ESM 609 Module 7
Spring 2006 Closure

(Prepared by Greg Kinney)

CLEAREST ITEMS

• I knew budgeting could be a complex issue, however the book did a good job of demonstrated how complex it could actually be when dealing with multiple levels of management. It was tasking when I had to create a budget for my superior and it was only him and me.
• The most useful thing in this chapter is to learn all the “lingo” and acronyms related to the subject of project budgeting. This at least might be helpful for an engineer to identify the type of budgeting process the firm he works for is using. 
• The clearest part of chapter seven was the explanation of different philosophies and methods of budget formulation.  In particular, the differences and benefits of top-down budgeting and bottom-up budgeting was clear and concise.  I thought it was neat to see how well chapter seven tied in with some of the previous chapters (especially chapter six).  You have to be a real “people” person to get things accomplished when you’re a project manager. 

• I ANSWERED THIS ONE as follows:  My perspective: I wouldn’t necessarily say you have to be a “people” person, commonly called an “extrovert” to be an effective PM.  Some of the best I’ve known were really quite introverted, though they could certainly be pushy.  They often exhibit quiet leadership of a very noncharismatic kind.  But the best are authentic, they are by and large smart, they have a bottom line orientation and an ability to follow through.  This is really an extension of the discussion with a fellow student a few weeks ago, in Module 3 (see the Spring 2006 Closure tab in that module) in which I laid out some of the great observations made by Peter Drucker on how charisma is overrated and how it is unrelated to, or even undermines, effectiveness.  See that discussion.

• Pretty much the whole chapter was interesting; probably useful also. What I get caught by is the mathematical treatments: Learning curve and templates using MAD and TS.
• Again, I enjoy the looks into the ‘other side of life’.  Government budgeting is an oxymoron.

MUDDIEST ITEMS
There were two questions that tackled very similar themes, so I will put them together—

As for G&A, I found a pretty good definition of general and administrative costs on a web search (see http://www.maxwideman.com/pmglossary/PMG_G00.htm#General%20and%20Administrative%20Costs ).   The definition found there is as follows:
“Expenses for the general management and administration of the business unit. For example, includes insurance, legal fees, bid and proposal, independent research and development, and interest expenses.”
You can see the problem based on the definition.  These costs are rolled up into business expense.  In a big enough company, it might be assigned to the corporate “service unit” or “business unit” (i.e., division).  Or there might be some spread among different divisions.  Either way, G&A expenses don’t get identified as a line item.  They are identified through after the fact analysis.

Also, different companies define G&A expenses in different ways.  At Alyeska Pipeline, I have seen some of the third party contract costs included in overall G&A.  Currently there is a great deal of executive interest in how our G&A compares with those of comparable companies, which is a legitimate interest if you’re trying to remain competitive.  What is critical, however, is to assure that our definition of G&A, which governs how we assemble our overall G&A costs, matches the definition of companies we’re benchmarking to.  This can be a real challenge because no two businesses are the same.

The answer isn’t complete without some discussion of the distinction (if any) between indirects (which first requires definition of direct costs) and overheads.  I’ve had to do some searching myself to come up with generally accepted definitions.  One website I found (http://uwadmnweb.uwyo.edu/sbir/comptips/understandcat.html ) defined direct and indirect costs as follows: 

What is a direct cost? The Federal Acquisition Regulations (FARs) states that "A direct cost is any cost that can be identified specifically with a particular final cost objective (or project)& " Conversely , the FARs define indirect costs to be "& any cost not directly identified with a single, final cost objective (or project), but identified with two or more final cost objectives & It is not subject to treatment as a direct cost."

This site also defines indirect cost groups as follows (note that this is built around a federal program, known as the Small Business Innovative Research or SBIR program, but the observations are generally valid):

