Internationally, the legitimacy of legal authorities is recognised as crucial for the state’s ability to function in a justifiable and effective manner. This applies, in particular, to the police. At National Police Day commemorations in Durban in January 2025, Minister of Police, Senzo Mchunu, placed a strong emphasis on building public trust in law enforcement agencies in the country and the need for the SAPS to work closely with communities to restore confidence in the police and achieve success in crime reduction efforts. The minister’s reflections drew attention to a persisting legitimacy challenge in the police-public relationship. Recent allegations of police corruption by KwaZulu-Natal (KZN) Police Commissioner Nhlanhla Mkhwanazi represents another decisive moment for policing in South Africa, the outcome of which will have a significant bearing on the project to rebuild public trust.

To provide further context to the extent and nature of this challenge, we examined representative survey data on trends in police confidence conducted through representative social surveys conducted by the Human Sciences Research Council (HSRC) since the late 1990s. Our research also outlined some of the variations and drivers of this policing perspective. It is our hope that this work will be used to design interventions to restore the public’s faith in the police.

The research

Views on crime and policing in the country have been a thematic priority in the South African Social Attitudes Survey (SASAS) series since its inception in 2003. This series is administered annually by the HSRC using face-to-face interviewing and has been designed to be nationally representative survey of the adult population aged 16 years and older. Each year, between 2500 and 3200 interviews are conducted countrywide. The data are weighted using Statistics South Africa’s most recent mid-year population estimates.

SASAS builds on earlier representative public opinion surveying at the Research Council that dates to the transition period of the early 1990s. On certain topics, this allows us to extend the period of analysis back to before the early 2000s.

National and provincial trends

The overall pattern of public confidence in the police over the 1998 to 2024/25 period is presented in the line graph in Figure 1.

Trust levels have remained relatively low over the course of this period. Not once during this 27-year interval did more than half the adult public say that they trusted the police. This suggests that the issue of police legitimacy is by no means a new one.

Over the 1998 to 2010 period, the average level of trust in the police was relatively static. It ranged between 39% and 42% in all but a few years.

This was followed by a sharp decline between 2011 and 2013, following the Marikana massacre of August 2012. However, confidence had almost returned to the 2011 level by the time of the 2015 survey round.

The 2016 to 2020 period was characterised by modest fluctuation between 31% and 35%. The hard COVID-19 lockdown experience, which including instances of police brutality in enforcing lockdown regulations, did not appear to have had an aggregate effect on confidence levels based on the 2020 survey results.

In 2021 public trust in the police dipped to a low of 27%. This appears to be linked to the July 2021 social unrest. Many criticised the poor performance of the police during the unrest. This was followed by a further 5 percentage point decline to 22% in 2022, with the 2023 and 2024/25 confidence levels almost unchanged, which may reflect increasing rates for certain crimes. The 2022, 2023 and 2024/25 figures are the lowest recorded in 27 years.

Figure 1: Confidence in the police, 1998-2024/5 (% trust/ strongly trust)

Figure 2 provides a slightly more nuanced picture, by overlaying the ‘distrust’, ‘neutral’ and ‘don’t’ know’ categories into the graph. The results show that patterns of trust and distrust were fairly similar in character over the 1998-2011 period. The shares expressing positive and negative sentiment were in a similar range, with regular cross-over effects and a slight lean towards the negative over the 2006-2009 period. After 2011, there has been a dramatic divergence in the pattern of responses, with distrust surging and trust collapsing to unprecedented lows. The shift to a clear pattern of net distrust is striking, peaking at 66% in 2023 and remaining high at 62% in 2024/25.

Neutral responses hovered around 12-19% over the years of surveying, with no significant long-term trend discernible. “Do not know” responses have consistently been very low (typically 0–2%), indicating most respondents held clear views on the police.

This graph reaffirms that has been a clear and sustained erosion of public trust in the police over the past quarter century, with distrust now 40 percentage points or approximately three times higher than trust. This suggests serious reputational challenges and growing alienation of the public from law enforcement authorities in the country.

