Saturday, March 17, 2012

4 October 2010 – C. Northcote Parkinson and the cabinet-maker’s art


One of laws of organizational structure formulated by historian and satirist C. Northcote Parkinson in his 1957 treatise (“Parkinson’s Law”) theorizes that there are two kinds of committees: those from which the individual has something to gain, and those to which he has something to contribute.  Of the former type, “the most deeply rooted and luxuriant are those which confer the most power and prestige upon their members.”  To this category he applies the classical term “Cabinets”.

According to Parkinson, the ideal size of a Cabinet is five, because that many decision-makers are “easy to collect and can act with competence, secrecy and speed”.  Parkinson theorizes that Cabinets tend to expand over time, sacrificing these virtues through the addition of new members.  He further predicts that every time a Cabinet grows beyond the point at which it is able to make decisions effectively, it goes through a natural and predictable life cycle, budding a new and smaller cabinet from within its own flesh, in order to reduce the complexity of the decision-making process.

Parkinson founded his thesis on observations of the cycle in British government where, in his words, “the plant has gone through this life cycle five times”.  The process began with the ‘English Council of the Crown’ (the original name of the House of Lords), which, when it was first heard of in the Middle Ages, had somewhere between twenty-nine and fifty members.  It had 60 members in 1601, 140 members in 1661, 220 in 1760, 400 in 1850, 650 in 1911, and 850 in 1952.  Its collective power and influence declined over time, as its size correspondingly grew.  Today its relevance to the governance process is limited at best.

The first “bud” was produced by the English Council in 1257.  Called “The Lords of the King’s Council”, it had fewer than 10 members, enabling it to take decisions effectively.  It had 11 members in 1378, and the same in 1410.  It had 20 members in 1433, 41 in 1504, and 172 by the time the Council stopped meeting.  Within the King’s Council, a second “bud”, the Privy Council stood up, with 9 members.  It  had 20 in 1540, 29 in 1547, and 44 in 1558; 47 in 1679, 67 in 1723, 200 in 1902, and 300 in 1951.  Within the Privy Council, the third “bud”, labelled the Cabinet Council, took over the core decision-making role in 1615, with 8 members.  It had 12 in 1700, and 20 in 1725.  It was superseded in 1740 by the fourth “bud”, Cabinet, with 5 members - Parkinson’s magic number.  The Cabinet had 12 members in 1801, 16 in 1885, 20 in 1900, 23 in 1939, and 24 in 1978.

Why is this interesting?  Because of some recent proposed changes to the internal governance structure of the Department of National Defence.  The most interesting changes, in my view, are the creation of 2 new executive-level committees: the Defence Strategic Executive Committee (DSX) and the Defence Finance Committee (DFC).  Both of these committees are being created to take decisions on matters that were formerly the province of the Defence Management Committee (DMC).  The reason these changes are interesting is because DMC - which is not being done away with - is co-chaired by the Deputy Minister (DM) and the Chief of Defence Staff (CDS), and has 20 members, and 3 attending members.  DSX, by contrast, consists only of the DM and CDS as co-chairs, with the Associate DM and the Vice-Chief of the Defence Staff (VCDS) as the only regular members.  DSX can call L1 leaders and managers and Environmental Chiefs of Staff (ECSs - the heads of the Army, Navy and Air Force) to “attend” on an “as-required” basis.

DFC is virtually identical in structure; it is chaired by the DM (with the CDS present in an “ex officio” capacity), with the Associate DM, the VCDS, and Assistant Deputy Minister (Financial and Corporate Services), or ADM(Fin CS), as the only regular members.  Chief of Programme (C Prog) attends but is not a member, and “appropriate advisors” may be called “on invitation”. 

These two sub-organizations have been created as “executive committees” of DMC (DSX is specifically described as such).  Between them, DSX (with 4 regular members) and DFC (with 4 regular members, plus the “ex officio” CDS) will henceforth make decisions on matters of strategy, defence priorities, the investment plan, and the strategic capability roadmap (DSX); and will make all strategic-level (L0) financial decisions for DND, and serve as the Departmental clearing-house for “resource issues” (DFC).  The controlling function of these executive committees shows that DFC is responsible for furnishing “financial supply” to DMC in response to “requests”, while DSX is responsible for providing “commander’s intent” to DCB, and “leadership” to DMC.

Given that the membership of DMC stands at somewhere between 20 and 25, depending on how one enumerates its 2 co-chairs, 20 members, and 3 attendees, it is fascinating to read Parkinson’s thesis, which, in the 1979 edition of his book, he articulates as follows: “A study of the British example would suggest that the point of ineffectiveness in a cabinet is reached when the total membership exceeds 20 or perhaps 25.” (Parkinson, 181-82).  This is not of course to assert that DMC was or is “ineffective”; those are Parkinson’s words, not mine, and I am merely citing this passage to highlight the fact that Parkinson identified, as the break-point for “budding” of a new successor executive committee, a predecessor committee size of “20 or perhaps 25”.  This is if nothing else a remarkable coincidence.  

Parkinson hastens to add that the study of what he calls “comitology” is by no means sufficiently advanced to entitle us to adduce the ideal number of members in a cabinet; but beyond his initial advocacy of about five members as optimum for decision-making purposes (which both of the new executive committees, in another remarkable coincidence, have settled upon), he notes that, based on the historical examples he has furnished, the number likely lies somewhere between “three (when a quorum is impossible to collect) and approximately twenty-one (when the whole organism begins to perish)…” (Parkinson, 185).  Based on a comparative analysis of the sizes of cabinets of 47 states in 1978, he concludes that the ideal number may be 8, since at that time only placid Iceland had a cabinet of this size.  “Attractive as this theory may seem at first sight”, he goes on to note, “it is open to one serious objection.  Eight was the number preferred by King Charles I for his Committee of States.  And look what happened to him!” (Parkinson, 185).

The application of Parkinson’s Laws to organizational structure is irresistibly fascinating to me, and has led me to write at greater length on the topic.  One of my papers on this subject can be found at Appendix 5, Annex O to the 2010 Report on Transformation of the Department of National Defence and the Canadian Forces.  The annexes haven't been published on-line, but the report itself has.  Copies of the report and annexes can be obtained by contacting




C. Northcote Parkinson, The Law (New York: John Murray Ltd., 1979; Penguin Books, 1981); esp. “Directors and Councils” (175-185).