The most common indirect cost groups are "Overhead" and "General and Administrative" (G&A) costs. Overhead costs are primarily those indirect costs associated with the conduct of the business (manufacturing, R&D, etc.). G&A costs are those primarily associated with the "general management and administration" of the business. Overhead cost examples could include costs for laboratories used for multiple projects, such as rent, utilities, insurance, supplies, expendable equipment and depreciation of laboratory capital equipment. Usually, labor is a significant overhead cost, and covers employee's time not chargeable to a specific "direct" project. Employee fringe benefits (e.g., paid leave time, health insurance, payroll taxes and "retirement" contributions) for "operations" employees usually are significant overhead costs also. G&A costs could include, for example, accounting, contract management, security, human resources, and overall business management. Labor and associated fringe benefit costs, and facility costs for G&A personnel, usually are important indirect costs. The Federal Acquisition Regulations (FAR) also allows Bid and Proposal (B&P) and Internal Research and Development (IR&D) expenses to be accumulated as G&A costs.
The FAR allow flexibility in establishing groups or categories for collecting indirect costs. The Federal Acquisition Regulations (FAR, 31-201b) states that: "Indirect costs shall be accumulated by LOGICAL (emphasis ours) cost groupings with due consideration of the REASONS for incurring such costs." It is important to have a rationale consistent with the FAR for accumulating indirect costs, and for designating costs as direct or indirect. The rationale should be applied uniformly to all projects. As noted earlier, overhead and G&A are common indirect cost groups used by small businesses engaged in SBIR contract/grant activity. However an acceptable practice used by some new small businesses is to have only one indirect cost group, namely overhead.

What this shows is that there is some flexibility in how indirects are defined and grouped, which perhaps adds to rather than clears up our confusion.  I can’t solve the problem, but hopefully I’ve pointed out some of the ways this is defined and group in practice, and the fact that there is some “fuzz” in application.

Q:  The question asking how to avoid letting budget planning from becoming a game is unclear. I never saw in the text where the term “game” is referenced. I answered the question assuming this is an interoffice situation where more than one person is bidding or planning the same project and time and money is being wasted. 

A:  Check out the top of page 332, where the term “zero-sum game” is used. This term comes from game theory, and refers to situations where anything gained by one party is lost by another party.  This situation is often present, particularly in office politics situations where there is competition between different departments for a piece of a fixed budget.  In a larger sense, the term “game” is generally defined as any situation where there are two entities (or “players”) in competition, where there is not a perfect understanding between the players of the intentions and future actions of the competition.  Budgeting provides an excellent example of gaming.  The people at the bottom have often learned that upper management is fond of 10% across the board cuts, “challenge” budgets and the like.  They adapt through first-strike approaches, inflating their estimates in advance to compensate.  Then upper management adopts other tactics, including punishing overestimates, and so on.  The point is that when you have that type of gaming going on, there really isn’t a whole lot of collaboration going on.  It is organizational illness.

Q:  The muddiest part of chapter seven could be tied to project three as well.  I understand the philosophy but the practical application of what I read is more difficult.  I know that project three is intended to facilitate the practical side of the chapter. I’m hoping I can get a better understanding of the details surrounding activity budgeting versus program budgeting.  As I read through chapter seven, I also get a little confused when they talk about learning curves.  I understand the logic but things get hazy during the formulation side.  I think it gets clearer after I read it a couple of times.  It takes time to really assimilate the information.  In any event, I do see the importance of learning effects in the budget estimation process. 
A:  The book offered an opportunity to do some calculations estimating learning curve effects, but we didn’t really go there.  This is an abstraction that may be difficult to defend in the budgeting process, but it is defensible at least on a conceptual level.  Perhaps in future courses, we should consider doing an exercise or two on this.

Q:  This chapter has reinforced the theory that budget data from the organization’s past projects and industry cost data are useful in forecasting budgets on short-duration projects. Also important is the distinction made on estimation for projects with multi-year life cycles. Here, traditional methods are not as valuable since probability of inflation, new construction alternatives, and availability of personnel may severely impact original estimates, especially here in Alaska.  However, estimation at many local construction companies that I’ve worked with seem to rely solely on labor cost data and supplier material quotes. What specific tools can be used to improve this die-hard method, short of allotting a higher percentage for contingencies or praying to the change-order gods?
A: You’ve asked a great question and I’ve had to think awhile to come up with a good answer to this.  Part of the problem is that I can’t in good conscience recommend tools that are inappropriate to the users.  The fact is that many contractors at the local level are relatively small businesses, headed and staffed by people who got there from the construction trades (just like me, before I went on to college).  The staff is usually pretty small, and sometimes the staffers get training and education in more advanced methods, but more often they don’t.  Their advantage is they are close to the work; those who are successful begin to develop a feel for how to hedge and spread risks, how to “load” their bids, and they learn a lot about site and job conditions that can heavily impact profitability.  This is an outcome of the school of hard knocks, and I would never argue that planning tools and software take the place of that.  The problem is that most who try just don’t survive the process.  So, what can they do to improve their estimates?

