Proportional Representation Society of Australia
	Tel. +61429176725	www.prsa.org.au
	info@prsa.org.au	2013-07-14

See the PDF version here.

The following copy of the historically important, and very clear and informative 30-page paper

"DEMOCRATIC REPRESENTATION UNDER THE HARE-CLARK SYSTEM - The Need for Seven-Member Electorates",

by the late Dr George Howatt, M.A, Ph.D, of the University of Tasmania was digitized by the Proportional Representation Society of Australia.

The statement below on this page appeared on a note inserted into reprints of this Parliamentary Paper.

"This brochure was tabled in the Tasmanian House of Assembly in May, 1958, to provide background for the discussion on whether the electorates under the Hare-Clark system should return six or seven members. Subsequently it was reprinted, with different pagination, as Parliamentary Paper No. 22 of 1958.

A bill to provide for seven-member electorates passed the House of Assembly on May 29, 1958, and the Legislative Council on December 3, 1958. It became Act No. 91 of 1958 and was reserved for Royal Assent on January 9, 1959. The first election using seven-member electorates was held on May 2, 1959.

An appraisal of the change to seven-member electorates, based on the results of this election, is contained in an article, entitled "Seven-Seat Electorate Defended", in "The Mercury", Hobart, of June 4, 1959.

Proposals for further Parliamentary reforms for the House of Assembly are contained in a monograph by the writer, entitled Fixing Responsibility of Governing (Hobart: Government Printer, 1960, 91 pages): this was reprinted as Parliamentary Paper No. 17 of 1960 (Tas.).

During the reprinting of the attached, slight alterations were made to improve the layout of some tables and to make typographical and editorial corrections.

G.H.

Hobart.

August, 1962"

1958. (No. 22.)

PARLIAMENT OF TASMANIA.

________________

DEMOCRATIC REPRESENTATION

UNDER THE

HARE-CLARK SYSTEM

The Need for Seven-Member Electorates

Paper by

George Howatt, M.A.

(University of Pennsylvania)

____________________

Presented to the House of Assembly by His Excellency's Command.

____________________

TASMANIA:

L. G. SHEA, GOVERNMENT PRINTER, HOBART.

_____

1958.

(No. 22.) 1958.

2

Acknowledgments

To Professors H. J. Abraham and W. A. Townsley, my academic advisers, respectively, of the University of Pennsylvania and the University of Tasmania.

To Messrs. H. J. Corrigan and A. J. Mazengarb, Assistant Chief Electoral Officer, Tasmania, and Commonwealth Divisional Returning Officer for Denison, respectively, who have given so generously of their time, their advice, and their exceptional knowledge of the Hare-Clark system.

To the University of Tasmania for a research grant under which this paper has been written.

To the many kind Tasmanians who have contributed to my under-standing on electoral matters and have helped to make this stay in Tasmania so enjoyable.

GEORGE HOWATT,

Hobart, May, 1958.

1958. (No. 22.)

3

FOREWORD.

Tasmania, the first community in the world to use the Hare system of proportional representation under adult franchise, has had a far longer experience with this method than any other place. First adopted in 1896 for electing the Hobart and Launceston members of the House of Assembly, the general features of the system were re-introduced after a lapse of a few years, for the election of all members of this House by the Electoral Act of 1907.

The "Hare " of Hare-Clark refers to an English barrister, Thomas Hare, who in 1857 laid down the chief features of the single transferable vote system of proportional representation. Mr. Justice A. I. Clark of Tasmania, when a member of the House of Assembly and Attorney-General made some original contributions to Hare's general principles. Because of this and Justice Clark's efforts in securing the adoption of the system in 1896, the method later became known as the Hare-Clark system.

Tasmania's special experience with this system is relatively unknown in most places of the world, and certain aspects of the Hare-Clark method need elucidation even in Tasmania, for example the subject of this paper, which discusses why the election of seven members per electorate instead of six is needed for the proper functioning of the Hare-Clark system. This paper contains some of the material which its writer gave to the house of Assembly Select Committee on Electoral Reform last year.

Mr. Howatt was a lecturer in political science at Lehigh University, U.S.A., before coming to Australia on a Fulbright scholarship to study election systems here. He is a specialist in this field and has written an M.A. thesis on proportional representation in American city elections and many articles on electoral subjects.

I commend this study to all members of Parliament, not only in Tasmania but in Mainland States, and to students of Political Science everywhere.

W. A. TOWNSLEY,

Professor of Political Science,

University of Tasmania.

(No. 22.) 1958.

4

TABLE OF CONTENTS.

	Page
Acknowledgments ………………………………………………………….	2
Foreword ……………………………………………………………………	3
Part I. A Potentially Model System with One Defect …………………	5
Part II. The Overall Merit of Hare-Clark is Unquestioned …………..	6
Advantages of the Hare-Clark System Summarised … … …	7
Objections to Hare-Clark …………………………………………	15
Part III. Why Six-Member Electorates are Faulty ………………………	16
Summary of Part III. ………………………………………………..	19
Part IV. Why the Solution Depends on Seven-Member Electorates ….	20
Three-Member Electorates ………………………………………	20
Five-Member Electorates ……………………………………….	21
Practical Considerations ………………………………………..	22
Application of Seven-Member Quota to Past Elections … …	23
Independents and Minor Party Candidates ………………….	26
Part V. Other Considerations …………………………………………….	26
Hare-Clark is Impartial Politically ……………………………..	26
Seven-Member Electorates could Reduce Cost of Government	27

(No. 22.)

5

Part I ^(a).

A POTENTIALLY MODEL SYSTEM WITH ONE

DEFECT.

In the past 100 years, during which the advance of universal suffrage has made its greatest strides, there has been, unfortunately, a steady decline in the promise once held for democratic government. Some thinkers now question even the premises of democracy; others are looking for specific ways of improving democratic forms and practices in the hope of preserving the philosophy of "government of the people, by the people, and for the people".

In particular there is widespread dissatisfaction with the election methods and practices existing in most places of the world. The limited number of persons who know of the Hare-Clark system and endorse its philosophy of representation would say that many of the shortcomings ascribed to democracy in general really arise from defects in electoral systems. Before we condemn democracy or despair of its value, they would say, we should give it a chance to demonstrate its capabilities by installing electoral systems which adequately reflect the wishes of the electors. Most election systems in the world today distort, suppress, or even reverse the wishes of the voters. With this misrepresentation, moreover, go many accompanying evils. Faulty electoral methods such as single-member electorate systems expose democracy to inherent weaknesses while preventing democracy from exhibiting its strengths.

The alleged failure of democracy, these supporters of basic electoral reform would say, is -not that democracy has been tried and found wanting, but that it has not been tried-except in a very few places, like Tasmania. In this State, thanks to fifty years' experience with the Tasmanian-pioneered Hare-Clark system of proportional representation, election results have mirrored the wishes of the people, and Tasmanians have been spared many of the tribulations caused by faulty electoral systems in use elsewhere. Moreover, even although the Australian mainland State electoral systems are distinctly inferior to Tasmania's Hare-Clark system, the general standard of Australian State and Federal electoral methods is so high by world standards that the general public in Tasmania is unaware of the seriously deficient methods used elsewhere in the world.

To the other States and countries which are suffering from faulty electoral methods, and many are suffering critically, the Hare-Clark system could serve as a model; but one grave existing defect must first be corrected.

One requirement of a truly democratic election system is that a majority of the electors should be assured of the right to return a majority of elected representatives. Only an electoral system like Hare-Clark can give this guarantee, and even it can do so only if an odd, rather than an even, number of members is chosen from each electorate.

In some circumstances the need for an odd number of representatives becomes more apparent than in others. When the chief contestants for all seats are candidates of two political parties of nearly equal strength, the need is most apparent. This is the situation which prevails in Tasmania today. When the Hare-Clark system was adopted in 1907 the present two-party system did not exist in Tasmania. Perhaps this could be one reason why an error was made in determining the electorate arrangements. The decision to have a thirty-member House and at the same time to utilise the five Federal electorate boundaries, meant that each Federal-State electoral division had to return six members.

Experience soon showed that it was a mistake to choose an even number of members from an electorate. In recent years the defect has become even more

As early as 1912 the mistake was widely commented upon ^(b). In recent years the_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) This paper is written primarily for well-informed Tasmanians, of whom some knowledge of the Hare-Clark system and Australian electoral methods is assumed.

(b) For example: On pages 4 and 5 of the Report on General Elections, 1912, Parliamentary Paper No. 11 of 1912.

(No. 22.) 1958.

6

apparent, resulting in the anomalous situation of equal party representation in the House of Assembly despite unequal voting strength in the country. For a party to receive a majority of four of the six seats in any electorate, a 58 per cent vote is required. This percentage is higher than either of the two parties can usually achieve in a large, multi-member electorate; consequently both parties have generally returned, and can be expected to return, three members each per electorate despite, unequal support among the electors. While this anomaly has caused severe criticism, the fault really does not lie in the Hare-Clark system itself but in the manner in which it is applied. Hence damage to the reputation of the Hare-Clark system and instability in Parliament will continue unless provision is made for an odd number of seats in each electorate.

The purpose of this pamphlet is to point out why the proper functioning of the Hare-Clark electoral mechanism requires the election of an odd number of members and to show why seven members instead of six should be chosen from each of the existing electorates.

The Hare-Clark system cannot serve as the model which the world needs until a change to seven-member electorates is made ^(a).

This change has been recommended by a Select Committee of the House of Assembly which last year after a careful study of numerous alternatives re-endorsed the Hare-Clark system. The Committee's conclusions were supported in unequivocal and forceful words by "The Mercury," which over a half-century ago was a leading force in the original battle for the establishment of the Hare-Clark system. In an editorial of October 30, 1957, it said -

" . . . The Hare-Clark system is mathematically the most efficient, and politically the fairest method of election. It has been subjected to considerable criticism in recent years because it apparently threatened to put the Assembly in perpetual deadlock, with 15 members on each side. But the system itself was never at fault; the method of using it was wrong."

"It is most gratifying that the committee, after careful comparisons with other electoral methods, has thoroughly vindicated the Hare-Clark system. In its view it is the best method of parliamentary election in the world - and this is no exaggeration . . ."

