Sample characteristics
Consistent with other telephone surveys, our sample displayed above average socio-economic status relative to the South African population, but was still characterised by considerable financial hardship and poor health (Table 1).
The sample had above-average education levels (68% had matriculated, 49% had a tertiary qualification). Although respondents reported higher household incomes than many South Africans, with 47% of respondents living in households with monthly income above R5,000 (approximately 300 United States (US) dollars), hunger was prevalent in 19% of the respondents’ households and 57% reported their household received government welfare grants. 22% of respondents reported having a chronic condition and 30% presented with depressive symptoms.
The age and gender distributions of respondents was 55% (18-34), 29% (35-49), 9% (50-59) and 7% (60-plus); 47% of respondents identified as female. 47% of respondents lived in township or informal areas. Approximately 19% of the sample reported having had COVID-19, and 46% lived with a vaccinated person. The vast majority held religion as important or very important to them (89%).
Table 1 Summary statistics: individual and Household Characteristics
Sample Characteristics
|
|
n
|
Mean/%
|
|
|
|
|
Vaccine rollout age categories
|
|
|
|
Aged 18 to 34 (%)
|
|
3,558
|
54.7
|
Aged 35 to 49 (%)
|
|
3,558
|
28.8
|
Aged 50 to 59 (%)
|
|
3,558
|
9.2
|
Aged 60 or above (%)
|
|
3,558
|
7.3
|
|
|
|
|
Demographic characteristics
|
|
|
|
Age (years)
|
|
3,557
|
37.5
|
Female (%)
|
|
3,607
|
46.9
|
Has a matric certificate (%)
|
|
3,545
|
67.7
|
Has a tertiary qualification (%)
|
|
3,096
|
49.4
|
Earned money recently (%)
|
|
3,608
|
60.0
|
Has medical insurance (%)
|
|
3,556
|
24.0
|
Has a chronic condition (%)
|
|
3,541
|
21.5
|
Depressive symptoms present (PHQ > 2) (%)
|
|
3,555
|
30.1
|
|
|
|
|
Religion is important to me:
|
|
|
|
Not at all (%)
|
|
3,513
|
6.8
|
Unimportant (%)
|
|
3,513
|
4.3
|
Yes (%)
|
|
3,513
|
29.2
|
Very important (%)
|
|
3,513
|
59.7
|
|
|
|
|
Household characteristics
|
|
|
|
Household size
|
|
3,589
|
4.4
|
Hunger in the household (past week) (%)
|
|
3,588
|
19.1
|
Household receives a government grant (%)
|
|
3,477
|
57.3
|
Household income last month (R)
|
|
2,505
|
11,496
|
Household income >= R5000 (%)
|
|
3,082
|
47.4
|
Own a running vehicle (%)
|
|
3,578
|
46.2
|
|
|
|
|
Household lives in:
|
|
|
|
Traditional area (%)
|
|
3,573
|
14.1
|
Township/informal (%)
|
|
3,573
|
46.8
|
Formal residential (%)
|
|
3,573
|
30.3
|
Farm/smallholding (%)
|
|
3,573
|
8.8
|
|
|
|
|
COVID-19
|
|
|
|
I have had COVID-19 (%)
|
|
3,496
|
18.6
|
Lives with vaccinated person (%)
|
|
3,590
|
45.8
|
Notes. Weighted descriptive statistics of individual and household characteristics are reported, as mean values where appropriate (age, household size, and household income), or percentages. The presence of depressive symptoms is defined as a PHQ-2 score above 2. Chronic illnesses refer to any of HIV, lung condition, heart condition, high blood pressure, or diabetes. Examples given for government grants included the state old age pension, child support grant, and the COVID-19 social relief of distress grant. CVACS Survey 2. Authors’ own calculations.
|
COVID-19 vaccination intentions and associated factors
Study participants were mostly reluctant to vaccinate. Only 19% intended to vaccinate as soon as possible, 20% intended to wait and see, 25% only if required, and the majority, 37%, reported they definitely would not (Table 2). 56% did not think they would be vaccinated by May, with this figure at 86% for the “definitely not” group. (Kruskal-Wallis 𝛸2=661, p<0.001). 20% of the “as soon as possible” group thought they would get very sick with COVID-19 in the next year, compared to only 3% of the “definitely not” group (𝛸2=126, p<0.001).
