CHAPTER 7

Law Six: Nine Women Can’t Make a Baby in One Month

Please accept that the terms “man-hour” or any multiple of it, is not intended to be a gender specific term but simply a unit of work; that is a period of time during which a human being, female or male, is engaged on a task. “Person-hours” hasn’t the same alliterative quality and sounds contrived. Also note that the wording of this law is derived from a phrase used in: ­Frederick Brook’s “The mythical man-month, essay 2.” Aug. 1995, Addison-Westley Publishing Company.

There are tasks within some projects whose duration may be shortened by increasing the number of workers engaged on them. If the walls of ten different rooms need to be painted then 10 painters will complete the work faster than 5, but they won’t do it in exactly half the time because there will inevitably be some social interaction and management overhead that will reduce the time engaged in putting paint on the walls. But many tasks within a project are sequential or interactive tasks and, in these cases, doubling the number of workers, as a reaction to falling behind program, is more likely to slow the task than to speed it up. This is because the complexity of coordinating and communicating with a larger group of heterogeneous workers rises exponentially as they are added to a team, the new workers have to be “brought up to speed” which means that both the existing and added workforce are slowed down.

It is not simply the number of workers that fall within this Law it is also inanimate processes. There are many processes that, for various reasons, take as long as they take and throwing resources at them does not change their duration. Many a home owner will have been frustrated in their house extension projects by having to wait for plastered walls to dry out sufficiently for painting or papering to start.

A classic example of a job on building sites that can’t be shortened, even by senior management shouting at it, is concrete curing (hardening) to a condition where work on it can be undertaken. Concrete curing is highly variable and depends on the ambient conditions of temperature and humidity, but a rule of thumb is that “normal” concrete in “normal” temperate conditions reaches its full strength in about 21 days but is hard enough to walk on in 3 days, but that does not mean you can safely drive a fork-lift across it. The curing process relies on the presence of moisture and, since it is an exothermic reaction, loss of moisture through heating up causes a reduction in strength and leads to poorer durability therefore, in large projects, it is a process which may need slowing down rather than speeding up and it takes as long as it takes.

Major civil engineering projects sometimes require a continuous concrete pour phase in the project plan; this requires a supreme level of project coordination and contingency planning, since any significant interruption might be absolutely disastrous to the project. In February 2014, in downtown Los Angeles, a project team started what is claimed to be the world’s largest continuous concrete pour. It involved more than 2,000 truckloads carrying over 16,000 m3 (21,200 cubic yards) of concrete, within under 24 hours, to form the foundation of a 73-story tower. Consider the risks of plant failure, weather or disruptive events and the detailed planning required to pull off such tasks. The 2,000 truckloads involved had only 90 minutes each to get to the site and discharge, or the concrete it contained would have been unusable; 19 pumps pulled the concrete from the trucks into more than a dozen hoses. This was a good example of a project task that was run at the maximum possible capacity in the space available and, once started, could not be speeded up or stopped; once complete the concrete mass had to be kept below a critical temperature for a cure period.

I was once involved in a project where a continuous pour was planned to create the foundations of a very large steam turbine and just as work commenced the Union representing the truck drivers demanded more money to cover working “unsociable hours”—this was a classic case of breaching Law Three, by having to deal with a crisis that could have been avoided with anticipation leading to prior briefing, discussion and negotiation.

The key features of Law 6 are:

• There is no such thing as a standard Man-hour or ­Woman-hour, in spite of that being implied by a project’s Gantt Chart. Doubling the number of man-hours allocated to a task will certainly double the labor costs but only in the special case of isolated subtasks will it halve the duration of the main task.
• Project task duration estimates may be made up of man-hours that range in effectiveness from zero to 100 percent although their unit cost will be the same. Note: effectiveness in this context is envisaged as each hour expended completing its calculated proportion of the task, thus if 10 hours has been allocated to paint a wall, at 100 percent effectiveness each hour sees 10 percent of the work done.
• The duration of some inanimate processes can’t be reduced or condensed within the same program slot because they are spatially incompatible. Floor painting is a classic ­example of a task that can be deeply frustrating when a project is ­running late, and furniture is about to be delivered on time (see Law 8). Turning up the heat in the oven above optimum is not a good way of reducing a pie’s cooking time.
• For IT project managers: unequivocally Adding manpower to a late software project makes it later.

It is worth quoting from Frederick Brook’s essay that is referenced in the first paragraph of this chapter:

Since software construction is inherently a systems effort—an exercise in complex interrelationships—communication effort is great, and it quickly dominates the decrease in individual task time brought about by partitioning [increasing the workers]. Adding more people then lengthens, not shortens, the schedule.

The Man-hour as an Estimating Unit

If you ask an experienced craftsperson to estimate the number of hours he or she alone would need to complete a well-defined and isolated task you could expect to get, not only a reasonably accurate forecast but also, within a contract, a commitment to complete the task within the quoted period. However, in this case the number of hours quoted rarely, if ever, equates to the duration of the job. In the case of jobs carried out by one individual, the duration will depend on the priority given to it over ­competing, concurrent, tasks.