PART II.

THE OVERALL MERIT OF HARE-CLARK IS

UNQUESTIONED.

Before examining the nature of the defect of the six-member quota and the reasons why the election of seven is required, lot us first consider the overall value of the Hare-Clark method of representation and election. Despite grave weaknesses inherent in six-member divisions, one should not lose sight of the extraordinary advantages of Hare-Clark voting in general. As Tasmanians have been so long accustomed to the benefits of a uniquely superior electoral system they are likely to take its advantages for granted and fail to realise the dangers and defects of other electoral methods.

Moreover, the key importance of the electoral method to democratic society must not be overlooked, for it is the features of the electoral system in vogue which determines to a large extent the type of person chosen to office. Those elected become at once our law makers, the official spokesmen for our society, and the men responsible for the administration of public policy.

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) Experience with seven-member electorates may show that further refinements can be made to the Hare-Clark system, as the last thought on electoral matters has not of course been formed. In fact, very little study or writing has been devoted so far to the science of elections. One reason for this lack is, of course, that electoral methods in the countries with single-member systems or first-past-the-post voting are so slapdash and crude that there is not much which is worthy of, or lends itself to, scientific study.

1958. (No. 22)

7

Although the following summary calls attention to numerous reasons for the superiority of the Hare-Clark method, the listing is not at all complete ^(a). Moreover, this list compares Hare-Clark only with the single-member electorate system, which has been the only alternative seriously suggested for choosing the Tasmanian House of Assembly. It is hoped, however, that enough of the advantages of Hare-Clark are reviewed to show that this system is incomparably more satisfactory than single-seat electorates in serving the purposes of elections and in contributing to the improved functioning of democratic society. If one believes in democratic representation, it would appear that the advantages of Hare-Clark are a matter of demonstrated fact rather than merely of opinion.

Advantages of the Hare-Clark System

Summarised.

A. THE HARE-CLARK SYSTEM PROVIDES FOR THE FAIR AND ACCURATE REPRESENTATION OF THE VOTERS. SINGLE-MEMBER ELECTORATES DO NOT.

1. HARE-CLARK ASSURES ^(b) THAT THE WISHES OF THE ELECTORS ARE FAITHFULLY REFLECTED IN THE ELECTION OF MEMBERS OF THE HOUSE OF ASSEMBLY.

On this major criterion of a democratic election system - whether it succeeds in reflecting, not distorting nor reversing the wishes of the electors - Hare-Clark succeeds pre-eminently and single-member electorates fail disastrously.

It is normal in elections conducted under single-member constituencies for nearly half of the " electors " to vote for defeated candidates. In these circumstances, in other words, nearly half of the voters are only " would be " electors, and are " represented " in Parliament by members whom they voted against and do not want. Under Hare-Clark close to three-fourths of the voters see their first choice elected to Parliament, and the votes of most of the remaining electors are effective in electing a candidate high in the elector's preference and within the party favoured by the elector.

The Hare-Clark system is purposely designed to give effect to the wishes of as many electors as is technically possible. In contrast, single electorates normally cause, in fact, the wastage of nearly half the votes and, in effect, therefore the disfranchisement of nearly half the voters.

This failure inherent in single-member electorates is not, moreover, just an abstract theoretical fault; it directly gives rise to practical consequences of gravest magnitude, as noted in the sections which follow.

2. HARE-CLARK CAN NORMALLY BE EXPECTED TO PREVENT A GOVERNMENT FROM BEING ELECTED BY A MINORITY OF THE VOTERS.

Too often single-member electorates exemplify the phenomenon of the democratic process in reverse-conducting a poll and then giving victory to the side with the smaller total vote. Frequent examples from all over the globe show that single-member electorates do allow a party with a minority of the votes in the country to obtain a majority of the seats in the legislature, while its opponent with a larger overall vote receives only a minority of the seats.

A case in point is furnished by the Union of South Africa, where in the general elections for Parliament in 1948 and 1953 the present Nationalist Party Government obtained a huge majority of seats with a total vote much smaller than that given by the electors of the United Party Opposition. The party, therefore, which has been ruling South Africa since 1948 would not have been in Government, but in Opposition, if an electoral system

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) A comprehensive and documented analysis of the advantages and disadvantages of the Hare-Clark system, with consideration of its less tangible benefits and deeper political effects, is to be included in another study of the Hare-Clark system now under preparation by the writer. The dynamics of the operation of Hare-Clark and the less obvious reasons for its success are of course too lengthy for discussion in this paper.

(b) Subject to some exceptions resulting from use of the present six-member electorates.

(No. 22) 1958.

8

reflecting the wishes of the people had been in effect. The "winning" party in those elections was not the majority choice of the voters of that country. Instead it represented only a minority of the electors, to whom the single-member system often gives a majority of the seats.

Variation in size of electorates, gerrymandering, and the first-past-the-post system of voting sometimes make it possible for the minority to govern the majority. None of these contributing conditions, however, was the cause in South Africa. The majority of the electors did not obtain a majority of the seats simply because of an inherent weakness of the single-member system: namely, the necessary cutting-up of the country arbitrarily into a multitude of small divisions which return only one member each. The election of only one member per division frequently means that one party, by winning its seats with larger majorities than its opponents, obtains a greater total vote but loses to the less popular party. This happens because the latter, though obtaining smaller majorities, has more instances of them because its vote, although being thinner, is spread more widely.

If subdividing a country into single-seat electorates happens to result in a distribution which is a fair cross-section of political opinion, then the seats won by the respective parties may be in accordance with their popular strength. Sometimes this does occur under single-member systems, but only if the number of safe seats for one party is cancelled by an equal number of similarly safe seats for the other party, with the remaining seats being "swingers".

When these particular circumstances arise and act to compensate for the crudeness and limitations of single-seat electorates, this "system" may sometimes give the appearance of affording representation to parties in accordance with public opinion. However, the situation illustrated by South Africa where the "winner loses and the loser wins" is always possible or probable under a system of single-member electorates. In contrast, the Hare-Clark system is scientifically and intentionally designed to prevent such thwarting of the democratic process and to assure that a majority of the voters return a majority of the elected representatives ^(a).

3.   HARE-CLARK AVOIDS THE VAGARIES, GAMBLES, AND DISTORTIONS WHICH GENERALLY RESULT FROM ELECTIONS CONDUCTED UNDER SINGLE-MEMBER CONSTITUENCIES.

These capricious results are the rule rather than the exception and are found wherever single-seat electorates prevail. For instance, in the British House of Commons elections: Although the Labor Party's percentage of the total vote increased from 46 per cent in 1950 to 48.5 per cent in 1951, its representation decreased by 19 seats in the 1951 elections. Moreover, in 1945 with 48 per cent of the total vote Labor obtained 62 per cent of the seats (excessively disproportionate to its strength in the country); yet by 1951, although the percentage of the total vote received by Labor had increased to more than its 1945 vote (in fact exceeding the Conservative Party total), the number of seats won by Labor decreased drastically from 399 (62 per cent) in 1945 to 296 (47 per cent) in 1951.

Examination of the voting behaviour in Australia and other democratic countries shows that generally public support for the respective political parties changes only slightly or moderately from election to election, but such changes result in drastic fluctuations in the representation of the parties in the legislature. With single-member electorates, a switch of very few votes from one election to another may severely reduce or enormously magnify a party's membership in Parliament. Under Hare-Clark, since public opinion is faithfully mirrored in the composition of the House of Assembly, stability in representation - not erratic fluctuation - is achieved. If Tasmanian electors continue to prefer the two-party system along present lines, a numerically large Opposition can always be assured, and the Government can be expected to have a workable majority ^(b).

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) The assurance that a majority of the voters within an electorate return a majority of the members requires that the number of seats be odd rather than even, as explained elsewhere in this paper.

(b) Provided an odd number of members is chosen from the electorates.

                                    1958. (No. 22)

9

B. HARE-CLARK AVOIDS GERRYMANDERING AND THE EVER-PRESENT HAZARDS NECESSARILY CONNECTED WITH DETERMINING BOUNDARIES FOR SINGLE-MEMBER ELECTORATES.

Since election results under single-member systems are often influenced as much by the location of the electoral boundaries as by the strength of the voting, it is not surprising that instances and allegations of gerrymandering are widespread in many places using the single-member system. Because of geographical features and such factors as shifts in population, what is alleged as gerrymandering may often be an unavoidable consequence of the limitations of single-member electorates. However, whether what appears as gerrymandering in its effects is intentional or unavoidable, the accusations of undeserved victories leave bitter feelings and bring discredit to democratic principles in general.

Gerrymandering has been widely alleged in almost all Mainland States at some time or other, and charges by the Opposition parties of gerrymandering in both South Australia and Queensland over the last generation have been so severe as to weaken dangerously the citizen's respect for the democratic process. Unfortunately, " winning by cheating " and "dishonest victories " can too often be truthfully applied to gerrymandering with single-member electorates.

Happily, Tasmanians know that gerrymandering has not occurred under Hare-Clark elections and charges of rigging the electoral system have never been made in connection with Hare-Clark. The basic features of this system, including the large, multi-member electorates, serve as a precaution against gerrymandering. Furthermore, the electorate boundaries for House of Assembly elections are established by the authority of another body, namely, the Commonwealth Government.

C. HARE-CLARK PROVIDES THE ELECTOR WITH A WIDER SELECTION OF CANDIDATES THAN ALMOST ANY OTHER METHOD OF PARLIAMENTARY ELECTIONS IN THE WORLD.

At present the Tasmanian voter can choose his representative to the House of Assembly from among at least a dozen and usually more candidates. With seven-member electorates instead of six the scope of selection can be expected to increase somewhat. In contrast, the voter under the single-member systems normally has a choice of only two or three candidates or, in the event of uncontested elections, no selection at all.

One practical effect of this severe limitation of choice is that the elector must normally choose his representative solely on the basis of party affiliations and regardless of the personal qualification of, candidates. When the elector is restricted to a choice between a party he considers unacceptable and an opposing candidate he considers inferior - and this is a common situation under single-member electorates - he generally chooses to accept the inferior candidate and to vote against the party policy which he dislikes.

