extract from: John Whyte 'How much discrimination was there under the Unionist regime, 1921-1968?', in Tom Gallagher and James O'Connell (eds.), Contemporary Irish Studies, School of Peace Studies, University of Bradford, Manchester University Press, 1983
Electoral practices
Three main allegations were made under this heading: that constituency boundaries for Northern Ireland parliamentary elections were gerrymandered; that the local government franchise was unfairly weighted in favour of the unionists; and that local government electoral boundaries were gerrymandered. Each will be discussed in turn. In a comprehensive examination it would also be necessary to discuss the franchise for parliamentary elections, since there were sporadic complaints about the provision of second votes for graduates of Queen's University and for owners of business premises; but this issue is rarely mentioned in the literature, and in any case was resolved by the Northern Ireland government when, off its own bat and in advance of the civil rights disturbances, it put through legislation abolishing these franchises in 1968.[2] I shall also omit elections to the Westminster parliament, because these took place under Westminster legislation and did not provoke charges of unfairness.
(a) Gerrymandering of parliamentary constituencies.
The first two general elections to the Northern Ireland House of Commons, in 1921 and 1925, were held under proportional representation, the forty-eight non-university seats being distributed between nine constituencies, with between four and eight seats each. (There was also a constituency for Queen's University graduates, with four seats.) In 1929 the Northern Ireland government carried an Act abolishing PR, except for the university constituency, and establishing forty-eight single-member constituencies, whose boundaries were laid down in the legislation. This rearrangement was attacked, at the time and since, as being unfair to the minority (e.g. Harrison, 1938: 80-1; Gallagher, 1957: 255-63).
This topic has recently been re-examined by three academics- two geographers (Osborne, 1979; Pringle, 1980) and a historian (Buckland, 1979). They find some grounds for criticising the effect on nationalist representation. Osborne, whose treatment is the most comprehensive, examines three counties where complaints of gerrymandering were made. He concludes that they were unfounded in Antrim, but possible in Armagh and definite in Fermanagh, where the nationalist majority gained only one seat out of three (Osborne, 1979: 48-53).
These authors agree, however, that such complaints can be exaggerated. Osborne (ibid.: 53) states that 'it is hardly possible to call the 1929 redistribution a general exercise in gerrymandering'. Even under PR, nationalists of all hues had gained only twelve seats; under the new scheme they continued to gain eleven. Indeed, Buckland, who has had the advantage of going through the government files on the redistribution scheme, finds that the nationalists were not intended to lose at all. The parliamentary draftsman imagined that he had left them with twelve seats, losses of one seat each in Antrim and Armagh being compensated by a gain of two seats in Belfast (Buckland, 1979: 241). A weakness in his calculations led to them making only one gain in Belfast.
Osborne, Pringle and Buckland focus their criticisms, not on what the 1929 reapportionment did to nationalists, but on its other effects. The abolition of PR generally weakens small parties, and the main losers in Northern Ireland were Labour, independent unionists, and other groups, who fell from eight seats in 1925 to four in 1929, although their share of the vote increased (Elliott, 1973: 89-90). The result was, in Buckland's words, 'to fossilise and stereotype attitudes and alliances in Northern Ireland' (1979: 226); or as Osborne puts it, the new constituencies 'coralled electors into neat areas of Nationalist or Unionist dominance', providing 'a structural prop to the communal divisions of Northern Ireland' (1979: 54-5). Uncontested elections became frequent, and, as Pringle shows (1980: 202), power within the unionist party shifted to a narrower and more affluent circle. In these ways, rather than by directly attacking the nationalists, the changes of 1929 harmed political life in Northern Ireland.[3]
(b) The local government franchise.
In Northern Ireland the local government franchise was not fully democratic. A small number of property owners, amounting to about one and a half per cent of the electorate (White, 1969), had more than one vote. A much larger number of adults, amounting to over a quarter of the parliamentary electorate in 1961 (Elliott, 1971: 792), had no vote. This was because the local government franchise was, with some exceptions, limited to owners or tenants of a dwelling, and spouses of such owner or tenant. Thus several categories of adult, such as lodgers, or grown-up children still living in the parental home, were disfranchised. These peculiar franchises were not invented by the unionist regime: they were part of the general law of the United Kingdom as it stood in 1921. But whereas in Britain they were abolished in 1945, in Northern Ireland they were retained and even strengthened. Nationalists argued that they were retained so as to reduce the anti-unionist vote (Gallagher, 1957: 238; CSJ, 1969: 13).