Beliefs about infection risk, and vaccine efficacy and safety showed stark and consistently significant divergence by vaccination intentions, especially between the “as soon as possible” and “definitely not” groups. A majority (72%) of the “as soon as possible” group believed the vaccine prevented death from COVID-19, compared to only 11% of the “definitely not”s (𝛸2=494, p<0.001). Additionally, 41% of the “definitely not” group believed the vaccine could kill, compared to 12% of the “as soon as possible”s, a significant difference (𝛸2=163, p<0.001). Many more of the “definitely not” group (60%) thought the vaccine would harm people, compared to 13% of the “as soon as possible” group (𝛸2=171, p<0.001). Vaccination intentions were significantly associated with the belief that the vaccine is safe for the babies of pregnant or breastfeeding mothers (𝛸2=123, p<0.001), or safe for those with a chronic illness (𝛸2=92, p<0.001), but only a minority reported holding these beliefs (22% and 25% respectively). The majority perceived their risk of becoming very ill from COVID-19 as low, while a minority believed the vaccine is safe or effective. Little difference existed across the groups in terms of rates of chronic illness, with 21.5% in the “as soon as possible” group reporting suffering from a chronic illness, with a comparable 19% in the “definitely not” group.
Table 2 Differences in vaccination beliefs and attitudes across vaccination intention groups
|
|
Do you intend to get vaccinated?
|
|
|
Intentions group:
|
All
|
As Soon as Possible
|
Wait and See
|
If Required
|
Definitely Not
|
Z or χ²
|
Sig.
|
|
|
|
|
|
|
|
|
Percentage of sample
|
100
|
18.8
|
19.7
|
24.6
|
36.9
|
|
|
Number of observations
|
3608
|
680
|
712
|
828
|
1294
|
|
|
|
|
|
|
|
|
|
|
Will you be vaccinated by May?
|
|
|
|
|
|
χ² = 661
|
***
|
Yes (%)
|
29.0
|
77.4
|
31.6
|
25.3
|
4.5
|
|
|
No (%)
|
55.8
|
11.5
|
40.8
|
58.5
|
85.7
|
|
|
I don't know (%)
|
14.5
|
9.8
|
26.7
|
15.7
|
9.1
|
|
|
I will get very sick with COVID-19 this year:
|
|
|
|
|
χ² = 126
|
***
|
Yes (%)
|
8.0
|
20.2
|
7.4
|
6.6
|
3.3
|
|
|
No (%)
|
74.2
|
54.7
|
70.9
|
76.6
|
85.4
|
|
|
I don't know (%)
|
17.8
|
25.1
|
21.6
|
16.8
|
11.2
|
|
|
|
|
|
|
|
|
|
|
Vaccine Efficacy and Safety
|
|
|
|
|
|
|
|
The vaccine will stop me dying from COVID-19:
|
|
|
|
|
χ² = 494
|
***
|
Yes (%)
|
31.5
|
72.0
|
36.7
|
28.5
|
11.2
|
|
|
No (%)
|
61.0
|
22.7
|
52.2
|
64.2
|
83.7
|
|
|
I don't know (%)
|
7.5
|
5.3
|
11.1
|
7.3
|
5.2
|
|
|
I believe the vaccine can kill you:
|
|
|
|
|
|
χ² = 163
|
***
|
Yes (%)
|
28.1
|
12.2
|
19.4
|
29.1
|
41.3
|
|
|
It might (%)
|
29.7
|
25.2
|
36.1
|
32.6
|
26.2
|
|
|
No (%)
|
33.5
|
57.3
|
36.2
|
31.6
|
21.8
|
|
|
I don't know (%)
|
8.7
|
5.3
|
8.3
|
6.8
|
10.7
|
|
|
The vaccine will harm or keep people healthy:
|
|
|
|
|
χ² = 171
|
***
|
Healthy (%)
|
30.6
|
70.6
|
33.4
|
25.0
|
12.9
|
|
|
Neither (%)
|
12.7
|
8.0
|
16.9
|
13.9
|
12.2
|
|
|
It will harm (%)
|
40.4
|
12.7
|
30.9
|
41.8
|
60.3
|
|
|
I don't know (%)
|
16.4
|
8.6
|
18.9
|
19.4
|
14.6
|
|
|
The vaccine is safe for the babies of
|
|
|
|
|
|
|
|
pregnant/breastfeeding mothers:
|
|
|
|
|
|
χ² = 123
|
***
|
Yes (%)
|
21.8
|
43.8
|
25.8
|
17.7
|
10.7
|
|
|
No (%)
|
56.1
|
33.0
|
49.5
|
59.6
|
70.6
|
|
|
I don't know (%)
|
22.1
|
23.2
|
24.8
|
22.7
|
18.7
|
|
|
The vaccine is safe for chronic illness:
|
|
|
|
|
|
χ² = 92
|
***
|
Yes (%)
|
25.2
|
51.6
|
25.2
|
24.8
|
11.3
|
|
|
No (%)
|
52.8
|
28.2
|
48.2
|
57.1
|
66.0
|
|
|
I don't know (%)
|
22.1
|
20.2
|
26.6
|
18.1
|
22.7
|
|
|
|
|
|
|
|
|
|
|
Table 2 – continued
|
|
Do you intend to get vaccinated?