The voter under Hare-Clark is spared from such invidious choices. With many candidates representing each party he can readily both endorse the political policy he favours and also vote for the candidate he considers most suitable. Assessment on the basis of the personal ability, character, and individuality of the candidates is taken for granted under Hare-Clark, but denied, for all practical purposes, to voters under single-member systems.

When the elector can pick and choose among a half-dozen or more candidates of his own party, as under Hare-Clark, he can exercise a significant discernment which results in more highly selected Parliamentarians. With the wide selection of candidates now available in House of Assembly elections, the appeal to electors here is to vote for someone, not against someone, as is the common practice under the single-member system. If the House of Assembly were chosen from single-member electorates, it could be expected that the present calibre of individual members would decline.

The numerous endorsements by each party under Hare-Clark elections enable both parties to nominate candidates attractive to a wide range of electors. When almost every Tasmanian elector can find a candidate to his liking from among the wide choice

                                 (No. 22.) 1958.

10

offered by both parties, he is not inclined to look outside the two major parties for representation. The satisfying representation thus offered by the Hare-Clark system would seem to be an important reason why the two-party system is, and has been, stronger in Tasmania than any Mainland State and most places overseas.

D. HARE-CLARK ELIMINATES UNCONTESTED SEATS AND SHAM ELECTIONS.

The uncontested return of candidates, though common enough in most single-member systems, is quite undesirable from the standpoint of the public welfare. An immediate consequence, of course, is that even the small electoral privilege existing under single-member representation disappears altogether. The problem of citizen apathy, normally serious enough under the best of circumstances, is further aggravated.

Only a little less objectionable than unopposed elections is the so-called "safe" seat-so named because the party's majority in the electorate concerned is large enough to make the seat a certainty for that party. While this situation means easy comfort f or the occupant of the safe seat, voters of the opposition party in this electorate are virtually disfranchised permanently. Retention of the seat comes to depend less on winning the support of the electors and more on pleasing the small number of party selectors who determine the endorsements.

Elections are a sham if the results are a foregone conclusion; yet this type of sham election is extremely common, normally accounting for more than two-thirds of the seats in a typical single-member State or country. Under Hare-Clark, of course, no seats are uncontested, and none is safe in the respect that a candidate can expect to be returned to office without working for it or without having earned support from the electors.

Genuine competition always exists, therefore, for seats in the House of Assembly. Since this means that Tasmania's M.H.A.s must keep on their toes much more than their counterparts under single electorates, more effective representation is given to Tasmanian electors. The Parliamentarian who favours the single-member system primarily for the fact that it gives him a safe seat should remember that the fundamental purpose of a democratic election system is to provide satisfactory representation for the electors, not personal convenience for the elected. All candidates under Hare-Clark must work for their votes to win a seat in Parliament - because there are no "walkovers" into office via unopposed and safe seats; better service to the electors is one result ^(a).

How the application of single-member electorates to the Division of Denison would create four safe seats is indicated in Table I. On the basis of the last State election returns, the use of this system would result in three "blue ribbon" Labor party seats, namely, Hobart Central and Hobart West, Hobart North, and Moonah, and one "blue ribbon" Liberal party seat. The remaining two seats would probably be "swingers", with Hobart East and Hobart South favourable to Labor and New Town favourable to the Liberal Party ^(b).

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) Because all Federal seats in Tasmania could be considered " swingers " most Tasmanians probably have not experienced and may not know the evils of "safe" seats.

(b) The six proposed single-member electorates are formed from existing subdivisions. Moonah, New Town, and Hobart North are large enough to make three electorates on their own. The remaining three electorates are obtained by combining adjoining subdivisions-one of the proposed electorates being made joining Hobart Central and West, another by combining Hobart East and South, and the last by joining Nelson and Queenborough.

                                1958. (No. 22)

11

TABLE 1 ^(a)

Application of House of Assembly October 1956 Election Returns to Single-Member Electorate System Formed by Dividing the Existing Division of Denison into Six Electorates Based on Present Subdivisional Boundaries ^(b)
Probable Electorates	Labor Party	Liberal Party	A.C.L.P. (c)	Bower, Wedd	Bound, Lloyd	Total Valid	Informal	Total Cast
Hobart Central …	1,196	423	31	136	11	1,797	98	1,895
Hobart West …...	1,722	1,221	86	222	21	3,272	167	3,439
Totals	2,918	1.644	117	358	32	5,069	265	5,334
Hobart North …..	2,357	1,429	94	373	20	4,273	238	4,511
Hobart East ……	621	352	43	56	11	1,083	79	1,162
Hobart South ….	2,008	1,803	118	294	31	4,254	185	4,439
Totals	2,629	2,155	161	350	42	5,337	264	5,601
Moonah ………..	3,318	1,744	137	397	12	5,608	230	5,838
New Town ……..	2,095	2,557	167	399	19	5,237	169	5,406
Nelson …………	589	1,931	60	100	7	2,687	80	2,767
Queenborough …	761	877	43	108	4	1,793	88	1,881
Totals	1,350	2,808	103	208	11	4,480	168	4,648
Total Ordinary Votes	14,667	12,337	779	2,085	136	30.004	1,334	31,338
Absent …………	1,506	1,209	97	206	11	3,029	132	3,161
Section 116A …..	14	18	─	3	─	35	1	36
Postal …………..	650	753	24	62	3	1,492	29	1,521
GRAND TOTAL	16,837	14,317	900	2,356	150	34,560	1,469	36,056

It will be noted that the figures of this election would yield three blue ribbon Labor seats, one blue ribbon Liberal, and two swinging seats, with one of these going to the Labor Party and the other to the Liberal Party.

(a) All election figures in this paper are taken from official reports of the Tasmanian State Electoral Department. The returns shown in this table are first preference votes.

(b) For further explanation of proposed, illustrative electorates, see footnote (b), p. 10.

(c) A.C.L.P. signifies Anti-Communist Labor Party.

________________________________________________________________________________________

E. HARE-CLARK REDUCES THE EXAGGERATED INFLUENCE AND BALANCE-OF-POWER POSITION OFTEN GIVEN TO SMALL POLITICAL FORCES AND PRESSURE GROUPS DURING ELECTIONS WAGED IN SINGLE-MEMBER ELECTORATES

Single-member electorates produce the "worst of two worlds" - on one hand, safe and uncontested seats, which in effect disfranchise large sections of the population; and on the other, "swinging" seats which magnify enormously the power of small groups.

Although the Hare-Clark system gives representation to all contesting groups in proportion to their size provided they can reach a quota, a group that cannot attract a quota of supporters may have little or, possibly, even no influence. In contrast, in any marginal electorate in a single-member system, the seat may swing from one party or the other because of the switch of only a few votes.

                                    (No. 22.) 1958.

12

When tiny sections of the population know that their support can throw, or assist in throwing, the election to one candidate or the other, and perhaps to one party or the other, they acquire a balance-of-power strength out of all proportion to their size and importance. The problem of exaggeration of pressure groups because of the single-member system of representation is particularly serious in a country like the United States where the parties are not responsible nor disciplined in the British sense. The Hare-Clark system, with its quota method of election and its large electorates, helps to keep the influence of small groups more in proportion to their size.

F. HARE-CLARK RELIEVES PARLIAMENTARIANS FROM MUCH NARROW PAROCHIAL PRESSURE AND CAUSES THEM TO BROADEN THEIR OUTLOOKS.

Large electorates necessitate a broad point of view and militate against the narrow parochialism characteristic of single-member electorates, where "parish-pump" pressures are notorious. When a member is responsible solely to a relatively tiny constituency it becomes difficult-if not at times politically hazardous-to take a broad view when local vested interests are involved. Even in the smallest and most homogeneous of the Hare-Clark electorates, namely, Denison, it is more in the general interest that members speak for all of Denison than represent possible individual single-member subdivisions thereof, such as Sandy Bay or Moonah or Hobart Central. The generalisation "the smaller electorate, the smaller the member" points out an advantage of the Hare-Clark provision for larger electorates.

G. HARE-CLARK ENCOURAGES THE PEOPLE TO CHOOSE MEMBERS OF HIGHER CALIBRE.

This results not only from use of large electorates which necessitate a broad point of view but also from the greater competitiveness for political survival under Hare-Clark, which, providing no safe seats, also confronts each member with many rival candidates. With competition keen at election time as well as between elections, far heavier demands are made on the members under Hare-Clark than under single-member systems. This heavier responsibility on Tasmanian M.H.A.s benefits the public because it results in better service to the electors and also tends to discourage weaker candidates from standing for Parliament.

Moreover, the large multi-member electorates under Hare-Clark cause all Parliamentarians to confront a much wider set of problems than under one-member electorates. The effect on outlook and knowledge necessarily is broadening. For candidates unable or unwilling to face this challenge of a big Hare-Clark electorate, the appeal of the small single-seat district can, understandably, be strong. In the public interest, however, it is better to have a system like Hare-Clark which leads candidates to concern themselves with broader public policies and the more important needs of their large electorates than concentrating on narrowly local interests. The candidate of limited ability whose success in single-member electorates depends heavily on facility at "handshaking and baby-kissing" among a local following will find political life more difficult under Hare-Clark conditions.

H. HARE-CLARK AVOIDS THE TWIN EVILS, COMMONLY OBSERVED IN SINGLE-MEMBER SYSTEMS, OF COMPELLING THE DEFEAT OF ABLE CANDIDATES OR ENFORCING THE ELECTION OF INFERIOR ONES.

Since the single-member system permits the election of only one candidate per electorate, this means of course that when more than two outstanding candidates nominate from the same electoral district one must be defeated and when two unqualified candidates nominate for another seat one must be returned. This is not merely a theoretical objection, but a practical weakness of the single-member system observed in some degree at every general election.

This inevitable defeat of some excellent candidates and the equally inevitable election of some inferior ones is avoided by the multi-member feature of Hare-Clark, which has permitted, for example, the electors to choose both the Premier and the Leader of the Opposition from the same electorate, as illustrated in Denison, 1950-55. The simultaneous election of the strong candidates of both parties is taken for granted under Hare-Clark.

