UNDER THE MICROSCOPE:PARKINSON'S LAW – "Work expands so as to fill the time available for its completion" – was coined in 1958 by C Northgate Parkinson, a naval historian, theorist of bureaucracy and humorist. This "law" is now part of mainstream thinking. When you first hear it you immediately think that it has a ring of truth. Parkinson's Law has now stood the rigorous test of mathematical analysis, as reported by Mark Buchanan in New Scientist (January 10th, 2009).
Buchanan describes how Parkinson originally got his idea. He was working as Major Parkinson in 1944 at the joint army/navy headquarters in Britain and he was being deluged by paperwork from above. The chief of base went on leave, his deputy fell sick and Major Parkinson was left to carry on alone. The flood of paperwork suddenly stopped but otherwise things went on smoothly. Parkinson later mused: “There had never been anything to do. We’d just been making work for each other”. In 1958 Parkinson wrote a satirical book on human behaviour – Parkinson’s Law: The Pursuit of Progress.
A powerful example quoted by Parkinson to illustrate his law concerns the civil service. Civil servant A feels overwhelmed by work (the feeling may or may not be illusory). He has three options; he can resign, share the work with colleagues at his level, or appoint two juniors. If he resigns he must accept a reduced pension. If he shares the work with someone at his own level that person will become a powerful rival for promotion. So he appoints two juniors, B and C. If he appointed only one junior, the subordinate would quickly assume almost equal status with A in their own minds and would therefore be a rival. Better to have the junior competing with another junior. Pretty soon, either B or C will feel overworked and two more subordinates will be appointed. Five people are now doing the work of one and all are busy bouncing e-mails, phone calls and faxes off each other and passing correspondence up and down the line. The take-home message is that adding to staff in any large hierarchial organisation doesn’t necessarily get more work done.
Parkinson backed up his idea with historical research. He noted that between 1914 and 1928 the number of administrators in the British admiralty rose by almost 80 per cent while the number of sailors fell by one-third and the number of ships by two-thirds. Parkinson concluded that in any hierarchical management structure, people in authority need subordinates and those extra bodies have to be kept occupied regardless of how much there actually is to do.
But is Parkinson’s Law a real law or just a cynical slogan with a ring of truth? Physicists at the Medical University of Vienna have mathematically recreated the bureaucratic dynamics described qualitatively by Parkinson and have mathematically confirmed his predictions. The researchers, P Klimek, R Hanel and S Thurner (Cornell University Library, August 12th, 2008), devised equations using quantities such as promotion and drop-out rates in a hierarchical body and demonstrated the conditions that breed ever greater bureaucracy – high probability of promotion coupled with the hiring of subordinates.
Parkinson proposed a number of variations on his idea. One was that the amount of time taken to reach a decision on a capital project is inversely proportional to the cost of the project. His reasoned that that when costs greatly exceed the incomes of the decision-makers they switch off – what’s another million here or there. But when costs are closer to amounts they are used to in their own lives they are more willing to ponder closely over spending a few thousand pounds.
Parkinson also thought about committees – how many members can a committee have and still remain effective? He argued that committees are unable to reach consensus when they have more than 20 members. Experience seems to have taught this lesson in practice. For example, the highest executive bodies, including cabinets, of most countries have between 13 and 20 members.
Klimek and colleagues analysed committee size using a simple network model and confirmed that the magic upper number is about 20 if the committee hopes to achieve consensus. However there is one number less than 20 that is problematic – number eight. Computer simulations show that a committee of eight people has a high probability of deadlock. The reasons for this are unclear, but Parkinson anticipated this result, noting that no nation had a cabinet of eight members. This remains true today and most committees charged with making important decisions tend to have a membership that falls on either side of eight. Buchanan reminds us that we ignore Parkinson at our peril. King Charles I had a Council of State of eight members. He eventually lost his head because of bad decision-making.
William Reville is associate professor of biochemistry and public awareness of science officer at UCC – http:\\understandingscience.ucc.ie