|
|
|
|
Intentions group:
|
All
|
As Soon as Possible
|
Wait and See
|
If Required
|
Definitely Not
|
Z or χ²
|
Sig.
|
|
|
|
|
|
|
|
|
|
|
Socio-economic status & chronic illness
|
|
|
|
|
|
|
|
|
Household income >= R5000 (%)
|
47.4
|
35.9
|
40.8
|
51.3
|
55.8
|
Z = -11.3
|
***
|
|
Has a chronic condition (%)
|
21.5
|
22.3
|
26.2
|
21.3
|
19.0
|
Z = 2.0
|
*
|
|
|
|
|
|
|
|
|
|
|
Percent responding yes to each potential
|
|
|
|
|
|
|
|
|
reason why the respondent is unvaccinated
|
|
|
|
|
|
|
|
God or the ancestors will protect me (%)
|
51.7
|
39.6
|
52.1
|
54.6
|
54.9
|
Z = -5.5
|
***
|
|
Vaccination site is too far away (%)
|
15.8
|
31.7
|
19.0
|
15.5
|
6.7
|
Z = 15.0
|
***
|
|
My body is strong enough to fight the disease (%)
|
64.6
|
49.0
|
62.3
|
68.2
|
72.9
|
Z = -13.5
|
***
|
|
My religious leader is against the vaccine (%)
|
8.7
|
7.9
|
8.7
|
7.7
|
9.9
|
Z = -1.6
|
|
|
No time to go get vaccinated (%)
|
22.7
|
46.4
|
26.9
|
22.7
|
8.0
|
Z = 18.6
|
***
|
|
My risk of being infected is very low (%)
|
48.7
|
36.3
|
46.7
|
52.3
|
54.7
|
Z = -5.5
|
***
|
|
|
|
|
|
|
|
|
|
|
Notes: Weighted descriptive statistics of beliefs relating to reasons for not vaccinating, vaccine efficacy, and safety in the full sample, and 4 vaccination intentions groups. These are respondents who intend to get vaccinated: as soon as possible, to wait and see, to get vaccinated only if forced, or to definitely not get vaccinated. Test statistics are Kruskal-Wallis Chi-squared or Mann-Whitney Rank Sum Z-scores. Significance levels are reported as follows: * represents 0.01<=p<0.05,** represents 0.001<=p<0.01, *** represents p<0.001. Authors' own calculations. CVACS Survey 2.
|
|
Vaccination intentions varied significantly by both socio-economic status and physical health. Only 36% of the “as soon as possible” group reported household income above R5000, compared to 56% of the “definitely not” group (Z=-11.3, p <0.001). The prevalence of chronic illness similarly differed by vaccination intentions, although in a more limited range: from 19% for the “definitely not” group, to 26% of those intending to “wait and see” (Z=2.0, 0.01<=p<0.05).
Reasons for not getting vaccinated
When asked why they were unvaccinated, and given the opportunity to choose multiple reasons, the top three reasons chosen by all respondents were a “body strong enough to fight the disease” (65%), protection from “God or the ancestors” (52%) and a “low risk of being infected” (49%). The opinions of religious leaders were not significantly related to vaccination intentions. Large and significant variation in all other reasons by vaccination intentions was observed, with access concerns highest among the “as soon as possible” group (but low in the full sample), and belief in low risk of COVID-19 infection and a strong immune system most prevalent among the “definitely not”s.