                                   1958. (No. 22.)

13

I. HARE-CLARK PROVIDES AN OPPORTUNITY FOR MINOR PARTY AND NON-PARTY REPRESENTATION.

As the underlying principle of the Hare-Clark philosophy of representation is securing the best possible reflection of public opinion, provision for minority representation is a necessary feature of its electoral machinery. The fair representation of the whole of society cannot be assured unless opportunity is provided for the representation of its parts. One requirement, therefore, is multi-member electorates, not single-member ones, for there of course is no way to apportion only one seat between two or more competitors. Multi-member seats permit a fair apportioning of representation, whether divided exclusively between two major parties or shared with minor party candidates. Although the remarkably wide choice of candidates under Hare-Clark provides exceptional satisfaction for the electors within the two-party system, should the electors feel a need for representation outside of the two major parties, the Hare-Clark system makes it available to them provided they can poll 14.3% of the total votes in a six-member electorate or 12.5% in a seven-member one.

While these percentages represent, when expressed in terms of actual votes, close to twice as many votes as would be required for election under the single-member system ^(a), the large, multimember electorates provide a wider area from which to attract the necessary quota. Hence, non-party candidates with substantial support in more than a local area can, and sometimes do, win election under Hare-Clark, as seen in Table VI. With single-member electorates, substantial minorities are excluded unless they happen to be concentrated geographically.

In respect to representation of minorities, Hare-Clark and the single-member system offer sharp contrast: Hare-Clark, which offers a systematic and assured way for non-major party representation, makes this potential representation less necessary because of the high satisfaction resulting from the extensive selection of candidates furnished by the two major parties. On the other hand, single-member electorates, which make no planned provision for minority or non-party representation, increase the need for it because of the restricted expression and poorer satisfaction which they give the electors. Single-member electorates therefore both create the need for better representation and frustrate the realisation of it.

J. HARE-CLARK AVOIDS HIGH ADMINISTRATIVE COSTS AND TECHNICAL DIFFICULTIES OF THE SINGLE-MEMBER SYSTEM.

As these advantages of the Hare-Clark system apply with particular force to Tasmania, this topic is discussed with special reference to this State later on. Two general advantages of Hare-Clark from technical consideration can be mentioned now. First, the Hare-Clark method permits extensive economies in administration because of the larger unit of representation. One example is electoral rolls: They are more easily maintained, especially under a system of compulsory enrolment as, in Australia, if subdivisional and divisional units are fewer and larger. Another example: The larger the electorate, the less frequent the need for redistributions.

Second, Hare-Clark features remove one inevitable and formidable technical drawback to single-member electorates, namely, the difficulty of establishing and maintaining satisfactory electoral boundaries. In order to lay out a plan of single-member districts of approximately equal size with attempted due regard to such guiding factors as community of interest, means of transportation, and physical features, many arbitrary demarcations are required. The much larger electoral units employed under Hare-Clark permit more logical electoral divisions.

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) For example: With a total -State-wide formal vote of 160,000 and 35 single-member electorates of equal size, the quota for election would be 2,286 votes. With the same vote and five seven-member Hare-Clark electorates the quota would be 4,001 votes.

                                  (No. 22.) 1958.

14

When plans for redistribution are announced in most Mainland States the cry of gerrymandering is normally raised by the Opposition. If this charge refers to intentional gerrymandering, it is true in some but not all cases. Many times the cause for complaint often arises simply from the necessities of cutting up the map arbitrarily to compensate for population changes within electorates. When districts are small, a large measure of objectionable, but unavoidable, boundary making is required. With the increase in electorate size the effect of population movement is reduced and at the same time the possibility of establishing more logical boundaries is improved.

K. HARE-CLARK ELIMINATES THE UNDUE INSECURITY OF MEMBERSHIP FACING MANY PARLIAMENTARIANS ELECTED UNDER THE SINGLE-MEMBER SYSTEM.

Snug security for some and "sudden death" for others is the rule of the single-member system. Not all members of Parliaments chosen from single-member electorates have safe seats. On the contrary, the occupants of "swinging" seats are put under a jeopardy which no member under Hare-Clark needs to fear. A capable, deserving member under the Hare-Clark system has reasonable expectation of being returned. But even a slight swing in public opinion under the single-member system, perhaps caused by reverses in party popularity completely beyond the control of individual members, can sweep from office those party members not lucky enough to have safe seats.

The extremes of unwarranted security for some and unreasonable jeopardy to others do not help to attract worthy prospective candidates into politics under single-member conditions. Moreover, members cannot give their best service if they live always under the disquieting possibility or likelihood of being "tossed out" at the next election-perhaps through no fault of their own.

Illustrations of drastic fluctuations in Parliamentary membership under single electorates are not difficult to find, for instance, Ceylon. Although the percentage of the total vote received by the United National Party declined, from the Parliamentary elections of 1952 to those of 1956, by 16%, its representation fell from 54 seats in 1952 to eight seats in 1956, or from 57% to 8% respectively. While the vagaries of single-member electorates inflated the representation obtained by the U.N.P. in 1952, the electoral gamble severely under-represented it in 1956.

Another example of the single-member system causing "sudden death" for sitting members is furnished by the two general elections for the Canadian House of Commons in the past year. As a result of the June, 1957, elections Liberal Party membership in Parliament dropped from 171 to 104 and that of the Progressive Conservative Party increased from 50 to 110. The effects of the March, 1958, elections showed even greater fluctuation, as the Progressive Conservatives gained in seats from 110 to 209 and the Liberals fell from 104 to 47. Within a one-year period, therefore, Liberal Party Parliamentary membership dropped from 171 to 47 and that of its chief opponent rose from 50 to 209 ^(a).

L.  HARE-CLARK AVOIDS THE "REGIMENTED DOWN-THE-TICKET "VOTING PRACTISED IN AUSTRALIAN SENATE ELECTIONS ON THE MAINLAND AND SOMETIMES IN TASMANIA.

Under the Hare-Clark system the elector makes his selections with fullest freedom, uninfluenced by the numbered type of how-to-vote cards generally used in Senate elections.

The heart of the success of the Hare-Clark system could be said to be this unhampered freedom of the electors to pick and choose as they please. If the listing of candidates on the ballot-paper were determined by "mutual consent", as in the Senate elections, and combined with the use of numbered how-to-vote cards, a great measure of the value of Hare-Clark would be destroyed. In contrast, the Hare-Clark system provides for an alphabetical listing of candidates, and no attempt is made by the political parties to suggest to their supporters any prescribed order for marking preferences. If voting "to order" "down the ticket", as in Senate election style, were followed, the choice of members of Parliament would pass, for all practical purposes, from the voters to political party management.

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) Figures for 1958 are from latest Press reports available, which noted the results for 264 of the 265 seats in the Canadian House of Commons, one seat being in doubt.

                                   1958. (No. 22.)

15

As the Hare-Clark system now is, it provides the Tasmanian elector with a more effective vote among a wider range of candidates than any other method of Parliamentary election in the world.

The free selection under the Hare-Clark system assures competition among candidates and keeps the elector sovereign. This unequalled privilege of choice is therefore one of the most significant values of the voters' franchise in Tasmania and sets Hare-Clark in a class apart from the Senate election system, which otherwise follows the Hare-Clark system in most features. May it be hoped that the superior Hare-Clark example will serve as a model and incentive for improving the Australian Senate system.

Objections to Hare-Clark ^(a) .

Little on earth is perfect, and this generalisation applies to Hare-Clark as well, even although its overall drawbacks are relatively slight. Besides, almost all complaints are from the Parliamentarian's point of view, not the elector's. One objection is that the multi-member electorates involve the members in costly travel required for attending to their duties in serving large constituencies.

From the angle of expense, this claim is quite valid, particularly for members outside Denison. The answer, however, is not to cut up the State into 30 or 35 single electorates-and bring the evils of single-member representation into Tasmania - but to provide an adequate travel allowance for members so that these costs do not draw too heavily from their modest salaries ^(b). In a small single-member electorate much recompense for travel would not be justified, but a large Hare-Clark electorate is another question and warrants increased subsidisation of travel expenses. There is no denying that the advantages of large electorates do necessitate greater cost in travel.

Another complaint is that the large Hare-Clark electorates cover such a wide scope of different industries, interests, and problems that it is difficult for members to keep abreast of the many diverse activities on which they are expected to be informed. Small, single-member divisions, it is claimed, would save members from the necessity of getting acquainted with this wide range of activity. The demands on members under Hare-Clark are greater than they would be under small electorates. From the elector's viewpoint, however, this complaint becomes an advantage.

The contact with varied and more complex problems under Hare-Clark is most educational and militates against the parochialism so characteristic of small electorates. Both consequences are as they should be-from the public's point of view. In their interest it is best for the electoral system to be geared to attract the more capable and energetic candidates; if weaker members find the large Hare-Clark electorates too difficult to represent, the adjustment should not be to tailor the electorates to suit these members, but to let the seats be filled by those to whom the large electorates are not too great a challenge.

A further objection to Hare-Clark made by some members is that the competition for No. 1 votes results in undesirable manifestations of rivalry between candidates of the same party. What solution do these members sometimes suggest? - use of single-member electorates. And why? - for the avowed purpose of eliminating as much competition between persons as possible, in order to make contests more strictly a struggle between parties.

This suggested "solution" to reduce rivalry for No. 1 votes, therefore, is to take from the electors their present breadth of choice in order to suit the private conveniences of a few members who wish to by-pass as much competition as they can. The elimination of the right of electors to pick and choose among candidates would strike at the heart of the Hare-Clark system.

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) As footnoted on P. 5, the subject of advantages and disadvantages of Hare-Clark is being analysed in a documented study now under preparation. Necessarily this section of the paper must be concerned with conclusions, not detailed evidence.

(b) Contributions to charities, societies and clubs, plus travel expenses and other costs involved in serving their electorates, consume a far greater part of a member's recompense than the average person realises.

                                   (No. 22.) 1958.

16

     The opportunity for electors to select between candidates as well as between parties has always been considered an integral part of the Hare-Clark philosophy of representation; attempts to abrogate this opportunity of choice are simply an undisguised effort to abridge democracy and curtail the special electoral privilege which distinguishes Hare-Clark from lesser systems.

