When analysed by language, chi-squared analyses indicated that the Tswana group was most likely to indicate that everyone should receive special help to find jobs and be reintegrated into society. The Xhosa speaking group was most likely to say that those who fought against apartheid should receive this kind of help and Afrikaans-speaking South Africans were most likely to indicate that nobody should receive help.

Analyses indicated that socio-economic classification was another variable showing a significant relation to responses to the question of whether people for whom the SANDF can no longer cater, should receive help in finding jobs and in reintegration into society. Squatters, respondents in rural areas and in the black townships were most likely to answer that everyone should receive help. Respondents in the former urban areas for white people were most likely to indicate that nobody should receive this kind of help, probably reflecting in part, a basic class difference in attitudes towards the issue.

ANC supporters are especially in favour of helping former soldiers to find jobs and with reintegration into society. Those respondents who had a positive view of the legitimacy of the Government of National Unity (GNU)5 felt that everyone should receive help (see Figure 2).

In another question on rationalisation, respondents were asked about retrenchment: "When the armed forces retrench people, who should be the first to leave?" Although many respondents were unsure (28%), the largest single proportion who expressed a viewpoint, indicated that those who are the least well trained (13%) and those who joined last (12%) should be the first to leave. Another reason that featured significantly was that those who are the least capable according to the command structure, should be the first to leave (11%). This was followed by "white soldiers" (8%). The first three reasons - late joining, least trained and least effective - shows signs of having connotations with labour relations practices.

UNIONISATION

Since the mid-eighties, the state has faced increasing pressure to grant its employees the same fundamental rights as the rest of society. Despite restrictions placed on trade union activity and strike action and the creation of alternative channels to address both collective and individual grievances, nearly all state departments have been affected by labour unrest, including the security forces.

In 1989, the Police and Prisons Civil Rights Union (POPCRU) was formed. Amid debate, members of the South African Police (now the South African Police Service) and the Department of Correctional Services embarked on protest action to gain the right to form and belong to trade unions. The first incidents of labour unrest in the Department of Correctional Services occurred at Johannesburg Prison during February 1990 and soon spread to other prisons thereafter. After a relative period of quiet, labour unrest again occurred in prisons in February 1993.6

Initially the SAP was not affected by labour unrest to the same degree as the Department of Correctional Services. According to Heinecken, "... the first occurrence of trade union activity coincided with the uprisings in Correctional Services in March 1990..."7 following which 39 policemen, believed to be linked to POPCRU, were fired under the Police Act for striking. On 21 August 1993 approximately 500 police assistants and prison warders of POPCRU, some armed, marched on SAP headquarters in Johannesburg. Action and counter-action escalated until 13 October 1993, when 88 protesting policemen were arrested for mutiny.8

Uniformed personnel of the military, police and correctional services have, until recently, been prohibited from belonging to trade unions. Towards the end of 1991, the SAP introduced a system of discussion forums to provide for members of the same rank and post to discuss work related matters. The Correctional Services instituted a work station discussion system for uniformed and civilian personnel, based on the principles of Quality Circles. The government acceded to the increased demand for public service unionisation by promulgating the Public Service Relations Act No. 102 of 1993. This legislation allowed public service employees to form trade unions, while restricting this right to civilian employees of the Security Departments. This meant that POPCRU was denied de jure recognition and labour matters affecting its members could only be addressed through ministerially approved structures as stipulated in the Police Act No. 7 of 1958 and the Correctional Services Act No. 8 of 1958. The fact that POPCRU and SAPU were there to stay and were gearing themselves for recognition set the stage for debate on unionisation in the Security Forces.

In September 1993 the Department of Correctional Services announced that uniformed members would also be included under the Public Service Labour Relations Act, although they were still denied the right to strike. The Police submitted an amendment to the Police Act that empowered the Minister to promulgate regulations allowing trade union activity and subsequently engaged in active attempts to form an employee body to represent policemen and women. The SADF resisted unionisation, although it did amend the Defence Act in September 1993 to allow members to join "... any professional or vocational institute, society, association or like body approved by the Minister."9

The Department of Defence has not been subject to the same degree of labour unrest as the rest of the security forces. Since 1994 and particularly after the acceptance of the new Public Service Labour Relations Act (1993) that has granted SANDF civilian employees labour rights, incidences of labour dissatisfaction have increased among both civilian and uniformed personnel. The most significant was the protest by 32 soldiers of 21 Infantry Battalion stationed at Lenz in August 1994, who were later convicted of mutiny. Although protesters have had meetings with senior officers, the Minister of Defence and even with the President, there is marked disapproval of their actions. The South African National Defence Union (SANDU) has recently attempted to challenge the restrictions in the Defence Act, prohibiting uniformed members from joining a union or striking, as being unconstitutional.