Belief in the protection of God or the ancestors as a reason for not vaccinating was significantly larger for the “definitely not” group at 55%, compared to the “as soon as possible” group (40%) (55%) (Z=-5.5, p<0.001). Two access issues, related to distance or a lack of time, were cited more frequently by the “as soon as possible” group (32% and 46% respectively) compared to the “definitely not”s (7% and 8% respectively, Z=15.0 and Z=18.6, p<0.001). Overall, a minority of all respondents reported distance or time as a barrier (16% and 23% respectively).
Belief in a strong immune system (“my body is strong enough”) was highest among the “definitely not” group (73%), but was also reported by 49% of the “as soon as possible” group. This belief varied significantly across intentions (Z=-13.5, p<0.001). Nearly 49% of the sample estimated their risk of being infected with COVID-19 at very low, and this measure varied significantly by vaccination intentions (Z=-5.5, p<0.001).
Main reason for not vaccinating: qualitative responses
The qualitative evidence showed diametrically opposed views on vaccination across the vaccination intention groups, with very little overlap (Figure 1). The top 5 most frequently cited reasons by the “as soon as possible” group (totaling to more than 50% of responses) were concentrated in access and health related issues, which were scarcely mentioned by the “definitely not”s. Not needing or not trusting a vaccine was cited nearly 8 times as much by the “definitely not” group as the “as soon as possible”s. Conspiracies and fear were more common among the “definitely not’ group, but these were also cited by the “as soon as possible” group too, although in much lower proportions.
For the “as soon as possible” group, logistical and access reasons were predominant (16% cite no time, 11% cite being sick, 9% and 8% cite general site access and health issues respectively, and a further 3% state the vaccine site was too far). In addition, reflecting incorrect beliefs about vaccine safety, reasons for non-vaccination included being pregnant or breastfeeding (10%) or having a chronic illness (3%).
For the “definitely not” group, the top 4 themes differed in frequency by less than 1 percentage point and reflected dissimilar motivations. The top reason, “I do not need a vaccine" (9%) and “No reason” (9%) indicate certainty in the decision not to vaccinate. “I don’t trust it” (9%), “Vaccine kills or harms people” (8%) or “Side effects” (7%), “Vaccine still being tested” (7%) or “Vaccine not 100% effective” (6%) or “It is a conspiracy” (4%), reflected a major portion of this group who were distrusting, afraid, or skeptical. A deep divergence in views was apparent across intentions groups - the most frequent theme (“No time”) cited by 16% of the “as soon as possible” group did not feature in the 30 most frequently observed themes among the “definitely not” group. Similarly, the most frequently reason cited by the “definitely not” group (“I do not need a vaccine”, 9%) was only the 15th reason (1%) for the “as soon as possible” group. Similar differences can be seen for beliefs in vaccine efficacy, fear of harm from the vaccine, and other themes. 4 themes cited by the “definitely not” group were not mentioned at all by the “as soon as possible” group (“NA")
Mistrust and mandates
Trust in government information on COVID-19 was low, and this lack of trust differed significantly by vaccine intentions (Table 3). 40% did not trust information on COVID-19 from the government at all, with the level of mistrust largely and significantly varied across the intentions groups, ranging from 10% in the “as soon as possible” to 64% in the “definitely not” group (Kendall’s 𝜏=-0.41, p<0.001). Support (including strong support) for mandates was low (25%), and varied significantly, ranging from 59% for “as soon as possible” to 9% for “definitely not” (𝜏=0.35, p<0.001), despite that 48% thought mandates would work fairly well or very well. These beliefs in the potential effectiveness of mandates differed significantly by intentions group (𝜏=0.28, p<0.001), and was lowest (29%) among the “definitely not” group.
Table 3 Differences in trust and attitudes to mandates across vaccination intentions groups
|
|
Do you intend to get vaccinated?
|
|
|
|
|
|
|
As Soon as
|
Wait
|
If
|
Definitely
|
Kendall's
|
|
|
|
Intentions group:
|
All
|
Possible
|
and See
|
Required
|
Not
|
Tau
|
Sig.
|
|
|
|
|
|
|
|
|
|
|
|
|
How much do you trust information
|
|
|
|
|
|
|
|
|
|
on COVID-19 from the government?
|
|
|
|
|
|
τ = -0.41
|
***
|
|
|
A lot (%)
|
22.9
|
54.8
|
25.2
|
18.4
|
8.9
|
|
|
|
|
A little (%)
|
34.4
|
34.0
|
44.0
|
40.6
|
24.5
|
|
|
|
|
Not at all (%)
|
40.0
|
9.5
|
28.6
|
37.5
|
63.8
|
|
|
|
|
I don't know (%)
|
2.7
|
1.7
|
2.1
|
3.4
|
2.8
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
How much do you support a mandate?