Figure 2: Patterns of trust and distrust in the police, 1998-2024/5 (%)

Substantial provincial variation in trust in the SAPS underlies the national trend described above (Figure 3 and Appendix Table 1). Trust in the South African Police Service declined across all provinces over the past two and a half decades, with notable disparities underlying this trend. Over this period, the Eastern Cape and Northern Cape consistently recorded the highest levels of trust, averaging 43% and 42% respectively, and reaching peaks of 69% and 72% in the early 2000s. In contrast, Gauteng and KwaZulu-Natal have exhibited the lowest average trust levels (29% and 34% on average over the period, respectively), with both provinces falling to just 18% in 2024/25. Provinces such as the Free State and Limpopo, which once reported relatively high trust, have also experienced marked declines, reaching 23% by 2024/25.

By the time of the 2024/25 survey round, no province recorded trust levels above 30%, with the Northern Cape (30%), North West (29%), and Western Cape (28%) ranking highest, but still reflecting low confidence overall. In almost all cases, the trust figures recorded in provinces over the last few years are among the lowest recorded since 1998. While provincial levels and trajectories have varied, the trend points to a deepening legitimacy crisis for SAPS across the country, with historically higher-trust provinces converging downward toward the levels of deep scepticism observed in Gauteng and KwaZulu-Natal.

Figure 3: Provincial trends in police confidence, 2011-2021 (% trust / strongly trust)

Source: HSRC Evaluation of Public opinion Programme, 1998-2001; HSRC South African Social Attitudes Survey (SASAS), 2003-2024/25

Factors associated with police confidence

Based on SASAS survey evidence over the years of assessment, various factors have been shown to exert an influence on public trust in, and legitimacy of, the police in South Africa. These are briefly summarised below.

Experiences of crime clearly matter. Those who had been a recent victim of crime displayed significantly lower levels of trust in the police (an average 10 percentage point difference between 1999 and 2023/25). Fear of crime has a similar effect (average of 14 percentage point difference across the 1999-2024/25 period). Higher levels of fear are associated with lower trust in the police. This applies to classic measures such as fear of walking alone in one’s area after dark, as well as crime-specific fears (i.e., worrying about home robbery or violent assault). These associations have been found across multiple rounds of surveying.

Experiences of policing are also associated with confidence in the police. Negative police experiences by the public have a bearing on police judgements. Those reporting unsatisfactory personal contact with police officers expressed lower trust levels than those reporting satisfactory contact.

Well-publicised instances of police abuse or failure can also seem to reduce public confidence. A prominent example is the perceived ineffectiveness of the South African Police Service’s (SAPS) response to the 2021 social unrest. New SASAS data on attitudes towards the 2021 social unrest demonstrates that harsher assessments of the policing response to the unrest is indeed associated with lower levels of trust in the police.

Perceptions of police corruption also had a strong, negative effect on police confidence.

From a procedural justice perspective, past in-depth SASAS research has shown that the South African public strongly emphasises both police fairness and effectiveness as important constituent elements of their overall assessments of police confidence.

• Distributive unfairness: The more the police are seen to be treating South Africans unfairly, the more they are likely to view the police as untrustworthy.
• Procedural unfairness: Similarly, perceptions that the police treat people disrespectfully, are not impartial in their decision-making, or lack transparency in the actions (procedural unfairness) also undermine public confidence.
• Police effectiveness: Lastly, if the police are seen as ineffective (i.e., unsuccessful in preventing, reducing and responding to crime), this will also diminish confidence (Figure 4).

Figure 4: Association between police confidence levels and satisfaction with crime reduction efforts, 2015-2024/25 (% trust / strongly trust) HSRC SASAS 2015-2024/25

Democratic performance

Another factor that is influencing how the public views the police is broader evaluations of democratic performance and the trustworthiness of government. These are positively linked with police trust (Figure 5). Public confidence in democratic institutions has shown a strong downward trend over the past seventeen years, as part of general pattern of diminishing political support. Police confidence is not immune to the pull-down effect of these wider trends.