If candidates should fail to observe fair play with their party colleagues in campaigning for No. 1 votes among supporters of their own party, this condition is not really a fault of the Hare-Clark system, but the result of the absence of a candidate's code of rules for campaigning. With proper loyalty to one's party and regard for one's colleagues, the existence of competition for No. 1 votes need not be objectionable-except to members who want no competition at all. An electoral system cannot both provide the advantages of competition and at the same time eliminate it.

Fortunately, many members realise that the rivalry among members of the same party within the same electorate is healthy and desirable and results in better representation f or the electors. Furthermore, candidates who object to this competition within parties should realise that single-member electorates frequently put pressures on candidates in -several ways which multi-member electorates do not, namely: (a) If a seat is a safe one, competition for party endorsement is exceedingly severe; (b) If the seat is a "swinger", extreme uncertainty about tenure subjects the holder to undue pressures from small, but potentially balance-of-power constituent groups; (c) Pitting one candidate directly and specifically against a single opponent often gives rise to certain undesirable campaign conditions (like public mud slinging), which are comparatively rare under Hare-Clark.

In any case, a good deal of whatever intra-party rivalry now exists would be eliminated if seven-member districts were adopted, for the reasons noted in the report of the Select Committee last year ^(a) .

An objection to Hare-Clark held by some is that the man in the street does not understand the counting procedure. In view of the lack of printed material on Hare-Clark at the present time, this is not surprising. With education on the subject, however, the problem could be solved. In any case, there is little more need for the man in the street to understand the methods of counting votes (which is the job for the returning officer) than there is for him to know the intricacies of his watch or wireless set.

Part III.

WHY SIX-MEMBER ELECTORATES ARE FAULTY.

The use of six-member electorates is gravely defective because it cannot provide the premise fundamental to a democratic electoral system, namely, that a majority of the voters within an electorate shall be assured of the right of returning a majority of members in that electorate. Only a system like Hare-Clark can give an assurance of majority rule, and Hare-Clark can do it only if an odd, rather than even, number of members is chosen from the electorates.

As the electoral quota under Hare-Clark is based upon the number of seats to be filled, the existing six-member quota is 14.3 per cent of the valid votes cast in each electorate. Consequently one party can receive as few as 42.9 per cent of the votes (just three quotas) and obtain three seats, and the other party can receive almost 57-2 per cent and still obtain only three seats. If an odd number of members is chosen within an electorate, any party or any group obtaining a majority of the votes, no matter how slight, must win a majority of the seats in that electorate. Yet, when even numbers are chosen, if the majority party does not exceed the minority by at least one quota of votes, the minority party will receive representation equal to that of the majority.

___________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) Report of the House of Assembly Select Committee on Electoral Reform, Parliamentary Paper No. 59 of 1957, p. 3.

                                       1958. (No.22.)

17

Since the electoral quota for seven-member electorates would be 12.5 per cent, a party would win three seats if it obtains 37.5 per cent of the votes. If this party wins another quota of votes, thus obtaining a majority of the total votes of the electorate ^(a), it must win the fourth seat.

In Tasmania, as in other countries with a two-party system, differences in voting strength between the major parties, taken on total votes over large areas, are usually fairly small. Hence a difference of 14.3 per cent of the votes (the size of a six-member quota) would be exceptional, and the possibility of a party obtaining a fourth seat remote. The likelihood, therefore, of two parties receiving an equal number of three seats each per electorate, despite unequal voting strength in the electorate concerned, is theoretically very probable-and has been demonstrated by experience as well.

As seen in Table II, the Hare-Clark system has been used for 17 general elections since its adoption, making a total of 85 separate elections for all five electorates. Although six different representation patterns have occurred in these 85 elections, in 44 cases the representation per electorate was three seats for Labor and three for non-Labor. Excluding the exceptional years of 1909, 1931, and 1941, there have been only 12 cases in the history of Hare-Clark where one- party- was able to obtain four of the six seats. Since World War II only once has a party been able to win the fourth seat within any electorate.

An unjust feature of these 44 cases of 3-3 "dead heat" results is that in most cases the support given to the majority party was markedly greater than that received by the minor party. As seen in Table III, there have been only eight instances where the two parties were separated by less than 2 per cent of the total votes cast. Any difference in votes between the parties greater than 5 per cent over a large area like a six-member electorate, can be considered to be marked; yet almost one-third of the cases of equal party representation have occurred when the difference in total votes has been between 5.0 and 9.9 per cent. In one fourth of the cases the difference has exceeded 10.0 per cent.

Equal party representation on unequal voting strength is a denial of electoral justice to the voters. While this condition is not to be compared with the frequent injustice under single-member electorates, where the majority party commonly obtains only a minority of the seats, the fault is nevertheless grave and cannot be reconciled with precepts of democratic representation.

Thanks to the ingenious features of the existing Parliamentary "deadlocks" law, commonly known as the Lyons Bill ^(b) , some cases of equal party representation are dealt with in such a way as to provide both electoral justice and a workable Parliament, namely, in those instances when the two parties are divided 15-15 in the House and happen also to be supported by the electors in nearly equal voting strength. In cases like the elections of 1955 and 1956, however, when the Government party obtained State-wide majorities of approximately 11,000 votes ^(c), the concession of the equivalent of an extra vote in the House was not in reasonable proportion to its popular majority. Unless the voting strength of the parties in Parliament is reasonably proportional to the support they received in the country, electoral justice to the voters is not accorded.

In cases where the parties are not returned 15-15, the Parliamentary "deadlocks" law does not apply, and the defect of six-member electorates then looms even greater. It is quite possible as a consequence of six-member districts, for a

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) Because of the exhaustion of votes the last candidate is often elected without a quota. The party winning four seats sometimes might appear, therefore, to have less than a majority of votes at the conclusion of the count.

(b) Formally called the Constitution Act (No. 2) of 1954, this law provides that if the two parties are returned in equal numbers of 15 each the majority party forms the Government, and the Opposition party has the privilege of supplying the Speaker. The party which obtains the most primary votes on a State-wide basis is declared the majority party. If the minority party declines to provide the Speaker and the majority party does so, it is permitted to replace the member chosen as Speaker by a recount of the ballot-papers which elected the Speaker to Parliament.

(c) 11,403 votes in 1955 and 10,619 in 1956.

                                    (No. 22.) 1958.

18

party to receive a majority of the vote in every electorate but one, and yet receive fewer seats that the party with a minority of the total vote. A situation like this in fact occurred in 1928 when the Labor party received either an absolute or relative majority over the Nationalist party in State-wide totals and in each of the five electorates, yet lost to the Nationalists. This occurred because the Labor majorities were not large enough to gain the fourth seat in any electorate and because in one electorate (Franklin) Mr. B.J. Pearsall, and Independent, was narrowly elected, causing Labor to lose the third seat in that division. Despite Labor's lead over the Nationalists in popular votes, as seen in Table V, Labor lost to the Nationalists by 14 seats to 15, plus Mr. Pearsall, who supported the Nationalists in Government.

In 1950 the Liberal party potentially could have been in the position of Labor in 1928, as the victory by Mr. W.G. Wedd, an Independent, prevented the Liberals from winning a third seat in Denison. The representation after that election was 15 Labor, 14 Liberal, and Mr. Wedd, who supported Labor in Government.

TABLE II ^(a).
Representation of Parties and Independents in Respect to Frequency of Patterns of Representation Within Electorates, House of Assembly Elections, 1909-1956.
	Patterns of Representation ^(b)							Total Elector- ates
Year	3─L.	4─L.	2─L.	3─L.	2─L.		3─L.
	3─N.L.	2─N.L.	4─N.L.	2─N.L.	3─N.L.		1─N.L.
				1─Ind.	1─Ind.		2─Ind.
Frequency of Occurrence of Pattern
1909			5					5
1912	4		1					5
1913	4		1					5
1916	3		1	1				5
1919	2		2	1				5
1922	2		2			1		5
1925	2	1		1			1	5
1928	4					1		5
1931			4			1		5
1934	2			2		1		5
1937	2	3						5
1941		5						5
1946	2	1		2				5
1948	3			1			1	5
1950	4			1				5
1955	5							5
1956	5							5
	44	10	16	9		4	2	= 85

(a) Following the first Hare-Clark election of 1909, a two-party political system developed in Tasmanian State politics. The Labor party side of politics has continuously been known by this name or the fuller designation, the Australian Labor Party. The non-Labor party has been known by several names, first Liberal, then Nationalist, then United Australia, and now the Liberal Party. In Tables covering a period of years, the term "non-Labor" is used if the name of the non-Labor party changed during the period under reference.

      The patterns of representation refer to the number of seats won by the respective contestants within the five electorates. "L" denotes the seats won by the Labor party, "N.L." by the non-Labor party, and "Ind." by Independent candidates. Six seats are filled in each electorate. Since 1909 there have been only six different patterns of representation, or ways in which the seats have been divided, as noted in the table.

(b)

        1958. (No. 22)

19

In view of the loss of the third seat in Denison, the Liberal party could have had nearly a fourth quota in four electorates, or a possible State-wide majority of over 15,000 votes, and still have remained in Opposition.

The six-member electorate, by preventing the majority of the voters from generally returning a majority of the members from any electorate, can allow the anomalies cited above to occur. Use of a seven-member electorate will enable the majority party in in each division to obtain its rightful due, and the certainty of the majority party obtaining a majority of the seats can normally be expected.

TABLE III. ^(a).

Size Of Majority Recorded for Larger Party in Cases of 3-3 Division of Seats Between the Two Parties, House of Assembly Elections, 1909-1956.

