Zambezia (1979), VII (ii).RESEARCH REPORTA NOTE ON COMMITTEE BEHAVIOUR PATTERNSTHE PURPOSE OF a committee agenda is the structuring of the committee'sdeliberations. This report is primarily concerned with the reactions of acommittee to agendas of differing lengths.In 1958, Parkinson' propounded his now famous First Law Š 'Work ex-pands so as to fill the time available for its completion' Š and, with specificreference to committees, his Law of Triviality which stated that the timespent on any item of the agenda would be in inverse proportion to the sumof money involved. Both these propositions seemed so obviously true topeople with experience in management, administration and committee workthat they have since become part of Western folklore.Table IMEETINGS HELD DURING 1977-8Date of Number of Number ofMeeting Agenda Items DecisionsDuration of Minutes perMeeting Agenda Item(in minutes)4/ 8f1/ 9f14/ 929/ 9f6/1027/lOf3/1117/11*24/llf2/125/ If19/ 12/ 2f23/ 2*9/ 3f15/ 3*12/ 4f26/ 44/ 5f1/ 6f19/ 66/ 7fTOTALS: 22MEANS :3424102142513c/f+322710c/f+3123191231283151431831514,3382710214121822311145211242294192042233115,0155100100115451501155155901401301802514510165851001351401602 4451114,54,110,05,411,2752,25,75,611,643,39,525,06,35,05,928,36,69,646,68,87,7f Regular meetings,ŁMeetings held prior to another meeting, i.e. meetings with a fixed end-time.i G. N. Parkinson, Parkinson's Law or the Pursuit of Progress (London, J. Murray,1958).211212 RESEARCH REPORTThis note describes the behaviour of a committee of a Local Authorityduring 1977-8. While there is considerable evidence for the validity of thefirst Law, there is none for the validity of the Law of Triviality.The basic data is set out in Table I* which shows a difference betweenthe total number of agenda items and the total number of decisions (i.e.items on the minutes of the committee). This difference was, in fact, greaterthan is implied by the totals; on 27 October and 24 November, 13 and 10agenda items were carried forward, respectively, to the next meeting.A committee's ability to adjourn, re-convene or cut short a meeting,that is to vary the time available for the completion of its work, would leadone to expect only a sm'all variation in the time spent per item. In practice,if the agenda was short the committee spent an average of 26 minutes peritem; if however, the agenda was long, the average time per item droppedto 4, 5 minutes. It would seem that the committee reacted to a short agendaby perceiving a longer time available per item and talking longer; and thata long agenda produced the opposite result.Table IIAVERAGE TIME PER AGENDA ITEMNumber of Number of Total Number of Total Time Time per ItemAgenda Items Meetings Agenda Items (in minutes) (in minutes)1-5 7 17 440 25,96-10 1 10 100 10,011-15 4 53 465 8,716-20 3 53 455 8,621-25 4 93 510 5.526-30 2 55 320 5^> 30 1 34 155 4,5TOTALS: 22 315 2 445 7,7The possibility of the relationship shown in Table II being a chancerelationship is extremely remote. It is equally unlikely that all the petty itemswere dealt with in meetings with long agendas (or, if one accepts the Lawof Triviality, in meetings with short agendas). The most likely explan&tionis that committee members, knowing that they have a short agenda, sub-consciously talk longer and that the chairman's control varies similarly.The relationship between the number of decisions taken and the timeper decision is similar to that between agenda items and time per item Šthe fewer the decisions, the longer it took to take them. With fewer than 5items on the agenda, a decision took 20 minutes; with more than 30 items,just over 4 minutes.2 The primary source of data is the minutes of one of the standing committees ofa Local Authority between August 1977 and July 1978.J. D. JORDAN213Table IIIAVERAGE TIME PER DECISIONNumber ofDecision)1- 56-1011-1516-2021-2526-30> 30TOTALS:Number ofMeetings713359122Total Number ofDecisions221037571115638331Total Time(in minutes)4401003803507552651552 445Time per Decision(in minutes)20,010,010,36,16,84,74,17,4At its regular meetings, the committee Bad before it the minutes of atleast three other committees. The members could therefore, if they wished,'lengthen their own agenda' by discussing items which appeared on theminutes of the other committees. This accounts, in the main, for the dif-ference between the number of agenda items and the number of decisionstaken by the committee. However, two-thirds of the extra items appear inthe minutes of meetings with an agenda shorter than the mean and onlyone third in the minutes of meetings with an agenda longer than the mean.The figures are set out in Table IV.Table IVEXTRA ITEMSNumber of Number of Total Number of Total NumberDecisions Meetings Agenda Items of Decisons Additional Items1-15> 15TOTALS:12102270222292*96235331261339 f*23 items carried forward from one meeting to the next and appearing twice inTables I and IIf 33 extra items on regular meetings.The ratio of agenda items to decisions is 1,06 in meetings with la longagenda and 1,37 in meetings with a short agenda. If the number of agendaitems can be taken to prognosticate the length of a meeting, and the numberof decisions as a measure of the work actually achieved, Table IV is a goodexample of work expanding to fill the time perceived to be available.The committee did not, however, act in accordance with the Law ofTriviality. There was no correlation between triviality and the time taken, bythe committee, to discuss the item. An item involving ZR$3 000 and anotheritem involving ZR$7,8 million both took about 25 minutes of the committee'stime; an item involving ZR$93 000 took 2 minutes while another involvingZR$74 000 took 35 minutes. Despite the general belief that small sumsgenerate more discussion than large sums, it is more probable that thedetermining factor is the level of understanding of the committee members.A well understood subject will generate more discussion th&n a subjectwhich is not understood.214 RESEARCH REPORTThe behaviour pattern of another of the standing committees was similar.An analysis of the agenda length and the time per agenda item for its regularmeetings during 1977-8 revealed an average agenda of 46 items with anaverage time per item of 4 minutes. However, if the agenda reached 55 itemsthe time per item fell to 3 minutes; and if the agenda only had 30 items,the time per item was 5 minutes. Although less marked than in the behaviourof the first committee (with an average agenda only one third as long) theassociation between short agendas and longer debate was present. Theopportunity to discuss extra items Was not available to the second committee.The figures presented in this note show that committee debate doesexpand to fill the perceived time available, but it should not be presumedthat this is undesirable. A sense of urgency occasioned by a long agenda canlead to incisive debate and decision-taking. On the other hand, a lack ofurgency occasioned by a short agenda can lead to deeper consideration.University of Rhodesia ]. D. JORDAN