In tasks employing several people there are two variables in the man-hour unit: people and hours; thus, a ten man-hour task of picking apples will take one hour with ten people and 10 hours with one and, depending on the contract agreed, the costs could vary between one and ten.

If you ask an experienced team manager for an estimate of the man-hours required for a well-defined task, the process of estimation will have to be based on the number and capabilities of the human resources ­available in the team and will invariably ignore, in the first iteration, availability of staff, so the estimate may be suffixed with the comment: “but we couldn’t start for 3 months because of current work load.” The project duration has the time of the delayed start added to it.

If the question is changed so that the task is given a fixed end date, the estimation in both cases is much more difficult since the only variables are either increasing the available men or women employed or increasing the hours worked on the task by individuals during each day, week, or month.

Increasing the hours worked over a standard day increases the unit cost of those man-hours due to higher rates paid for overtime or loss of staff given time off in lieu of extra pay. Therefore, a man-hour is not a standard unit of either cost or progress within a project and is only of real use as a cost estimating tool, used early in the process.

Overtime Working

If a man-hour worked during “normal” working hours can be highly variable in its effectiveness it is doubly true of overtime hours because the mental and physical efficiency of all of us varies throughout the 24-hour period. According to a 2008 study by the American Journal of Epidemiology that surveyed 2,214 middle-aged workers, a work week of 50 hours and above has a negative cognitive effect on productivity.

In addition to the problem of weariness induced by long work hours there is a great variation in personal diurnal preference, meaning people range from early risers “larks” to late night “owls.” Most people find that their performance is at peak at specific times of day. This preference is called one’s chronotype and has a marked and well documented effect on behavior and work efficiency. In order to increase the output of certain individuals, I have allowed them, for the period of key project tasks, to change their standard start time from 08:30 to 10:30 with commensurate later finishing times simply because their natural internal clock meant that they were fairly useless during early morning and much more efficient while working late in a quiet, near empty, office. Being of the opposite nature, I frequently chose during periods of peak load to work overtime by making a 05:00 start and finishing at the normal 17:00.

To overcome a project crisis, rather than simply catching up with a program slippage, such as repair of a major plant failure or meeting an unmovable shipping date, stints of unusually long overtime working are sometimes required; but the following warnings should be heeded and its legality under different national regulations must be taken into account.

• In Europe¹ employees only have to work overtime if their contract says so and even if it does, by law, they can’t usually be forced to work more than an average of 48 hours per week, averaged usually over a 17 week period. An employee can agree to work longer, and many members of project teams do, particularly when working away from home, but this must be part of a formal agreement signed by both employer and employee. The 40-hour week at a contracted hourly rate, thereafter overtime paid at an enhanced rate, has general acceptance worldwide.
• Excessive overtime working, let us say anything over 14 hours per day or working double shifts, must be a very short-term expedient with a defined achievable target understood by all involved. Not only does the quality of work inevitably drop off rapidly if extended periods of double shifting are worked, but the danger of accidents, even in an office environment, increases significantly. I know from personal experience that courts of law involved in personal accident claims are very unsympathetic to employers when dealing with industrial accidents, or post-shift vehicular accidents caused, even in part, by tiredness from long working hours.
• Excessive overtime working requires diligent management aware of the dangers and able to enforce regular rest-breaks and suitable nutrition. There are some critical tasks that are not suitable for over-tired brains, these should be identified and dealt with by persons, or at times, that are more suitable.

  On several occasions I had charge of a team engaged in the repair of ship’s engines; on one occasion we had worked continuously, albeit with breaks for food, for 50 hours in a noisy environment. The final task was for me to check the final shaft alignment readings, which with practice is a straightforward task requiring some basic arithmetic and good spatial awareness; on this occasion I could make absolutely no sense of the diagrams and numbers presented to me, although in every other respect I felt reasonably alert. Be aware that long hours can reduce mental function before any obvious bodily ­deterioration.

• Always have a minimum of 3, preferably 4, persons in a team when very long hours are being worked and set clear ­achievable and agreed targets for the work period.
• Recognize that at the end of the period of excessive overtime the staff involved will require time off and team productivity will drop while everyone recovers. It is truly a risky strategy only to be used when a clear and vital project target has to be achieved that can be followed by a week of team recovery.

Bureaucratic Processes are Resistant to Acceleration

There are bureaucratic processes within projects that are highly resistant to being hurried or effected by an increase in numbers of staff employed on them. How do you forecast the number of man-hours required by the nominated person responsible for tasks that involve third-party bureaucratic processes such as obtaining building permits?

One answer is: start very early in the project’s life and, once started, recognize that you probably have to take it at the pace determined by the bureaucratic process, rather than the project program.