Majorities for Labor.	Majorities for Non-Labor
0.5	0.2
0.6	0.8
1.0	2.2
1.6	2.3
1.6	3.0
1.8	3.6
2.0	3.6
2.2	4.0
2.3	5.6
3.0	7.1
3.5	11.4
5.6	12.0
6.0	12.4
6.1	12.7
6.6	14.6
7.0	15 cases
7.3
7.5
7.9	SUMMARY. Size of Majorities No. of Cases Under 2 per cent = 8 2.0 to 4.9 = 11 5.0 to 9.9 = 14 Above 10.0 = 11 44
8.3
8.4
8.9
9.2
11.2
11.4
11.8
13.5
14.9
15.4
29 cases

(a) The percentages of difference are between the major parties within electorates and are based on first preference votes. All votes included in tables of this paper refer to first preferences. Because cross-voting between parties is extremely small, calculations on first preferences are quite satisfactory for many types of analyses.

Summary of Part III.

Equal party representation on unequal voting strength is a denial of electoral justice to the voters. Six-member electorates, even with the present Parliamentary " deadlocks " law, cannot guarantee majority rule. Commendable though the " deadlocks " law is in mitigating one type of misrepresentation, it does not, and cannot, provide the correction for the many anomalies which result from six-member electorates.

(No. 22.) 1958.

20

TABLE IV.
Summary of Representation on Electorate Basis, House of Assembly Elections, 1909-1956
	Bass	Braddon	Denison	Franklin	Wilmot
1909 . . . .	2─4	2─4	2─4	2─4	2─4
1912 . . . .	(3.6%)	(5.6%)	(2.2%)	(14.6%)	2─4
1913 . . . .	(0.6%)	(2.2%)	(4.0%)	(12.4%)	2─4
1916 . . . .	(7.9%)	3─2─1	(0.8%)	(3.6%)	2─4
1919 . . . .	2─4	3─2─1	(11.4%)	(7.1%)	2─4
1922 . . . .	2─3─1	(3.0%)	(12.7%)	2─4	2─4
1925 . . . .	(0.2%)	(3.0%)	4─2	3─1─2	3─2─1
1928 . . . .	(9.2%)	(1.0%)	(0.5%)	2─3─1	(8.9%)
1931 . . . .	2─4	2─4	2─4	2─3─1	2─4
1934 . . . .	(1.6%)	(12.0%)	3─2─1	3─2─1	2─3─1
1937 . . . .	4─2	(15.4%)	4─2	4─2	(13.5%)
1941 . . . .	4─2	4─2	4─2	4─2	4─2
1946 . . . .	4─2	(1.8%)	3─2─1	3─2─1	(14.9%)
1948 . . . .	(11.8%)	(2.0%)	3─1─2	3─2─1	(6.0%)
1950 . . . .	(8.4%)	(5.6%)	3─2─1	(2.3%)	(2.3%)
1955 . . . .	(7.5%)	(6.6%)	(8.3%)	(3.5%)	(11.4%)
1956 . . . .	(7.0%)	(1.6%)	(7.3%)	(6.1%)	(11.2%)

Key:

1. Designation of seats won:

By Labor─in light figures.

By Non-Labor─in heavy figures.

By Independents─in italic figures.

2. In cases of 3-3 division of seats between the two parties, the size of the majority of larger party is indicated in percentages in appropriate type style to designate which party received the majority of votes.

Part IV.

WHY THE SOLUTION DEPENDS ON SEVEN-MEMBER ELECTORATES.

The return of an odd number of members within electorates is required, as noted in preceding sections, to assure the successful operation of the electoral mechanism of the Hare-Clark system. The next question is which odd number should be used, 3, 5, or 7? This section will endeavour to show that regard for both sound electoral principles and special conditions prevailing in Tasmania virtually limits the avenues of solution to one proposal, namely, the election of seven members from each of the existing five electorates.

Three-Member Electorates.

In respect to electoral principles, three-member districts are quite unsatisfactory. Firstly, the relatively small three-member electorates would almost certainly promote a parochialism now fortunately absent in House of Assembly electorates. Secondly, the resultant two to one division of party representation within electorates would introduce objectionable features. For example, on one hand, if the two parties in some electorates are almost evenly divided in voting strength, severe distortion of representation will result because one party, with only slightly more votes, must gain twice as many seats.

1958. (No. 22)

21

On the other hand, the smaller three-member districts would mean that some electorates would have unchallengeable majorities for one party or the other. Consequently, the certain foreknowledge of a 2-1 result in favour of the Liberal party in some electorates and the Labor party in others would greatly reduce competition between the parties in these electorates. Three-member electorates would introduce in a modified form some of the evils of "safe" seats under the single-member system.

Thirdly, three-member electorates would drastically curtail the desirably wide selection of candidates now available to the electors. Fourthly, use of a three-member electoral quota would necessitate an unduly large number of unutilised remaindered votes. With seven-member electorates the unrepresented fraction is 12.5%, but with three-member electorates, it is 25%.

TABLE V.
General Election Results, House of Assembly, May 30, 1928. (In Votes and Percentages within Electorates)
	Bass		Darwin		Denison		Franklin		Wilmot		Total Votes for Parties
Labor … …	9,386 (3)	54.6	8,758 (3)	46.7	7,545 (3)	40.9	7,886 (2)	43.5	8,274 (3)	50.9	41,831 (14)	47.2
Nationalist …	7,795 (3)	45.5	8,574 (3)	45.7	7,448 (3)	40.4	6,783 (3)	37.4	6,832 (3)	42.1	37,432 (15)	42.2
Independents	─		1,413	7.6	3,438	18.7	3,450 (1)	19.0	1,142	7.0	9,443 (1)	10.6
Totals per Electorate …	17,163		18,745		18,431		18,119		16,248		88,706

NOTE.─Number of seats won is indicated in parentheses.

Five-member Electorates.

The shortcomings of three-member electorates from the standpoint of principle apply in a lesser degree to five-member electorates. For example, a 3 - 2 ratio in dividing seats between two parties

                                 (No. 22.) 1958.

 22

of nearly equal strength is too rough for proportioning the representation closely. A 4 - 3 distinction supplies a much truer and more reliable result. The unrepresented fraction of votes within each electorate is now 14.3%; this amount would decrease to 12.5% with seven-member electorates, but increase to 16.7% if five members were chosen. Five-member electorates would reduce, rather than improve, the selection of candidates available to the elector.

Unless the present electoral boundaries were dispensed with, five-member electorates would require a 25-member House, which would obviously be too small for effective Parliamentary government under present-day conditions. If seven new electorate, returning five members each were established, smaller districts would result, giving rise to certain disadvantages because of existing, concentrations of political party strength in various areas of Tasmania. As noted by the Select Committee ^(a), the use of seven five-member electorates would probably mean that four or five of the electorates would have "safe" majorities for either one party or the other. Since the resultant 3 - 2 victories by one party in two of the electorates would be offset by 3 - 2 victories for the other party in two or three electorates in which it had a strong majority in voting strength, the effective decision as to which party would win the election would rest in two or three electorates only. The result probably would be "a much more frequent occurrence of small majorities for the governing party than would be the case with five seven-member electorates" ^(b) .

With five seven-member electorates instead of seven five-member ones, the variations in party voting strengths within electorates are smaller because larger electorates encompass a wider range of people, thus producing a more average result in terms of distribution of party voting strength. Fewer and larger electorates provide a better cross section of the State than smaller, more numerous ones. Consequently, the same fairly close balance between the two parties on a State-wide basis is found also within most of the present electorates, almost all of which could be considered "swingers" "Swinging" electorates result in more competition between parties and in the greater probability of the governing party having an adequately ample majority in Parliament. The existing five electorates are large enough to permit almost all of them to be a fair average of the State. Three-member or five-member electorates in a 30 or 35-member House are too small to encompass a fair cross section.

Practical Considerations.

In addition to superiority with regard to the paramount consideration of principles of representation, the election of seven members from existing electorates is desirable because of extremely important practical factors. The seven-member plan permits complete utilisation of joint Commonwealth-State electoral boundaries and rolls. This joint arrangement is most advantageous financially and also saves Tasmanian electors from the inconvenience of contending with separate enrolment systems for Federal and for House of Assembly purposes.

Any electoral changes incorporating either three-member or five-member districts with a 30- or 35-member House would necessitate separate State electoral boundaries. This change would mean the establishment of new boundaries initially, and periodic redistributions thereafter. Moreover- the use of electorates smaller than the present ones -would make the need for redistributions more frequent. Boundary changes involve appreciable costs and must: contribute to confusion and inconvenience for electors.

With seven five-member electorates joint Commonwealth-State rolls could be used, but some subdivisional changes would be necessary. With three-member districts most of the existing subdivisional boundaries would need to be altered and many new subdivisions created in order to make possible the use of joint rolls.

If single-member electorates were adopted for electing: the House of Assembly, separate boundaries would of course be required, and the present joint Commonwealth-State rolls could not be used without almost complete modification involving manifold

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

(a) Paragraiph 12 of the Select Committee Report on Electoral Reform, Parliamentary Paper No. 59 of' 1957.

(b) Ibid.

1958. (No. 22)

23

administrative problems ^(a). The cost alone of printing the latest joint rolls was ₤4,197 ^(b), and the work of maintaining and compiling the rolls is naturally far more expensive than printing them. Because of the unavoidable smallness of single-member districts in Tasmania, redistributions would need to be inordinately frequent. Factors of practicability would condemn the single-member system in Tasmania, if it were not already condemned on the basis of electoral injustice.

Application of Seven-member Quota to Past Elections.

Because of the extensive data on the Hare-Clark election result sheets it is quite possible to ascertain what results the seven-member quota would have produced if used in past elections ^(c). By re-working the original result sheets on the basis of the smaller seven-member quota, a clear-cut answer can be found in almost every case as to what the election outcome would have been if seven members instead of six had been chosen from each electorate. In elections where the difference in votes obtained by the respective parties was not close and where the vote for Independent or non-major party candidates was not large, the election results with a seven-member quota can be ascertained without re-working a full sheet.

_______________________________________________________________________________________

(a) Discussion by Mr. Edward Parkes, Chief Electoral Officer of Tasmania, on the incompatibility of using single-member electorates and joint Commonwealth-State electoral machinery is contained on pages 59-65 of the evidence attached to the Report of the Board of Inquiry on Parliamentary Deadlocks, Parliamentary Paper No. 6 of 1951.

(b) Information supplied by Commonwealth Electoral Office, Hobart.

