Project Management Institute

Project control

why do budget overruns occur?



Arild Sigurdsen

To most people, project control is about the control of costs, schedules and product quality. These three areas strongly interrelate. Although most of the comments and discussions in this article bear relevance to all three areas, I will concentrate on cost control. The main objective is to show why budget overruns may occur as the result of deficient cost control systems, dilution of responsibility, and the inability to preserve project knowledge.

To control something we need a reference point or a goal. In cost control this goal is the budget. Normally we would arrange the budget according to an appropriate Work Breakdown Structure (WBS) and probably also a purpose-made resource code.

The Cost Control Tools

In fear of something going wrong, projects—especially larger projects—are denied nothing in terms of expensive and sophisticated computerized project control systems. However, some people feel that the huge computer systems require too much time and effort to handle the system as such and thus diverge the attention from the real problems. Recently the trend has been to go for less advanced and less bulky systems. But it seems that some projects develop large budget overruns despite the cost or level of sophistication of the project control tools. I believe that the existing project control tools are not the solution to the problems.


On what basis is the selection of a project control system made? It is doubtful that many people select a project control system based on such trivial factors as screen colors and screen layouts. However, over the years, enormous amounts of money have been spent on non-productive features. In fact, some people have paid extra to have the system so packed with trivial or nice-to-have features that their productivity is actually reduced because the capacity of the system and the speed of operation is diminished.

Does it make any difference which computer system we choose for the job? In general, no. All systems apply more or less the same techniques. Still, some systems are easier to use than others, and it is a good idea

The Cost Control Techniques

The blame for budget overrun is often put on the lack of cost control. In many cases this is undoubtedly true, especially if the blame cannot be put on, for example, extended scope of work or force majeure. But, what is actually meant by cost control? Is it only control during the project execution phase, or does it also include the actual preparation of the control budgets? In discussions that often follow in the wake of disaster projects, many people seem to forget the latter, although there is every reason to believe that the majority of all cost overruns stem from bad capital cost estimates.

From the project's point of view, there is usually no problem pinpointing blame if a pump that was estimated to cost $1,000, for example, actually costs $1,500. However, it is an entirely different matter if workhour-related activities begin to show unhealthy development. A lot of nasty questions will be asked: Are we going in the right direction? Are we doing things in the correct sequence? Is productivity too low? A few changes and additions to the scope of work will confuse the situation even further.

A popular means of illustrating the present situation of the project is by calculating the “earned value.” However, correct calculations of earned value are totally dependent upon correct budgets and the correct evaluation of progress at the time in question. But we know these figures are not absolutely correct because the budget is hardly ever 100 percent correct, nor is the evaluated progress. Everyone who knows about project control knows this. Nevertheless, we act as if all of the numbers are exact.

Very much the focus of attention these days are discussions related to cost risk analysis. For many project people the analysis of risk involvement is not new, but a lot of work is required to make these methods generally accepted. Care must be taken lest we discredit these methods by making them too expert-dependent. A number of unfortunate examples of misuses have recently been discovered, misuses that may mostly be related to the lack of objectivity by the users and occasionally to the (deliberate) misinterpretation of the outcome of the analysis.

Often-heard criticism toward the present cost risk systems are that they are too difficult to understand, too time-consuming to use, and too costly to purchase. As previously mentioned, most systems lack the ability to permit a satisfactory degree of objectivity to be reflected in the analysis. Contrary to some experts in the field, I am convinced that systems that are entirely dependent upon subjective input often cause great harm.

Some people claim that cost control is of no use because “it costs what it costs anyway.” If they are right, it means that the destiny of the project is decided the day the scope of work is defined and the control budget approved. Therefore, whatever happens during the project execution phase has very little bearing on the final result. This view may sound somewhat extreme, but it has much truth in it. More people seem to agree with a proverb that expresses some of the same: “garbage in equals garbage out.”

The Role of the Board of Directors

If it costs what it costs, it means that most budget overruns are the result of bad cost estimates. But who is to blame? The cost estimating department that produced the bad estimate? The technical department that produced the bad input specifications? Or perhaps the board of directors that approved the cost budget?

Is it not true that many boards of directors are staffed by people who, generally speaking, have not the faintest idea about cost estimating or cost control? No, they are undoubtedly among the brightest people in the industry, but they are at the mercy of outside experts.

A common mistake made by boards of directors is that of waiting too long before the owners or, in the case of public projects, the public are informed about actual or potential budget overruns. There is always the hope that an early budget overrun can be corrected later. Hence, the delay.

The fact that the board of directors is not qualified to critically analyze a control budget is serious business, indeed. It means that the project is out of control to start with. It is always very disturbing when a board of directors gives the go-ahead for a major project based on so few key numbers— key numbers derived mainly from the budget estimate. There must be a better way!

