead and then the sum (plus the Infrastructure) is compounded with G&A. The net was an increase. The post is correct about the unfuneded added overhead activities. There are many added overhead costs that end of being charged directly. The people in CFO perform all kinds of accounting shenanigans to hide the true overhead rate.
Get rid of fake "central" organizations that deploy workers to full-time, ongoing jobs but prevent workers from being hired by the host organization. This forces the workers to have TWO sets of management, costs the host org roughly twice what it would cost to directly hire the worker (whether UC or contractor), and prevents the workers from any sort of career development with the host org. All this to pay "overhead" costs for the central org, which is top heavy with team leaders who do nothing because most workers are deployed out of the central org. Central orgs should exist for short-term projects where a group needs help on a one-time or limited basis. Long-term deployed workers should obviously be hired by that org. The issue of "shadow functions" is raised because of this forced, expensive, and redundant set up. In no way is this a "best business" practice when a host org pays 125K for a worker (who earns half that number) to pay for "overhead" in another group. This system does NOT work and is not fair to the deployed worker.
Speaking of overhead, the first stupid action that Nanos took that got my attention was the change to "one-size-fits-all" overhead. The problem with this is that theoretical/computational groups now get charged the same way that experimental groups do, so they are subsidizing the experiments! This is bad enough when all the money comes from DOE, but it is crippling when the funding source is external. Try to tell your sponsors that they should fund you over a cheaper researcher at a university because, well, they get a piece of DAHRT bundled with your work!
Second Issue: The Enterprise Project is a big waste. Yesterday evening I needed to do connect to the Computer User Registration databse as part of my OCSR duties. I was not impressed. Just to get to the login screen took a number of steps. Once connected the system is really slow and this was after 5:00PM. Then you get a tiny fixed size screen that is pretty much useless requiring you to scroll up, down and sideways endlessly to get to the data you are looking for. I recenty attended a open source conference with quite a few Fortune 100 companies in attendance. Everyone I spoke with regarding EP using Oracle stated that in the end it will not work and be way over budget. These folks saw that light and have implemented an open source software solution using LAMP for a fraction of the cost.

Regarding the Shadow Organizations I could not agree more that its just another way to fill the coffers of the centralized organization. The customer winds up on the short end of the stick so to speak.
If group money has been moved into the division and is then being handed back less a tremendous tax, then that needs to be changed. If the groups and the staff are charged with finding the programs, they should manage the money and have the obligation to include in the teams working on the programs the best people, not ones who for the buddies of the division leaders. Simply, if it's the group leader's job to bring in the money, it should the group leader's job to spend the money.

Once the division office tax is significantly decreased, the division leaders will have to decide how to fund all of those expensive yet worthless additional staff. They are worthless because they do not add to the value of the product since all they do is push paper from one place to another. This has become the chicken and egg problem - first DOE pushes paper down, LANL hires the staff to manage it, DOE says "good, there's staff so we'll push more paper", etc. Who is going to draw the line?

There needs to be a tremendous overhaul in the benefits part of LANL. Letters to the Reader's Forum enumerate problems with Delta Dental and United Health Care. Many of us do not understand why, when contracts are let, the providers can suddenly change the scope of the contracts. To many it appears that HR is unable to negotiate binding contracts and that LANL is not concerned enough about the employee benefits to ensure that the right people are involved in negotiating dental and medical insurance contracts.
In one letter, the author described becoming ill in Honolulu, going to a walkup clinic, being treated and filing a claim. UHC denied as out of network (I guess no one bothered to figure out how to insure LANL employees who travel outside NM). He saved them quite a bit of money by not going to an ER, but that was of little concern. In another letter, the writer described becoming ill, not being able to see her PCP (probably because of overbooking), and making the decision not go to an emergency room because the illness was not serious enough but going to a walk-in clinic. Of course, the charges were not covered. This person learned after the fact that a coworker visited a different urgent care facility and never had to pay any out-of-pocket costs. And UHC's response to all this was “you don’t have an urgent care in your area that is in network.” and "we would have denied your ER claim if you had proceeded to the ER because it wasn’t a life threatening situation.” The bottom line is that we again have poor health care insurance because the HR staff does such a poor job. It's interesting to note that there were no reported problems during the first year of this contract, but now that we are in the second year, UHC is beginning to show its true colors.

