Corruption and confidence:

Public perceptions of corruption


After six years of post-apartheid governance, South Africa is no longer able to escape some searching questions about the extent of corruption in a wide variety of public sector institutions and to consider its harmful effects on state capacity or perceptions towards state institutions. A HSRC survey revealed that the government has a lot to do in rooting out corruption.

Data on the extent of corruption is, by definition, largely perceptual. The Human Sciences Research Council (HSRC) recently began to track public perceptions of corruption. Its most recent survey, conducted in November l999 amongst a representative sample of 2700 adults, asked three direct questions about the extent of corruption and government responses to it. They were:
  • Do you think corruption amongst government officials is increasing or decreasing?
  • How do you feel about the priority the government gives to ensuring clean and honest conduct by public officials?
  • Are you aware that public officials are regulated and bound by a code of conduct?

The extent of corruption

Just over half (53%) of the respondents indicated that they thought corruption within state institutions was increasing, while just less than one quarter (22%) thought it was decreasing. A rather large segment (25%) said they did not know.

Within all four population groups, an outright majority (or, in one case, close to a majority) of respondents argued that corruption was increasing. The figures ranged from 45% amongst blacks, 63% amongst coloureds, 81% amongst Asians, 84% amongst whites. Amongst blacks, 27% felt corruption was decreasing with a similar "do not know" figure. Male respondents were slightly more likely to feel that corruption was increasing than female respondents (58% to 50% respectively). Since this was the first time that this question was asked in this way, no clear trends over time can be drawn (Table 1).

Table 1 Do you think corruption amongst government officials is increasing of decreasing?
(%, rounded off)

Black
Coloured
Asian
White
All respondents
Increasing
45
63
81
84
53
Decreasing
27
15
9
4
22
Do not know
27
22
10
13
25
Total
99
100
100
100
100

Of more interest than these aggregated figures is the connection between the perceptions of respondents and their political party affiliations. Supporters of the United Democratic Movement, closely followed by the Democratic Party have the most negative perceptions. Interestingly, IFP supporters have the most positive attitude (Table 2).

Table 2 Attitudes of supporters of major parties: Is corruption increasing or decreasing? (%)

ANC
DP
IFP
NNP
PAC
UDM
Increasing
48
85
30
76
41
89
Decreasing
27
7
41
10
54
6
Do not know
26
8
28
15
6
5
Total
100
100
100
100
100
10

Government attempts to control corruption

This question was designed to elicit what might be regarded as prescriptive comments by respondents towards government actions.

The data appears to suggest that the balance of opinion amongst South Africans is that the government is doing enough to control corruption. Just over one third (34%) believe that the government is giving "sufficient" priority while another 17% believe that the government is giving it "too high" a priority. Just under one third (30%) believe it gets "too low" a priority to ensure clean and honest conduct by public officials. The remaining 18% report a "do not know" response (Table 3).

Table 3 How do you feel about the priority the government gives to ensuring
clean and honest conduct by public officials? (%, rounded off)

Black
Coloured
Asian
White
All respondents
Too high
20
10
11
5
17
Sufficient
40
26
23
13
34
Too low
20
43
58
72
30
Do not know
20
21
8
10
19

Predictable responses emerge between the different population groups. Just over seven out of ten whites (72%) believed the government gave too low a priority to ensuring clean and honest conduct by public officials. This dropped to 58% amongst Asian and 43% amongst coloured respondents.

Amongst black respondents, 20% said government efforts to ensure clean and honest conduct by public officials got too high a priority; 40% said it received sufficient attention while another 20% said it received too low a priority. The remaining 20% said they did not know.

On a party political basis, very little difference emerged on this point between the views of IFP and ANC supporters. Almost identical proportions within each party occupied the middle ground, at a shade under 40%. DP supporters were the most sceptical about government efforts to ensure clean and honest conduct by public officials.

Knowledge of government actions

This question was designed to move beyond broad perceptions to actual knowledge of the government’s anti-corruption efforts by asking respondents if they knew that the actions of civil servants were bound by a code of conduct. This code forbids corrupt practices among civil servants.

All government departments are required to develop specific anti-corruption mechanisms. At a more macro policy level both the Department of Public Service and Administration, and the Office of the Public Service Commission are intimately involved in developing anti-corruption policies.

At a national, aggregated level, the clear majority of respondents (61%) said they did not know of such a code while 39% said they did. Clear differences in responses between population groups emerged.

Table 4 would appear to explain one point about a clear variation in attitudes amongst population groups elicited to the earlier questions. For example, white respondents, who claim to be the most knowledgeable about the code, clearly believe that such a code makes no impact on the spread of corruption. In general the more a population group knows about the code, the less positive its evaluation of trends in corruption becomes.

Table 4 Are you aware that public officials are regulated and bound by a code of conduct? (%)

Black
Coloured
Asian
White
Total Group
Yes
33
48
51
62
39
No
67
52
49
39
61

This being so, the implications of this point become rather interesting. Would the evaluative perceptions of blacks change accordingly as knowledge of the code increased?

Conclusion

These perceptions have much wider relevance for South Africa’s consolidation of democracy than might appear to be the case. One point only need be noted here: that of the effect of corruption on public perception. The literature on corruption generally notes that pervasive corruption decreases the trust that persons have in state structures.

Corruption limits the extent to which state structures can deliver services in an impartial and honest fashion. If democratising states want to encourage participatory processes as a component of good governance — as South Africa does — then citizens’ incentives to engage with institutions will need to be increased by reducing the level of corruption in state institutions.

However, other data in the HSRC’s annual public opinion survey does not yet indicate any decline in levels of trust which could be linked to these broad attitudes towards corruption. The one exception is that of the attitudes of white South Africans, but these can probably be explained more by a general alienation than as a consequence of corruption.

This data does suggest that national attitudes might be rather precariously balanced. If this hypothesis is correct it means that all government agencies have to engage in clear and decisive action to root out corruption.

Richard Humphries
Human Sciences Research Council