In their more extreme forms these claims were palpably false. A recent critic of the nationalist literature (Hewitt, 1981: 365) has no difficulty in disproving claims that a majority of those disfranchised were Catholics. But there is some ground for supposing that Catholics were over-represented among the disfranchised - as their lower average position on the socio-economic scale would lead one to expect. Gallagher (1957: 227-8) cites figures to show that in Derry city nationalists comprised 61 -6 per cent of parliamentary electors, but only 54-7 per cent of local government electors. A survey conducted in Belfast in 1967 showed that 51 per cent of local government electors identified with the unionist party, as against only 38 per cent of non-electors (Budge and O'Leary, 1973: 176)- though it is fair to say that the sample numbers were small, and that the authors themselves conclude that the bias revealed by their findings was 'so unsystematic as not to be the result of any concerted plan' (ibid.: 177).
On the whole, however, the nature of the local government franchise made only a slight difference to election results. Unionists held most local authorities by a majority so substantial that even if a change in the franchise cost them one or two seats, they would still retain control. An investigation by a Belfast Telegraph journalist (White, 1969) concluded that in only one local authority - Armagh Urban District - was unionist control so precarious that it could be overturned by a simple change in the franchise. This point was conceded by nationalists. As one source put it (CSJ, 1969: 13), a change in the franchise would be useless 'unless each vote is of equal value, in other words if there is no gerrymander'. The biggest effect of the unionists' insistence on retaining archaic franchises was that they made their opponent a present of the superb slogan 'one man, one vote'.
(c) Gerrymandering of local government electoral areas.
We come now to the principal complaint under the heading of 'electoral practices'. It is not denied that, even under the most favourable electoral system, nationalists could have controlled only a minority of local authorities in Northern Ireland. But the complaint is repeatedly made that the unionist regime gerrymandered the nationalists out of winning a large proportion even of those authorities which they could reasonably have hoped to control. In 1920 the British government had, as a safeguard to minorities north and south, introduced proportional representation for Irish local elections. But in 1922 the Northern Ireland government, as one of its earliest acts, abolished PR for local elections, and in 1923 it redrew the electoral boundaries in many areas. Following these changes, unionists made wholesale conquests of local authorities previously held by nationalists. Subsequent boundary changes won Omagh Urban District for the unionists in 1935, consolidated unionist control of Derry County Borough in 1936, and won Armagh Urban District for the unionists in 1946. The controversy revolves round the justice of the changes put through in 1922 and subsequently.
Here, as so often in Northern Ireland, truth has been obscured by exaggeration. Hewitt (1981: 363-5) shows that some of the more reckless charges of gerrymandering cannot be true. He also shows that even some of the more restrained complaints are vitiated by a failure to take account of the higher proportion of children in the Catholic population (ibid.: 365-7). In a number of areas a unionist council ruled over a population with a slight Catholic majority, and - if one assumes that a Catholic would never vote unionist, which in a border area was probably true - thus seemed prima facie to owe its control to gerrymandering. But when the proportion of adults of voting age is considered, some of these Catholic majorities disappear, and so the charge of gerrymandering becomes less plausible. Hewitt has even found two instances (Limavady and Ballycastle Rural Districts) where a nationalist council appears to have ruled an electoral majority of Protestants.
Another dubious yardstick sometimes used is to compare post-1922 election results with those obtained before the brief experiment with proportional representation in 1920. Several councils held by nationalists before 1920 were captured by unionists after 1922, as a result of the redrawing of electoral divisions. But unionists could reasonably claim that such redrawing was necessary. The pre-1920 boundaries derived from ones which had been drawn up for Poor Law guardian elections back in the 1 840s, since when there had been sweeping population changes, producing huge anomalies in the size of electoral divisions (Walmsley, 1959: 10; Buckland, 1979: 239). Thus when nationalist writers (e.g. Gallagher, 1957: 240-1; All-party Anti-partition Conference, 1950b) print maps comparing the pre-1920 and post-1922 boundaries, and take this as proof that the post-1922 arrangements were unfair, they make an unwarranted assumption. Maybe it was the pre-1920 situation that was unfair.