Figure 5: Association between police confidence levels and satisfaction with democracy, 2011-2024/25 (% trust / strongly trust) HSRC SASAS 2011-2024/25

Polishing the tarnished badge

From the above, it is evident that in a high-crime and socially-divided society, confidence in the police is make up of a mix of views on the fairness and effectiveness of the police, combined with experiences of crime (including worry about crime), experiences of policing, and even more general views on the functioning of democracy and government.

The risk is that low and diminishing confidence in the police, if left unchecked, will also continue to negatively shape views key elements of police legitimacy, such as a sense of shared moral values and the duty to obey the police.

What does the survey results suggest might need to be done to shore up confidence in the police in the country? The work of the Institute of Security Studies is essential here. Suggestions include:

• Dispensing with an excessively hierarchical culture,
  • Promoting competent and ethical SAPS leadership, as well as
  • Strengthening other parts of the overall system of police governance.

Key also is the implementation of a non-militaristic policing ethos. This should be framed around a service culture and use of minimal force. It also requires SAPS to put more measures in place to monitor and control the use of force, and promote a culture of police accountability.

These ideas warrant serious attention. They matter fundamentally for preventing further instances of police misuse of force, corruption among senior officials, and police ineffectiveness in handling crime. This is crucial for stemming and reversing the eroding confidence in the badge.

APPENDIX TABLE 1: Provincial trends in police confidence 1999-2024/25 (% trust / strongly trust)

	1999	2000	2001	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014
WC	43	32	40	49	37	38	47	39	34	47	42	37	30	25	32
EC	55	49	37	60	69	64	43	44	43	47	34	47	44	40	40
NC	62	50	50	72	60	53	43	45	39	46	45	45	47	32	40
FS	64	41	42	38	45	56	43	54	62	37	42	42	40	38	41
KZN	55	29	37	28	39	47	26	33	52	25	49	49	40	23	29
NW	49	31	50	46	54	45	30	29	41	56	37	37	31	23	41
GP	30	39	38	35	32	32	38	37	24	41	34	29	25	19	25
MP	41	31	47	30	50	48	50	45	41	39	46	43	41	45	41
LP	50	58	37	54	54	55	39	43	37	48	46	47	37	34	49
National	47	39	40	42	46	47	38	39	39	40	41	41	35	28	34
Lowest	30	29	37	28	32	32	26	29	24	25	34	29	25	19	25
Highest	64	58	50	72	69	64	50	54	62	56	49	49	47	45	49

	2015	2016	2017	2018	2020	2021	2022	2023	2024/5	Average (all years)	Range of values	Rank all years (1=high; 9=low)	Rank 2024/5 (1=high; 9=low)
WC	31	33	29	31	43	22	24	28	28	35	22-49%	7	3
EC	54	44	36	36	36	38	19	27	24	43	19-69%	1	5
NC	32	33	46	31	37	25	25	24	30	42	24-72%	2	1
FS	33	41	23	35	28	30	23	28	23	40	23-64%	5	7
KZN	41	36	37	28	30	27	13	17	18	34	13-55%	8	9
NW	35	31	31	20	25	35	26	19	29	36	19-56%	6	2
GP	33	26	36	29	31	19	19	16	18	29	16-41%	9	8
MP	39	32	37	39	47	50	28	24	26	40	24-50%	4	4
LP	42	40	41	37	37	21	40	28	23	42	21-58%	3	6
National	38	34	35	31	34	27	22	22	22	36	22-47%	…	…
Lowest	31	26	23	20	25	19	13	16	18	29	…	…	…
Highest	54	44	46	39	47	50	40	28	30	43	…	…	…

Source: HSRC Evaluation of Public opinion Programme, 1998-2001; HSRC South African Social Attitudes Survey (SASAS), 2003-2024/25