Many of the people who have great practical experience have real limitations in application of mathematical and modeling approaches.  Their best bet is to hire or partner with people who have more talent than they do at that kind of thing.  With your experience and education, you will be in an excellent position to help.  You still need labor cost data and material quotes, but as you know, there is a great deal of additional information you’ll need, and you can help fill in the gaps. 

In general, I believe that you need to begin with a work breakdown structure before you ever get to the labor rates and material costs.  I recommend that you begin with a deterministic approach – assign a single value to estimated duration of work, estimated labor hours or what not.  Depending on the kind of work you’re doing, you can use references such as the Means Construction Data series to estimate the task durations, which can be rolled up in your WBS.  (Of course, there are some things that don’t lend themselves to that kind of “parametric” approach: for earthwork jobs, for example, you have to have site knowledge of where the borrow pits are, what the distance is to the job site, what the material price is, etc.)  You have to build in some level of “fudge” factors like Alaskan cold weather performance, overall efficiencies (allowing slack for breakdowns, bathroom breaks, etc.) and the like – but not too much, or you will lose the bid.  Once you get the overall durations based on the WBS – and the material takeoff – you need to get information on quotes, delivery time and costs, etc.

After you get the deterministic estimate – and assuming you still have time before the bid due date – then you have an opportunity to bring in a more advanced approach.  With your WBS set up in a project management program (Microsoft Project or something superior like SureTrak), you can start to play with durations (minimum/expected/maximum), which lends itself to a ADM or PERT approach as in one of the Module 9 homework problems.  You can also do Crystal Ball type modeling on all of the labor elements to get you to the overall cost spread, which should factor into your contingency for the job and, indeed, into your decision as to whether or not this is a good bid.  Along the way, you can also play with inflation factors, especially in a job of longer duration.  My advice is not to do this in isolation, but to try to involve others in it.  This may be tough at first, since you’ll have to show them what your approach does and how it can be useful. 

When it comes to project bids, you are up against the competition and you cannot know their thinking.  You can only control your own perspective on costs and risks.  But you can’t help but be affected by what bids are successful and by your intelligence on whether the successful bids become profitable projects.  Try to learn all you can about successes and failures both inside and outside your business. 

As a final comment, what determines success and profitability is likely to be something, or some group of factors, that just can’t be captured by the useful estimating and scheduling tools we’re using in this course.  The “X” factor is, I would argue, almost always provided by the talent and imagination of the project team, especially but not exclusively the leaders.   It’s often some very innovative idea that makes project execution much more efficient and provides the competitive advantage needed to win bids and make money.  Sometimes it is very good planning and personnel management.  Sometimes it is a combination.  The point is that your estimating tools will never capture or replace that “X” factor, but should be designed to complement those and to help manage risks.

Q:  Muddiest thing: Biased estimates. How to identify a biased estimate. How to avoid them in the first place. 
A:  Bias is formed by a combination of two things: experience base and fundamental attitude.  I would argue we all have biases of one kind or another.  It is just a part of the human condition, and is neither good nor bad except as it manifests itself in destructive ways (as in biases against other ethnic groups, for instances).   In those destructive cases, values-related issues arise.

The kind of bias we’re describing here has to do with how aggressive we are in our approach to cost risks.  This has something to do with our fundamental attitudes toward risk in general, whether we are conservative (risk-averse) or aggressive (risk-taking) as it relates to cost issues.  These attitudes can and do evolve over time with experience.  But this chapter is not really trying to get into anyone’s psyche and to help with those attitudes.  Rather, the authors show an approach (Section 7.2) intended to improve the cost estimation process, which you might recognize as being derived from Statistical Process Control (SPC) procedures.  We didn’t do any homework to illustrate the use of this, but it likely is useful in helping estimators recognize their biases and begin to converge their estimates toward actual costs. 

The ideal is to have no bias, but I doubt it’s achievable.  The consequences of underbudgeting can be disastrous, so most estimates will want to bias estimates high for that reason.  There is also risk – there are variations in what the actual costs turn out to be – so you are probably wise to incorporate a bias.  You just want to make sure that your biases are within bounds, to improve your own craft and to improve the success of your enterprise.

Comment:  I guess I didn’t like this chapter because it seemed only to touch the high points on budgeting, I understand the next chapter will get into the meat and potatoes but this almost seemed like a preview.