Finally, I'd like to address the concept of "shadow organizations". Sig Hecker is responsible for the central organization structure in effect today. He was director when CCN (formerly C for Computing) Division decided to set up a team of system administrators who could be farmed out for way more money than what a Division/Group would pay. Hecker also was the one who started to move of all the property, HR, and budget people back to central organizations although some of that was not accomplished until the late 90's. As the 09:34 poster wrote, this is to no one's advantage except for the host organization who makes money from it. This structure was modeled on SUP (formerly BUS) because purchasing has to be centralized because of all the rules. But there is no reason for someone who tracks property, prints shippers, etc. not to work for the division. It would be cheaper. The argument is always one of backup - the host organization will provide someone to fill in when the property person is on vacation. But that is rarely successfull because that person does not sit in the regular office, has no account on the computer, can't get to the records, etc.

The bottom line is that people should be paid by the division/group for whom they work. No matter how many checks and balances there are, it is not to an employee's advantage to have two bosses if the two have conflicting goals. This is not structured to meet the needs of the employees but to support certain divisions by taxing other divisions.
A further problem with CCN (and other "central" orgs) is that it takes away decision-making authority that rightfully belongs to line management. This, coupled with an entrenched self-protective instinct, is very difficult to work with. I've interacted with several CCN staff who seem to believe they are the sole experts concerning everything involving a computer. Their willingness to try to make programmatic decisions amounts to having a "shadow director" in the NOC. This arrangement is not serving our mission well.
and throw IM into that same scenario for editing, design, computer, and web work. Centralization works for short projects, but long-term deployed workers should be hired by the group paying for their services. Why pay for a bunch of do-nothing team leaders and customer reps who provide NO value to anyone at the lab? Centralized services do not work if the hiring organization needs ongoing services. Let's use some common sense here and streamline services and cut out artificial organizations that serve no purpose other than to create layers un unneeded mangement and team leader jobs. Dump IM.
Two comments:

First - The new Director should look at the EP project and see why we have spent so much and received so little. Second - HR can't do S*it and they do not care about employees, so why do expect benefits to do step up to the plate?
while I agree that the administrative cost of running LANL is increasing. We have ourselves to blame in part. How about HR and Rich (I like Women) Marquez, hire a office leader for a group with 7 people and still have a group leader?
HR strongly needs an overhaul in the personnel department as well. They need their wings clipped. I very much dislike having inexperienced non-technical HR employment "specialists" culling the lists of applicants for positions I am to fill. Very frequently the clearly most highly qualified candidates are discretely removed by HR so that the pool of "qualified" candidates meets the quotas for AA/EEO or whatever. Please HR, allow the very best candidates make it to the interview selection candidate pool! The prospective personnel choices that HR presently allows, are not the ones that are the best for the Lab. HR should limit their scope to simply processing the hiring paperwork...else please go away and make paperclip chains or something more useless and less damaging.
Dear Dr. Kuckuck;

Dump HR
Dump PM
and what the hell does PS DIVISION do?
The cost of doing business is extremely high at LANL. I make around
$115 K per year, a nice salary to be sure. However, my cost to sponsors is
now running around $340 K per year after DOE, PMs, Divisions, G&A, etc.
all take a bite out of funding. This can't keep up much longer. Four years
ago, I only cost about $240 K per year. Old timers who are lucky enough
to be in the $150+ K salary bracket now costs almost $500 K per year once
all taxes and overhead are added. This is crazy. Even the new Director of
LANL will only receive a salary of $355 K for managing the Lab. No one
is worth nearly half a million dollars per year. These extremely high
costs clearly show just how bloated the Lab has become in the last few
I agree with the 5/7/2005 12:13:35 PM comment: The "one size fits all" Division Overhead rates are a disaster for theoretical work in any division that has experimental work. Sponsors do not want a multiplier of 2.25 on payroll and benefits for a person who sits at a desk doing computations. Of course, the Divisions take advantage of this by by putting all kinds of programmatic functions for experimental programs in the division overhead in order to spread these costs to theoretical programs and thus subsidize the experimental work at the expense of theoretical work.