(c) Examples of the method used in re-working the House of Assembly elections of 1956 can be seen by examining the scrutiny sheets submitted by the State Electoral Department. See pages 64-68 of the evidence attached to the Select Committee Report on Electoral Reform, P.P. No. 59 of 1957.

TABLE VI.
Summary of Comparative Results of Past Elections for House of Assembly Based on Return of Six and Seven Members from each Existing Electorate.
	Actual Results Six-Member Quota			Results Calculated with Seven-Member Quota
	A.L.P.	Non-Labor	Others	A.L.P.	Non-Labor	Others
1956	15	15	─	20	15	─
1955	15	15	─	20	15	─
1950	15	14	1	17	17	1
1948	15	12	3	18	14	3
1946	16	12	2	20	13	2
1941	20	10	─	23	12	─
1937	18	12	─	21	13	─*
1934	14	13	3 ^(a)	17	16	2 ^(b)
1931	10	19	1	12	20	3
1928 ^(c) (d)	14	15	1	17	15	3
1915	14	16	─	15	20	─
1909	12	18	─	14	21	─

* One seat not ascertained.

(a) 1934: Carruthers, Independent, elected without a quota as the sixth candidate in the original count, is not elected under the seven-member count.

(b) 1934: One Independent was Pearsall, regarded as Nationalist in sympathy; the other Independent was Becker, who was Independent Labor.

(c) For comment on 1928 election results, see text p. 18.

(d) The Liberal party won the elections of 1913, 1916, and 1919. In the elections of 1922 and 1925 the Country party contested and won seats as well as the Nationalists and the Labor party.

                             (No. 22.) 1958.

24

TABLE VI.
Comparative Results of Past Elections for House of Assembly Based on Return of Six and Seven Members from Each Existing Electorate
	Actual Results, Six-Member Quota				Results Calculated with Seven-Member Quota. Seat Allocation, Additional Members, Remarks.
1956 ^(a)
	A.L.P.	Lib.	Others	A.L.P.		Lib.	Others	Name, Party
Bass … …	3	3	─	4		3	─	Madden, A.L.P.
Braddon …	3	3	─	4			─	Taylor, A.L.P. Close contest ^(b)
Denison …	3	3	─	4		3	─	Miller, A.L.P.
Franklin …	3	3	─	4		3	─	Hand, A.L.P.
Wilmot …	3	3	─	4		3	─	Spurr, A.L.P.
	15	15		20		15

(a) Results fpr 1956 supplied by State Electoral Department to Select Committee on Electoral Reform.

(b) Note on Braddon: According to evidence available on official scrutiny sheet, Taylor would be elected. In view of closeness of margin between Taylor and Fidler (Lib.), however, it is possible that the latter, if aided by considerable exhaustion or cross-voting not detectable on the scrutiny sheet, could be elected instead of Taylor.

1955
	A.L.P.	Lib.	Others	A.L.P.	Lib.	Others	Name, Party
Bass … …	3	3	─	4	3	─	Atkins, A.L.P.
Braddon …	3	3	─	4	3	─	Ward, A.L.P.
Denison …	3	3	─	4	3	─	Bower, A.L.P.
Franklin …	3	3	─	4	3	─	Crawford, A.L.P.
Wilmot …	3	3	─	4	3	─	Fisher, A.L.P.
	15	15		20	15

1950
	A.L.P.	Lib.	Others	A.L.P.	Lib.	Others	Name, Party
Bass … …	3	3	─	4	3	─	Oliver or Atkins, A.L.P.
Braddon …	3	2	─	3	4	─	Breheny, Lib.
Denison …	3	3	1	3	3	1	Duncan, Lib.
Franklin …	3	3	─	3	4	─	Solomon, Lib.
Wilmot …	3	3	─	4	3	─	Fisher, A.L.P.
	15	14	1	17	17	1

1948
	A.L.P.	Lib.	Others	A.L.P.	Lib.	Others	Name, Party
Bass … …	3	3	─	4	3	─	Atkins, A.L.P.
Braddon …	3	3	─	3	4	─	Acheson, Lib.
Denison …	3	1	2	3	2	2	Duncan, Lib.
Franklin …	3	2	1	4	2	1	Worsley, A.L.P. Close contest ^(a)
Wilmot …	3	3	─	4	3	─	Burnell, A.L.P.
	15	12	3	18	14	3

(a) Whether Worsley, A.L.P., or Park, Lib., would win the seventh seat would depend on the distribution of Pearsall's preferences. As Mr. Pearsall, Ind., was runner-up in the original six-member count, the allocation of his preferences is not revealed on the result sheet. When asked how he thought his preferences would be distributed Mr. Pearsall stated that he believed Mr. Worsley would have received one-half of his (Pearsall's) preferences in Kingborough, and one-third of those from other subdivisions. Because of Worsley's lead over Park at the time of the exclusion of Pearsall, this would mean a narrow win by Worsley, although the margin would be close.

1946
	A.L.P.	Lib.	Others	A.L.P.	Lib.	Others	Name, Party
Bass … …	4	2	─	4	3	─	Hollingsworth, Lib. ^(a)
Braddon …	3	3	─	4	3	─	Lane ^(b), A.L.P.
Denison …	3	2	1	4	2	1	Tyler, A.L.P.
Franklin …	3	2	1	4	2	1	Hand, A.L.P.
Wilmot …	3	3	─	4	3	─	Burnell, A.L.P.
	16	12	2	20	13	2

(a) Note that A.L.P. won 4─2 in actual contest but would win 4─3 with seven-member count.

(b) Contest close between Acheson, Lib., and Lane, A.L.P., but result is not in doubt.

1958. (No. 22)

25

1941
	A.L.P.	Nat.	Others	A.L.P.	Nat.	Others	Name, Party
Bass … …	4	2	─	5	2	─	Welsh or Quintal, A.L.P.
Braddon …	4	2	─	4	3	─	Wright, Nat.
Denison …	4	2	─	5	2	─	Heerey, A.L.P.
Franklin …	4	2	─	4	3	─	Plummer, A.L.P.
Wilmot …	4	2	─	4	3	─	Foster, Nat.
	20	10		23	12

1937
	A.L.P.	Nat.	Others	A.L.P.	Nat.	Others	Name, Party
Bass … …	4	2	─	5 ^(a)	2	─	McElwee, A.L.P.
Braddon …	3	3	─	4	3	─	Aylett, A.L.P.
Denison …	4	2	─	4	2	─	(^b)
Franklin …	4	2	─	4	3	─	Pearsall, Nat.
Wilmot …	3	3	─	4	3	─	Spurr, A.L.P.
	18	12		21	13	─	(^b)

(a) Result for seventh seat was close, but on available figures Labor would have won five of the seven seats; with a little more cross-voting the result could have been 4─3.

(b) Winner of seventh seat in Denison not ascertained, but would have been either White, A.L.P., or Johnson, Nat.

1934.
	A.L.P.	Nat.	Others	A.L.P.	Nat.	Others	Name, Party
Bass … …	3	3	─	4	3	─	Lamp, A.L.P.
Braddon …	3	3	─	3	4	─	Wright, Nat.
Denison …	3	2	1	4	3	─	Mahoney, A.L.P. ^(a)
Franklin …	3	2	1	4	2	1	Frost, A.L.P.
Wilmot …	2	3	1	2	4	1	Best, Nat.
	14	13	3	17	16	2

(a) Carruthers, and Independent elected without a quota as the sixth candidate under a six-member count, is not elected under the seven-member count; Mahoney, A.L.P., is elected instead.

1931
	A.L.P.	Nat.	Others	A.L.P.	Nat.	Others	Name, Party
Bass … …	2	4	─	3	4	─	Becker, A.L.P.
Braddon …	2	4	─	2	4	1	Medwin, Ind.. ^(a)
Denison …	2	4	1	3	4	─	Cosgrove, A.L.P.
Franklin …	2	3	─	2	3	2	Pearsall, Ind.
Wilmot …	2	4	─	2	5	─	Best, Nat.
	10	19	1	12	20	3

(a) Medwin, Ind., was runner-up under six-member count.

(b) Pearsall, Ind., was runner-up under six-member count.

1928 ^(a)
	A.L.P.	Nat.	Others	A.L.P.	Nat.	Others	Name, Party
Bass … …	3	3	─	4	3	─	Davies, A.L.P.
Braddon …	3	3	─	3	3	1	Whitsitt, Ind.. ^(a)
Denison …	3	3	─	3	3	1	Mahoney, Ind. Labor.
Franklin …	2	3	1	3	3	1	Sheridan, A.L.P.
Wilmot …	3	3	─	4	3	─	Osborne, A.L.P.
	14	15	1	17	15	3

(a) For comment on 1928 election results, see text p. 18.

The practical demonstration of the effect of the seven-member quota on past elections, as seen in Tables VI. and VII., shows that the result is what one could expect as a consequence of the principles involved, namely, that the Opposition Party remains strong but that the Government generally has a better working majority.

                                   (No. 22.) 1958.

26

Independents and Minor Party Candidates.

The use of the seven-member quota for the last ten elections since 1928 indicates that since 1934 there would have been no change in respect to the election of Independent or minor party candidates. In 1934, Mr. D. A. Carruthers, an Independent elected is the last candidate without a full quota, as noted in Table VII, would not have been elected under a seven-member quota. One Independent Labor candidate and one Independent in 1928 and two Independents in 1931 would have been elected as a consequence of the smaller seven-member quota. Of the 50 additional seats which would have been filled since 1928 if seven-member electorates had been used, the net gain by non-major party candidates would, therefore, have been four.

Part V.

OTHER CONSIDERATIONS.

HARE-CLARK IS IMPARTIAL POLITICALLY.

The Hare-Clark system is singularly neutral in its operation. All parties and all candidates are treated with scrupulous impartiality. Whether candidates are from the Government or the Opposition, or from a major party or a minor one, or standing as Independents, all must reach the same quota of votes in order to win.

Moreover, the fortunes of Hare-Clark have in the past been identified with both parties. It owes its very adoption to the determined efforts by the non-Labor political parties in Tasmania in face of strong Labor party opposition in 1906. When the Labor party first came to power under Premier John Earle in 1914, extensive, but unsuccessful, efforts were made to have Hare-Clark replaced by a party list system of representation.