Another yardstick is available, however, which is less open to objection. This is to compare the electoral results obtained after 1922 with those obtained under PR in 1920. Even under PR there can be some discrepancy between the proportion of votes and the proportion of seats won by different parties, owing to the vagaries of electoral boundaries; but the discrepancy is much less than under the first-past-the-post system, and, since the 1920 boundaries were drawn up by the British administration, it is unlikely that they were stacked against the unionists. It would be unfair to compare the 1920 results with those immediately after 1922, because the nationalists boycotted many elections through the '20s, but even if we compare them with the results for the '30s and later, when nationalists were trying to win as many seats as the system would permit, the changes are startling enough. The following councils, which nationalists won under PR, were captured by unionists under the post-1922 electoral arrangements:
| Londonderry County Borough | Tyrone County |
| Fermanagh County | Enniskillen Urban District [4] |
| Cookstown Rural District | Dungannon Rural District |
| Lisnaskea Rural District | Magherafelt Rural District |
| Omagh Rural District | Strabane Rural District |
| Omagh Urban District (from 1935) | Armagh Urban District (from 1946) |
In addition, Beleek Rural District was captured by amalgamating it with unionist-controlled Irvinestown Rural District; and Castlederg and Downpatrick Rural Districts, which divided fifty-fifty under PR, became safely unionist under the post-1922 arrangements. This gives a total of thirteen councils formerly under nationalist control, and two evenly divided, which the unionists won. The nationalists were left with the following councils which they had won under PR and retained after 1922:
| Ballycastle Urban District | Downpatrick Urban District |
| Keady Urban District | Newry Urban District |
| Strabane Urban District | Warrenpoint Urban District |
| Ballycastle Rural District | Kilkeel Rural District |
| Newry No. 1 Rural District | Newry No. 2 Rural District |
To these they added after the war Limavady Rural District, which they had not held under PR. This made a total of eleven local authorities in nationalist hands out of seventy-three. Not only was this a smaller number than the unionists won from them after the abolition of PR, but they were less important. The post-1922 electoral changes cost the nationalists control of a county borough and two counties; the largest local authority left in their hands was Newry Urban District, with a population of 12,000. The change is startling enough to raise the strongest suspicions of gerrymandering.
The fate of Londonderry County Borough aroused the most bitterness. It had a substantial, and growing, Catholic majority - by 1961 Catholics were more than 60 per cent even among the adult population (Hewitt, 1981: 366). Yet unionists won back control under the ward division imposed in 1923, and when, after some years, it looked as if the nationalists might capture one of the unionist wards, the boundaries were redrawn so as to perpetuate unionist rule (Buckland, 1979: 243-6).
The stock unionist defence for the post-1922 arrangements (Walmsley, 1959: 9-10; UUP, 1969: 12) is that local government electoral boundaries were drawn so as to take account not only of population but of ratable value. This was justified on the ground that those who paid the most rates were entitled to the biggest say in the conduct of local government. Thus unionists, who were on average richer than nationalists, could legitimately find themselves more favourably represented. This, however, is a dubious defence. Democratic theory does not in general permit that the rich should be more strongly represented than the poor; the unionists themselves did not make such a provision in parliamentary elections. Furthermore, as time went by, it became less and less true that large ratepayers contributed the bulk of local government finance. Grants from the Northern Ireland government became increasingly important, until by 1969 they provided three-quarters of the revenue of local authorities (Cameron, 1969: para. 141). Thus, as the Cameron commission concluded (ibid.), 'such validity as this argument ever possessed is one which is rapidly losing any force which it might have had'. If electoral boundaries were drawn so as to over-represent the rich, this was not a refutation of the charge of gerrymandering: it was a description of how the gerrymandering was achieved.
In any case, attempts to defend the post-1922 arrangements crumble before Buckland's (1979) discoveries in official papers. He shows that the Northern Ireland government did not even attempt to be fair. The 'sole concern' of the Ministry of Home Affairs was 'how to give effect to the views of the Unionist rank and file' (ibid.: 233), and the reorganisation in controversial districts was 'virtually dictated by local Unionists' (ibid.: 239). The Derry redistribution of 1936 was designed by the Ministry of Home Affairs, who did the job better than the Derry unionists had been able to do it for themselves. 'Throughout the discussions between ministry officials and Londonderry Unionists there was never any question that the government should not assist the latter' (ibid.: 245). Buckland's conclusions are particularly weighty because his previous writings (on unionism in the period 1886-1921) had shown him as sympathetic to the unionist cause.
To sum up on electoral arrangements. Charges that parliamentary constituencies were gerrymandered against the nationalists have only slight validity, whatever other criticisms might be made of the effects of abolishing PR. The peculiarities of local government franchise were also of little effect. But when it comes to gerrymandering of local government boundaries, criticism is much more firmly based. Nationalists were manipulated out of control in a number of councils where they had a majority of electors. This is one of the clearest areas of discrimination in the whole field of controversy.