|
|
|
|
|
τ = = 0.35
|
***
|
|
|
Strongly support (%)
|
11.2
|
32.5
|
10.6
|
6.6
|
4.1
|
|
|
|
|
Support (%)
|
13.6
|
26.9
|
16.0
|
14.9
|
4.8
|
|
|
|
|
Oppose (%)
|
23.2
|
18.1
|
26.6
|
23.8
|
23.9
|
|
|
|
|
Strongly oppose (%)
|
49.9
|
21.0
|
45.4
|
52.1
|
65.9
|
|
|
|
|
I don't know (%)
|
2.1
|
1.5
|
1.4
|
2.5
|
1.3
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
How well will a mandate work?
|
|
|
|
|
|
τ = = 0.28
|
***
|
|
|
Very well (%)
|
25.4
|
48.9
|
29.4
|
25.6
|
11.7
|
|
|
|
|
Fairly well (%)
|
22.3
|
23.6
|
25.1
|
25.8
|
17.7
|
|
|
|
|
Not at all well (%)
|
46.3
|
21.3
|
40.3
|
43.3
|
65.2
|
|
|
|
|
I don't know (%)
|
6.0
|
6.2
|
5.3
|
5.4
|
5.3
|
|
|
|
|
Weighted descriptive statistics of beliefs relating to trust in government COVID-19 information, and mandates in the full sample, and 4 vaccination intentions groups. These are respondents who intend to get vaccinated: as soon as possible, to wait and see, to get vaccinated only if forced, or to definitely not get vaccinated. Tests of association between vaccination intentions and these beliefs are reported from Kendall's Tau tests. Significance levels are reported as follows: * represents 0.01<=p<0.05,** represents 0.001<=p<0.01, *** represents p<0.001. Authors' own calculations. CVACS Survey 2.
|
|
|
|
|
|
|
|
|
Main reason for strongly opposing a mandate: qualitative responses
788 and 1,665 respondents opposed or strongly opposed mandates. More than half of the 1,665 respondents strongly opposed to mandates gave the reason that they felt it was the individual’s right to decide to be vaccinated. Fears of harm from the vaccine, and feeling that vaccines were not needed or were not effective also emerged as respondent themes, but to a much lesser degree (see Table 4).
Table 4 Reasons behind lack of support for mandates
Why do you feel that way about mandates?
|
%
|
|
|
It is the individual's right to decide
|
54.1
|
Vaccine(s) are not needed
|
5.2
|
Fear of harm from vaccines/heard negative rumours
|
4.7
|
Vaccines are not (100%) effective
|
4.6
|
President or Government said vaccines wouldn't be mandated
|
4.6
|
Mandates limit freedom
|
3.3
|
Vaccines developed too fast/not tested/lack evidence or data
|
2.2
|
Do not believe in COVID-19 or the vaccine
|
1.8
|
Distrust information on COVID-19 or vaccines
|
1.6
|
Too sick/can't get the vaccine
|
1.4
|
Mandates/vaccine are a conspiracy
|
1.3
|
Mandates do not consider religious/cultural differences
|
1.3
|
Vaccines are unfair if used as a condition of employment
|
1.2
|
Distrust in government/political motivation
|
1.1
|
People lack (correct) information about vaccines/COVID-19
|
1.0
|
Vaccines discriminate against and stigmatise the unvaccinated
|
1.0
|
Mandates will not work
|
0.5
|
Government should prioritize other things
|
0.2
|
|
|
N
|
1,665
|
CVACS Survey 2. Thematically coded main reasons given to open ended question about why the person is strongly opposed to mandates. 6.5% of responses are uncategorised, and a further 1.5% had no reason or did not know. 0.7% of these respondents showed a misunderstanding of the question and gave answers in support of mandates.
|
The remaining reasons given by respondents were scattered over several themes (with no clear concentration in any one category), although many were very similar to the reasons given for not vaccinating among the “definitely not” group in Figure 1. 5% thought vaccines were not needed, nearly 5% feared harm from vaccines, and nearly 5% thought vaccines were not 100% effective. Referring to a previous speech by President Cyril Ramaphosa, 5% were strongly opposed as the president had said vaccination would not be mandated. Mandates limiting freedom, and vaccines being insufficiently tested were the last meaningfully substantial reasons given (3% and 2% respectively).