The Project Management

Responsibility for the daily running of the project is assigned to a project manager. For obvious reasons, the project control manager works very closely with the project manager. However, particularly for large projects or in larger companies, the chief cost estimator—the person responsible for preparing the control budget—is usually not involved in the execution phase. It is expected that the project manager accepts and approves the control budget, but we all know that this is only a formality. How can he or she accept or approve something as complex as a cost estimate without detailed knowledge? It is the same as the situation faced by the board members.

Today's cost control/project control software does not provide the techniques necessary to prevent devastating and all-to-frequent cost overruns from occurring. It is simply an advance calculation tool that includes a powerful report-writer or report generator. Obviously, these tools make life a lot easier for the project management, but the mentioned rule still stands: garbage in equals garbage out. Fortunately, some project people have a special talent for discovering potential trouble early enough for corrective actions to have an effect. This skill is not totally reliable, but some people are definitely blessed with more of it than others.

Most people who have accomplished success in projects know that there is some luck involved too. This is why very few successful project managers and project control people remain in the job for too long. They often use a successful project to open doors to obtain another job—one less dubious and unpredictable. This is very understandable, especially concerning people who risk being pushed into the darkness if they fail to perform to their superior's satisfaction, and perform every time.

Some Recommended Actions

I believe that to a larger or smaller degree anarchy rules in certain areas of project management. There are no common sets of rules with regard to systematization of the project breakdown structures (WBS), resource codes, cost estimating techniques, cost control techniques. Even within individual companies, it often looks as if standardization is of no importance. The major reason seems to be that professional education is almost non-existent—and no one can decide what a curriculum should be anyway. Consequently, everyone does what they feel is best.

Why this is so? After all, bad project management—in its widest sense—has been blamed for most of the big disasters within the field of capital projects. One would think that at least the project owners would be concerned about reducing the risk of budget overruns.

The reason why nothing is being done is quite simple. To most people there seems to be nothing they can do. Because of the huge number of variables involved in project control, smaller and larger overruns are simply unavoidable. So why spend time and effort on developing project control systems—or join a project management organization—when there are only marginal gains anyway?

Giving in to this type of reasoning can be devastating, especially since it is not true. A lot can be done, and is being done. However, individuals, organizations, and companies are pulling in different directions, thereby making it difficult for outsiders or newcomers to determine what is essential and what is not. What is needed is a major locomotive that can help a sufficient number of competent organizations and individuals move in the same direction. If it takes government organization, then so be it. Below are some of the issues that such a locomotive should attempt to get moving.

First, a proper framework or foundation, in the form of a standard cost coding system, should be established. One research project showed that most companies involved in capital projects do not have an in-house cost coding system. Therefore, the advantage of preparing standardized cost reports and systematization of historic data, for example, is difficult, if not impossible, to organize properly.

Most capital industries can apply the same general cost coding structure. Let us hope that some of the “majors,” or any of the expert organizations, take the initiative to develop a general code that all companies in the industry can apply freely. The international shipbuilding industry appears to have brought this problem closer to a solution than most other capital cost industries.

When calculating costs, project owners have every reason to be suspicious of consultant companies that apply the same installation rate of X-workhours-per-feet-of-pipe no matter how long or short the pipe is. The reason for this is that such cost values are extremely inaccurate since they usually stem from mean or average computations of historic cost data. Numerous cost overruns could have been avoided if simple regression analysis techniques (e.g., the least square regression technique) had been applied instead. There is really no problem to do this; with modern computers, this kind of calculation is easier and faster to perform than any of the “antique” methods.

It is astonishing to me to see how many companies, owners, or contractors still risk their entire future on the quality of publicly available cost data rather than build their own library of data. Off-the-shelf cost data may be necessary to get started, but should be replaced with your own, historical, much more reliable data as soon as possible.

Cost estimating is not, and never will be, based upon exact cost data. The same applies to cost control. We need to focus more on cost risk analysis techniques in all phases of the project, starting with the cost estimates and continuing throughout the project development stage.

Cost risk analysis is more than just adding a fixed risk contingency value of 10 percent to the base estimate, for example. To do only this is definitely asking for trouble.

Likewise concerning project control, traditional project control systems are based upon single-point forecasts rather than probabilistic measures. No wonder we never discover the buildup of disasters until it is too late.

Going back to the subject of the education and training of project personnel, an increasing number of universities and colleges are teaching project management topics. This is fine. Contrary to all the private organizations and companies that teach project management topics, the universities seem to communicate better. For the benefit of the whole industry, we should investigate the possibilities of increasing the communication between the universities and industry. The professional organizations—for example, the Project Management Institute, AACE International, and Internet organizations—have a major role to play in this context.

There is a long way to travel before the largest pitfalls in cost control have been removed. They may never be completely removed. What we nonetheless have every reason to hope for is that the situation today can be drastically improved. Perhaps we may soon see the light at the end of the tunnel. However, to quote another familiar phrase: we must be careful that the light we see is not that of another locomotive traveling in the opposite direction.img


Arild Sigurdsen, Cera Consult, Lier, Norway, is a consultant in project management/controls and related computer applications, who has more than 22 years experience in offshore process construction projects.

This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI or any listed author.

PM Network ● April 1995



Related Content