494 PROCEEDINGS OF SECTION G. 



the last count, and there may be several candidates elected or re- 

 jected at the last count. A slight extension of the foregoing 

 reasoning leads us to the following results: — * 



The candidates may be divided into four classes — 



(a) Candidates elected before last count; 



{b) Candidates elected at last count, 



(c) Candidates eliminated before last count; 



((I) Candidates rejected at last count. 



All first preference votes given to group (h) have an " eflfective 

 weight " unity. All first preference votes given to gi'oup (d) have 

 an " effective weight " zero. 



Cases (a) and (c) must be further considered. Take first a 

 candidate in group (a). If the Droop quota and the Gregory 

 system of transfer are employed, then the votes are transferred in 

 a packet, the fractional value of whose papers is F where F =z — 7.- 

 V being the number of votes polled and D the- Droop quota. If 

 ■r of these ultimately find their way into group (d), they are 

 wasted, and we get the "effective weight" of a first preference 

 vote in group (a) is 1 — - 



Lastly, consider case (c). Here the votes are transferred with 

 undiminished value, consequently the result is similar in form to 

 the last, F having the value unity. , The " effective weight," 

 therefore, of a first preference vote in group (c) is 1 — - 



It may be noticed, in passing, that if a candidate in group (a) 

 just obtains a quota on the first count, then F = 0, and the 

 " effective weight " of his first preference .votes is unity. 



The Hare system thus destroys the principle of " one vote one 

 value," which is an essential factor of a fair electoral system. 

 This is a fundamental defect entirely different in kind from faulti- 

 ness in some practical detail, like distribution in quota excess. 

 It is a defect, moreover, which cannot be remedied, since it is im- 

 possible to say beforehand the order in which the candidates will 

 be elected or eliminated. The Hare system, therefore, cannot be 

 accepted in itself as an entirely satisfactory substitute for the 

 system which it was proposed to displace. lb has made, however, 

 one noteworthy contribution to the problem under discussion — 

 ,j»he principle of large electoral areas. 



The comparative merits of the Hare and Droop quotas have 

 been the subject of frequent discussion. They both, under certain 

 «ircumstances, lead to disproportionate representation between 

 parties — the Hare quota because of the wastage of votes on suc- 

 cessful candidates, and the Droop quota through causes inherent 