With regard to the above, respondents to the survey were asked: "In view of the fact that they perform an essential service and some people think the service they render should be uninterrupted, should people in the SANDF be allowed to form trade unions?" The largest single proportion of respondents (45%) indicated that they should be allowed to form trade unions. A sizeable proportion (40%), however, felt they should not.

Chi-squared analyses of the response according to population group revealed that 52 per cent of African respondents, 51 per cent of Asian respondents and 39 per cent of coloured respondents, representing the largest single proportion in each case, were in favour of allowing trade unions in the SANDF. Almost three-quarters of white respondents (73%) were not in favour of trade unions in the SANDF (see Figure 3). The proportion of respondents in favour of the SANDF allowing trade unions increased according to educational levels. Respondents with no or little education are the only group where the proportion of those in favour of the SANDF allowing trade unions is marginally smaller those against the SANDF allowing trade unions.

Analyses according to support for political parties indicate that respondents supporting the traditionally black political parties were in favour of allowing members of the SANDF to form trade unions: 61 per cent of PAC, 59 per cent of IFP and 53 per cent of ANC supporters expressed themselves favourably. Respondents supporting the traditionally white parties were against the SANDF allowing it. Supporters of the right-wing parties were most inclined to oppose the formation of trade unions within the SANDF, with 88 per cent opposed, followed by supporters of the NP (53%) and the DP (50%). The response should probably be seen as partly related to different attitudes regarding trade unions among black and white people in South Africa (see Figure 4).

Another question was: "In view of the fact that they perform an essential service and some people think the service they render should be uninterrupted, should the members of the SANDF be allowed to strike or demonstrate?" More than half (56%) of the respondents were against allowing members of the SANDF to strike or demonstrate. This probably partially reflects a fear that the country and its people will be unprotected and vulnerable to attack if SANDF members go on strike or demonstrate, or that internal conditions could deteriorate without the military being there to protect them, as was the case in Bophuthatswana and Ciskei before the April 1994 elections. Chi-squared analyses of these responses by population group indicated that more than half of the respondents from each population group were opposed to SANDF members being allowed to strike. Analysed by support for political party, respondents supporting right-wing parties (95%), the DP (79%) and the NP (70%) are the most likely to feel that SANDF members should not be allowed to strike or demonstrate.

Respondents were asked: "Do you think that working conditions for the members of the police service have improved or deteriorated since trade unions have been allowed?" The response revealed that 39 per cent of respondents indicated that conditions in the police service had improved or improved a lot since trade unions were allowed, while 24 per cent stated that conditions had deteriorated or deteriorated a lot. Analysed according to population group, data indicated that the largest single proportion of African (48%), coloured (33%) and Asian (36%) respondents expressing a definite viewpoint, claimed that working conditions in the police improved since trade unions were allowed. In the case of white respondents, by far the largest single proportion (46%) said that working conditions had deteriorated in that time. A feature of this response was the large proportion of respondents who were uncertain.

Chi-squared analyses according to support for political parties indicated that supporters of the traditionally black political parties were most inclined to say that working conditions in the police service improved since trade unions were allowed. Fifty per cent of ANC supporters, 43 per cent of PAC supporters and 52 per cent of IFP supporters, in each case the largest single proportion expressing an opinion, indicated that working conditions in the police improved since trade unions were allowed. On the other hand, the greatest single proportion of supporters of traditionally white parties indicated that working conditions deteriorated in the police since the advent of trade unions. Especially the supporters of the right-wing parties (53%), but also of the NP (33%) and the DP (29%) were inclined towards this view. Many respondents, however, were uncertain in their responses to this question.