Another issue is the "infrastructure" tax. This replaced the Space Charge. Space Charge had a significant benefit in that each group paid for the space that it used and there was some incentive to use less space. Now, with the Infrastructure Tax applied to payroll, the incentive is to grab as much space a possible.

Here are my suggestions. First establish and adhere to the principle that taxes should be applied in a very strong relationship to benefit.

1. Reinstate the space charge. Groups and Programs should pay for the space that they occupy, no more, no less.

2. Each group should have its own specific overhead rate. If a group has a mix of theoretical and experimental activities, then that group should have two (or more) overhead rates. If this is too complex for the LANL accountants, then we should get better accountants.

3. Do not apply taxes (G&A and Division Overhead) to payroll and benefits rates. Instead, institute a flat tax. For instance, the G&A tax should be a fixed dollar amount applied to each employee. After all, the TSM making a salary of $125K uses no more HR resources than the OS-4 making $35K.

4. Labor costs are now charged in $10K bands. With benefits costing about 20% of payroll and a tax multiplier of as much as 2.50 in some cases, giving an employee a raise from $89.5K to $90.5K costs you $30K. Labor costs should be charged in precise proportion to the employees' salaries. Again, if the accountants cannot figure out how to do this, then we should get better accountants.
Matrix management needs to be revisited and probably eliminated. As it operates today we have replicate organizations (line and program) managing and fighting over the same dollars and the same people. Organizations that have merged the two functions, like the old NIS Division that grew to be about 10% of the Laboratory, were so successful that they frightened senior managers who think that success in management is determined by how many boxes they under their command.
The Operations Task Force proposes yet another matrix organization with "Centers of Excellence" AKA HSR, FM, ENG, CFO, SUP, etc. The Line organizations will have to receive "deployed" personnel from these Centers of Excellence to do their work.

So now, the programs have the money, the centers of excellence have the people, and the folks trying to get the work done have to bargain with both sides. This is supposed to make things better.

Cheaper would be a better question. Does Scott Adams work here?
the idea of deployed workers doesn't work. Why should deployed workers have 2 sets of "managers" ? Even after years of being deployed, the worker is still part of the central organization even tho his/her loyalty is to the "host" org. It just doesn't make sense even without considering the costs involved (see above). Keep your centralized orgs for small projects, but as soon as a worker is deployed on a full-time basis, it should be up to the host org to hire that person directly. It's not uncommon for deployed workers to have been at their jobs for more than 5 years but still part of the central org. That's nuts.
On the leadership issue: Several years ago, starting with John Browne, a decision was made (maybe implicitly) that the most important quality of a group or division level manager, was the ability to "get along" with peers. On the surface, this sounds great, but if you think about it, it was implemented in a way that prevents leadership. Leaders who were willing to stand up for what they believe in (isn't that what leadership is all about?) were labelled as "not getting along", and were shifted aside, while "nice" people, who never had the guts to stand up for their beliefs, and who were unwilling to criticize when they saw junk science, became "leaders".

In order to continue a tradition of excellence, we need to have leaders who have the scientific background and the guts to stand up for what they believe in. In addition they need the "emotional intelligence" to do this in ways that are acceptable in our Los Alamos society.
5/9 6:57 has it right on. Advances are made in an open forum where anything and everything can (and should be) critiqued and criticized. This should be done with respect, not abuse!!! With leaders like these, ER would be dead, HR would stand up for us, (you complete the list...).
Post a Comment

<< Home

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