Jobs that appear in the project Gantt Chart entitled, for example: Obtain Fire Department Certification or Obtain Planning Permission should be entered and considered as required by rather than a fixed event date because such jobs rely on bureaucratic processes whose end date, or even achievement, cannot be definitively determined but can only be estimated by using precedent.

Sometimes bureaucratic processes can be speeded up by the judicious use of personal contacts in positions of power.

However, if through personal contacts in high places or guile, you believe you are able to short circuit the bureaucratic process remember that it is wise to be very sure that your plan is going to work. It should be considered a “once only policy” since the system has a memory and may wreak revenge next time.

While working on a project in Moscow in the 1970s I was told by a work colleague that, in Russia in the days of the USSR, when people meet for the first time, either in a social or work context, both parties immediately start exploring the question: how can this new contact be used in my favor to obtain an advantage in dealing with the bureaucratic processes?

This is a good policy for a Project Manager to follow throughout his or her career, and it supports the advice a friend gave his daughter on going up to University: “Whatever grade of degree you obtain just ensure that, while at University, you form lifelong friendships with one each of a student lawyer, dentist, and tax accountant.”

Trying to force the pace of any bureaucratic process can, if done in a culturally inappropriate manner, back-fire badly. Being forceful and demanding rarely works. Most bureaucratic processes are handled by, or through, holders of quite uninspiring clerical roles, who, if so minded, can wield wide powers to delay, passively or actively, completion of their part of your task.

The lessons to be learnt are:

• Head-on fights with any bureaucratic process will nearly always turn into a prolonged war of attrition, so are best avoided from the very start.
• Going “over the head” of functionaries is a dangerous tactic which makes enemies, so if you use it be very sure it is going to work.
• It is recommended that logic or commonsense is not used to try to understand some bureaucratic delays—that way can lead to madness—just comply with the requirements of the process.

My daughter was born in a small South Pacific country a short time before I had the opportunity to fly back to the UK on a business trip; we needed her to have a passport quickly and we were told that a full British Passport would take some weeks to obtain with no chance to speed the process. We decided therefore to get her a local passport which would only take a few days, so armed with photographs (at eight hours old), birth certificate and more, I presented myself at the Passport office in the capital and was assured that I would be able to pick up the document in three days. When I got back to my office the next day I had a phone call from a clerk in the Passport office.

Clerk: Mr. Martyr I am processing your daughter’s Passport and I have a question.

Me: OK How can I help you?

Clerk: Your daughter has not signed the application form—is your daughter illiterate?

This is the dangerous point in such conversations and if badly ­handled can cause major delays in the process; keep to the script and do not invoke commonsense or ask such questions as “have you looked at the photograph”?

The correct answer was given: “Yes unfortunately she is unable to write”

Clerk: This is no problem, but we will require her thumb print in two places on her application form.

Me: Thank you for your understanding—I will collect the forms and arrange for her thumb print to be applied today.

We still have the Passport bearing a very small inky smear that could have been made by anything but, in the event, satisfied the bureaucratic process.

“Place of Work” Rules Applied to IT Projects

As a nonspecialist project manager, but one whose multidisciplinary projects have had a substantial software development and commissioning content, I have observed a significant change in the task over the period of 25 years. Increasingly, software routines are configured by using proven subroutines or modules in both control and data analysis programs. There are now highly effective software-based analytic and modeling tools provided to engineers such as MATLAB and Simulink that allow them to assemble highly complex routines using existing blocks, whose contents may be the products of hundreds of man-hours of work.

The more proven blocks of software that exist the more the problems of IT projects are that of understanding the question rather than providing an answer; this is all about law Zero. All this means that progress of the software content of a project is becoming easier for nonspecialists to monitor, but it is still true that there is no such thing as an accrued standard man-hour with which to measure how far advanced is a cerebral, rather than physical, task.

If a task within an IT project is specific to one team member and is best carried out within the individual’s best working hours then, providing team discipline is not adversely affected some latitude should be given. When one is straining everything to recover a slipped project program having a key and trusted employee spend their best working period sitting in a commuting traffic jam, on their way to an office, may not be the best use of resources. However, other than in very special circumstances I do not recommend allowing work to be done in employees’ or subcontractor’s homes because of lack of intellectual and physical security. Allowing an employee to work from home on a project should be the exception and done only when you can define his or her work to a specific task and an agreed number of work-hours.

An avoidable threat to any project is the adoption of a Bring You Own Device (BYOD) policy; when mobile project staff use their own computers. Company issued laptops, while they are still vulnerable to loss or theft, can at least have corporate security systems installed. The company has no such control over data stored on machines where it is shared with private e-mail and social media.

Remember

• Increasing the number of people employed on a task increases the labor cost but may also increase, rather than reduce, the task duration.
• The man-hour is not a standard unit either of work or task duration, it’s an estimating convention.
• Excessive or prolonged overtime working carries risks for safety and efficiency drop-off.
• Bureaucratic processes are resistant to acceleration and should be programmed as “required by.”

¹ The EU’s Working Time Directive (2003/88/EC).