Respondents were also asked: "Do you think the effectiveness of the SAPS has improved or deteriorated since trade unions have been allowed?" The largest single proportion of respondents (35%) felt that effectiveness improved or improved a lot, while 25 per cent thought it deteriorated or deteriorated a lot. Analyses indicated that African, coloured and Indian respondents were inclined to feel that the effectiveness of the police service had improved (respectively 43%, 30% and 39%). A mere seven per cent of white respondents thought that the effectiveness of the SAPS had improved while 51 per cent felt it had deteriorated since trade unions were allowed. A feature of the data was the large proportion of respondents who were inclined towards an ‘uncertain’ response. Unfortunately other reasons for improvement or deterioration during the same period were not tested (see Figure 5).

Chi-squared analyses of the data by support for political party of responses to the question of the influence of trade unions on the effectiveness of the police service revealed a substantial difference between those supporting traditionally black and white parties. The largest single proportion of supporters of traditionally black parties expressing an opinion stated that the efficiency of the police improved since the start of trade unions. The proportions of respondents inclined to this view were 46 per cent of ANC supporters, 29 per cent of PAC supporters and 37 per cent of IFP supporters. Supporters of the traditionally white parties took the opposite view, with only fifteen per cent of DP supporters, eighteen per cent of NP supporters and a mere four per cent of the supporters of the right-wing thinking the effectiveness of the police had improved since trade unions were allowed (see Figure 6).

The possibility of future unionisation in the security forces places the government in a position where it must produce acceptable forums for negotiation, mediation and arbitration as soon as possible, as it is only through the acceptance of alternative forms of bargaining that the possible embarrassment of prolonged security force labour unrest in the future can be avoided.

CONCLUSION

Although the questions included in the Omnibus Survey did not intend to be either exhaustive or definitive, the results pointed to important human resources factors that should be considered further. Attitudes towards these issues need to be considered when policy is formulated in future and may warrant further and more extensive surveys. The general public does not seem to be adequately informed about these issues and efforts should be made to involve all South Africans in discussions and debates.

Respondents expressed themselves unequivocally in favour of help for all soldiers being demobilised. A thoroughly planned programme of assistance and re-integration aimed at all soldiers being demobilised, appears to be required. In view of the preference expressed by respondents that all soldiers should be helped, continued planning for such assistance or preferential treatment only for those soldiers who fought against apartheid, may be unwise.

Although the largest single group of respondents expressed the view that members of the SANDF should be allowed to form trade unions, a sizeable proportion were against this. With more than half of the respondents against allowing members of the SANDF to strike or demonstrate, another form of personnel association should probably be allowed in the SANDF and a formal mediation and arbitration system announced. Moreover, the security forces need clear policies in respect of unionisation and labour matters.

The survey has shown that future human resources policies of the security services in general, and the SANDF in particular, will require careful planning, adequate negotiation involving all groups and, above all else, the realisation that South Africans will continue to expect the government to have dependable, disciplined forces available, when they are required.

ENDNOTES 

See the analysis and description of other areas in the survey by C. Schutte and J. Cilliers, Public Opinion regarding the South African Defence Industry, South African Participation in Peacekeeping, and Women in the Security Services, African Security Review 4(4), 1995, pp. 47-54.



Rural areas in former self-governing and TBVC states; squatters in former non-white urban areas (metro and non-metro); hostels and hotels; former urban areas for coloured people; former urban areas for Asians; former urban areas for Black people; former urban areas (non-metro) for white people; Former urban areas (metro) for white people; and rural areas, excluding the self-governing and TBVC states.



Detailed tables containing these figures are available at the Centre for Socio-Political Analysis of the HSRC.



Only statistical relations showing significance at the 1% level are included in relations considered by means of chi-squared analyses. Initially, dependent variables were analysed according to combinations of the following independent variables: gender, age, province, socio-economic category, an index of various social and economic indicators, marital status, educational level, employment, satisfaction with political situation, satisfaction with economic situation, feeling of security, perceived rise or fall in standard of living, favourable to attacks on the police, favourable to protest action, an index of the perceived legitimacy of the government, the government’s control over crime, improvements or not to personal life circumstances and the perceived manner in which the government has governed.



This variable was constructed out of the following sub-items: respondents’ rating of the government according to fairness, effectiveness, transparency, strength, honourableness and credibility.



L. Heinecken, The effect of industrial democracy and unionisation on the South African Security Forces, paper presented at the conference of the South African Political Science Association, Bloemfontein, 20-22 October 1993, p. 13-14.



Ibid., p. 14.



Ibid., p. 15.



Ibid., p. 25.