Since then Hare-Clark has been continued by both parties. In 1951 a Board of Enquiry on Parliamentary Deadlocks, appointed from outside Parliament and headed by Professor T. Hytten, Vice-Chancellor of the University of Tasmania, recommended the continuation of Hare-Clark, with a change to seven-member electorates. In 1954 a Bill providing for seven-member electorates was introduced by Mr. L. V. McPartlan, Independent member for Denison. It passed the House of Assembly with the support of the Government, but was lost in the Legislative Council. Last year the House of Assembly Select Committee on Electoral Reform (as noted elsewhere in this paper) emphatically re-endorsed the Hare-Clark system and urged the adoption of seven-member electorates.

The friends and foes of Hare-Clark have come from both parties. Although the present Labor party Government is supporting the Hare-Clark system and recommending its improvement, the chief antagonist to the system is a Labor member (Dr. J. F. Gaha, M.H.A.), who acknowledges that he has been "an implacable enemy of the Hare-Clark system for many years and has not changed his views" ^(a).

One of the most ardent supporters of Hare-Clark, on the other hand, has been Mr. J. G. Breheny, M.H.A, a Liberal party member. When many others were attacking Hare-Clark in 1955, Mr. Breheny expressed his convictions without equivocation: "In no circumstances will I support the proposal to abandon the fairest and most democratic electoral system in the world to revert to the malpractice, injustice, and anomalies inseparable from the single electoral system with which electors have been so painfully familiar in the Labor states of New South Wales and Queensland for more than a quarter century."

In the same way that the Hare-Clark system is in itself impartial, so is the change from six to seven members per electorate. The plain fact is that the seventh seat in any electorate will go to whichever party polls the majority vote in that electorate. The winning of the five additional seats will, therefore, be decided strictly and solely by the electors within the respective five Commonwealth-State divisions. The "swinging" nature of all these divisions is shown in Table VIII. The margins between the two parties are close enough in all electorates for either party to consider that it has a good chance of winning 20-15 or 19-16 at the next election.

(a) From poll declaration speech reported in "The Mercury," Feb. 26, 1955.

1958. (No. 22)

27

On the basis of past returns and the known reliability of the Hare-Clark system, the pattern of the next election result can be foreseen with reasonable certainty: with a seven-member quota both parties can expect to win a minimum of three seats each, the fourth seat going to the larger party. In contrast, if 35 single-member electorates were used the outcome of the election could well depend not so much on the support given by the electors but on the distribution of the population in relation to boundary lines.

Seven-member Electorates could reduce Cost of Government.

A permanent continuation of 15 - 15 "dead heat" divisions of party representation can be expected on the basis of present voting trends. This condition not only causes electoral injustice (as explained elsewhere) but also promotes costly instability in government. The last State election, which would have been averted by seven-member electorates, cost ₤11,482 for minimum official administrative expenses alone. The expense of possible unnecessary elections resulting from six-member electorates could be greater even than the cost of salaries and expenses of five additional members.

Moreover, the stabilising of Parliament through elimination of 15 - 15 divisions in party representation could decrease materially the direct and indirect costs of government by making Parliament more effective. For example, the present even balance between the parties means that any one member in either party holds a potential balance-of-power influence over his party in Parliament. This interferes with the citizen's right, under the British tradition of responsible political parties, to expect parties to exercise responsibility collectively without undue influence by individual persons within the parties.

The small increase in House membership required for correcting the electoral machinery is more than warranted because of the increased work-load on Parliamentarians since the fixing of the present House membership over 50 years ago. As noted in Figure 1, the population of Tasmania has increased by more than two-thirds since 1907; with this increase in population has also come an even greater increase in the demands on members' time.

TABLE VIII.
Amount of "Swing" Required to Return Four Liberal Party Candidates in Each Electorate
On Basis of Figures of October, 1956, General Election ^(a)
1. Votes for Major Parties on Electorate Basis ^( b) :
	Bass	Braddon	Denison	Franklin	Wilmot	Total
Labor	16,556	15,302	16,837	14,566	16,835	80,096
Liberal	14,292	14,814	14,317	12,751	13,303	69,477
	30,848	30,116	31,154	27,317	30,138	149,573
2. Labor Majorities:
	2,264	448	2,520	1,815	3,532	10,619
3. Swing in Votes Required for Winning of Fourth Seat by Liberals^(c):
In Numbers	1,133	245	1,261	908	1,767	5,310
In Percentage	3.67	0.81	4.05	3.32	5.87	3.55

(a) According to first preferences. Because of the extremely limited amount of cross-voting the first preference figures are quite satisfactory for calculating the swing required.

(b) The preferences of Independent and Anti-Communist Labor Party candidates, which were not large in any case, were shared between the two major parties in so nearly equal amounts that it was not considered necessary to make adjustments for them in ascertaining the amount of swing- required. The sharing of non-major party preferences by the two major parties was in the Liberals' favor in Franklin by approximately 200 votes, and in Labor's favor in Denison by a similar amount. The Liberal gain in non-major party preferences in Braddon was approximately equal to Labor's gain from this source in Bass.

(c) These calculations assume that no Independent or minor party candidates will be elected. On the basis of the past two State elections this assumption is quite warranted.

(d) This calculation on a State-wide basis shows the amount of swing required for the Liberals to win all of the five additional seats. For them to win all seats except Wilmot, where the Labor majority is largest, would require a smaller swing.

                      (No. 22.) 1958.

28

Governmental services and activities hardly thought of 50 years ago are taken for granted now. Indications of the increase in governmental activity can be seen in the rise in expenditure and in the size of the public service. In 1907 annual appropriations by the Tasmanian State Government were £912,000; for 1957-58 £21,900,000 ^(a) In 1913 there were 376 permanent officers employed under the Public Service Act; as of March 1958 the number of corresponding employees is 2,326 ^(a) - more than a six-fold increase. These figures on public service employees do not include the large number of persons not covered by the Public Service Act, such as employees of the Hydro-Electric Commission, Metropolitan Transport Trust, the Railways and others. Expansion of the work in these areas represents a similar increase in the number of employees.

There is no question that the enlargement of government activity has been enormous and has placed correspondingly larger responsibilities on members of Parliament. At present there are not enough of them available for the proper study of public questions, which become increasingly specialised as years go by. Moreover, there are not enough members to offer satisfactory competition for appointment to Cabinet; now it is almost a matter of putting every available person into a job.

____________________________________________________________________________________________________________________________

(a) Figures supplied by the State Treasury Department, Hobart.

FIGURE 1.
Comparison of Sizes of House of Assembly, Tasmania, at Various Stages of Development in Respect of Population, 1856-1957.
Year	Size of House (a)	Population	Number of Head of Population Per Member of House of Assembly
1856	30	89,821
1870	32	100,886
1885	36	128,860
1893	37	154,895
1898	38	172,900
1901	35	181,832
1907	30	190,745
1954	30	308,752
1957	30	329,539
1957	35	329,539
	(proposed)
			Head of Population

(a) Data on size of House for years 1856-1907 is taken from E. L. Piesse, Bibliography of Proportional Representation for Tasmania, Hobart: Government Printer, 1913. p. 41.

(b) Population for years 1856-1907 are given for the years cited on basis of 5-year averages for the quinquennium of which the designated year is a part. The population for 1957 is an estimate as of 30th September, 1957; this date is the most recent best date for calculation purposes. All population figures are from publications of the Commonwealth Bureau of Census and Statistics.

                                   1958. (No. 22)

29

Figure 2.
Comparison of Sizes of Legislatures in the Leading English-Speaking Federations.
For States or Provinces with Populations under 1,000,000.
States or Provinces	Population	Size of Lower Houses (a)
		AUSTRALIA (b)
Tasmania	308,752
Tasmania	308,752
South Australia	797,094
West. Australia	639,771

CANADA (c)
Prince Edward	99,285
Newfoundland	415,074
Nova Scotia	694,717
New Brunswick	554,616
Manitoba	850,040

UNITED STATES (d)
Delaware	318,085
Nevada	160,083
Wyoming	290,529
Idaho	588,637
Utah	688,862
New Mexico	681,187
South Dakota	652,740
Arizona	749,587
Montana	591,024
Rhode Island	791,896
North Dakota	619,636
Maine	913,774
Vermont	377,747
New Hampshire	533,242

The fact that the proper working of Parliamentary government requires a certain minimum number of members was pointed out as long ago as 1914 by Mr. R.M. Johnston, I.S.O., Government Statistician, in giving evidence to a Select Committee considering electoral changes. To quote Mr. Johnston: "Five sevens would be better than 31. The number of representatives of any State, however small the population, should never fall below a practical working minimum number. Theoretically there ought to be not less than 35 representatives, even in a small State like this. The question of economy should not be a difficult matter" (a).

        Even in 1906 when the question of adopting the Hare-Clark system was being debated in the House of Assembly, there was criticism of the proposed move for reducing the membership of the House, then 35, to the present 30. In opposing this move, "The Mercury" stated in an editorial of September 26, 1906" "… The saving in annual expenditure would be trifling, and there is no reason to believe that among the smaller number would be found a greater aggregate of wisdom.

(a) Parliamentary Paper No. 31 of 1914, p. 6.

                                        No. 22. 1958.

30

       Moreover, with the probability of a serious crisis arising in consequence of the financial and other relations between the Commonwealth and the States, it seems to us a very inopportune time to do anything which may have the effect of weakening the power and influence of the Tasmanian Legislature".

For persons who may consider a 35-member House too large, Figure 2, showing the sizes of the smaller American States and Canadian provinces, will furnish interesting comparisons and contrasts.

Because the reasons for a 35-member House are obvious enough it may seem a waste of the readers' time to record even the above incomplete account. A reluctance however, by the public and by Parliament, to increase the membership of the House to 35 has prevented the establishment of seven-member electorates ever since 1912. Instability in Parliament and electoral injustice to the voters will continue unless seven-member electorates are adopted. Six-member electorates, even with the existing Parliamentary "deadlocks" law cannot guarantee majority rule.

* * * * * * * * * * *