1. Dr Ben Roberts, Chief Research Specialist, Human Sciences Research Council
2. Ms Jarè Struwig, Chief Research Manager, Human Sciences Research Council
3. Prof. Narnia Bohler-Muller, Executive Director, Human Sciences Research Council
4. Dr Steven Gordon, Senior Research Specialist, Human Sciences Research Council
5. Ms Thobeka Zondi, Ms. Samela Mtyingizane and Mr. Ngqapheli Mchunu
Doctoral researchers, Human Sciences Research Council
Today marks the beginning of the 4IRSA Digital Economy Summit, which aims to bring together different government, market and other stakeholders to constructively engage in discussions about the opportunities and pitfalls that rapid technological change – the so-called Fourth Industrial Revolution – is likely to present in the South African context. It will also serve as a key forum for identifying the kinds of interventions required to minimise the human costs of such societal developments.
Against this background the Human Sciences Research Council’s Democracy, Governance and Service Delivery (DGSD) research programme developed and fielded a module of experimental questions on public attitudes towards the Fourth Industrial Revolution as part of its 2018/19 annual round of the South African Social Attitudes Survey (SASAS). This forms part of a larger programme of SASAS research on the public understanding of science.
This nationally representative series has been designed to examine the underlying values and preferences of South Africans aged 16 years and older. The survey was conducted between November 2018 and February 2019, and consists of 2,736 respondents living in private residence. The data are weighted to be representative of the adult public, and are benchmarked to the latest Mid-Year Population Estimates produced by Statistics South Africa.
What follows is an overview of a selection of high-level findings emerging from the 4IR survey module.
A considerable 4IR knowledge deficit still exists
The survey asked South Africans to self-evaluate their general level of knowledge of new technologies and scientific advances that are occurring. Close to three-fifths (58%) expressed moderate to high levels of knowledge, while 39% reported that lacked or had a circumscribed understanding of such matters (Fig. 1). To further assess levels of knowledge, respondents were asked about their levels of familiarity with the following three concepts: ‘social media platforms such as Facebook, YouTube and Twitter’, ‘artificial intelligence (AI)’ and ‘driverless cars’.
While half the adult public said they knew enough about social media to explain it conceptually to someone else, slightly less than a third of adults were able to do so in relation to AI (18%) and driverless cars (17%). In the latter instances, close to half the adult public had not even heard of the concept. A total of 47% believed that they could not explain any of the three concepts, 34% could explain one concept, 9% two concepts and only 11% reported that they could explain all three. These findings imply that between 40-50% of the adult public lacks a basic knowledge of 4IR technologies.
Room for improvement in computer and Internet technology skills
To assess the self-rated proficiency of South African adults in the use of computer and Internet technologies (ICTs), survey participants were asked whether they consider themselves to be sufficiently skilled (i) in their daily life and (ii) to do their job (where applicable). A total of 59% felt that they possessed adequate skills levels generally, while 51% of employed South Africans regarded their ICT proficiency to be of a suitable for their particular jobs. Conversely, 35% and 52% respectively felt that their ICT skills level was not adequate for the workplace.
South Africans display a moderately positive view of the impact of digital technologies
In terms of the envisaged impact that recent computer and Internet technologies currently have on the economy, society and personal wellbeing, South Africans could be considered cautiously optimistic (Fig. 2). Around half of South Africans (48-52%) believe that such technological advances are beneficial in economically, societally and personally, whereas approximately a fifth express reservations (18-21%).
Rise of the robots: A more sceptical view of the likely impact of the 4IR on the labour market
In South Africa, we find that there is broad-based recognition among the public that automation will have a bearing on the workplace, and a sizeable majority of the employed are concerned that it will affect them (Table 1). In the years to come, automation is expected to have an appreciable effect on the labour market, with robots and AI able to perform certain types of roles currently performed by people. Three-quarters (73%) of South African adults believe that in the next 10 years machines or computer programmes will assume many of the jobs presently done by humans. In addition, we find that 6 in 10 workers (62%) are very or quite worried that such automation will threaten their own job position. In comparison with findings from the 2017 round of the British Social Attitudes Survey, we find that South Africans exhibit almost equivalent views to Britons on the likelihood of automation impacting on the labour market (73% vs. 75%), though local workers demonstrate vastly higher levels of worry about the personal job impact of automation than is evident in the UK (62% vs. 10%).
There are signs of an age effect underlying views on the labour markets impact of automation. Those aged 16-25 years are most likely to report that automation will replace many existing jobs (78%), while those aged 56-65 and 66 years and older present a slightly lower degree of conviction in this regard (64%). The age-based variation is more distinct in the case of worry about automation threatening one’s personal job security. South African workers aged 55 years and below all show higher levels of worry than older age groups (61-70% vs. 46-49%), with the highest reported level of concern evident again among those aged 16-25 years. It is interesting to note that the lower worry among workers in the oldest age groups is due in part to a higher level of uncertainty about the job impact of automation and not exclusively due to lack of concern.
Low levels of cultural acceptance of technological change
In order to gauge how culturally accepting South Africans are of technological change, we asked respondents to rate how uncomfortable or comfortable they felt with four situations involving the use of robots, namely: (i) having a medical operation performed on you by a robot; (ii) factories where workers are replaced by robots; (iii) receiving goods delivered by a drone or a robot; and (iv) being driven in a driverless car or taxi in traffic. They provided scores using a scale from 1 to 10, where 1 means ‘totally uncomfortable’ and 10 ‘totally comfortable’ (Fig. 3).
On average, the South African public is not especially accepting of any of these four scenarios, with only 14-18% expressing comfort (opting for a value of 7 through 10 on the scale), 19-23% stating they would be moderately comfortable, while 59-67% were uncomfortable with such propositions. The mean scores on the scales ranged from a low of 3.33 out of 10 in the case of medical operations performed by robots to a high of 3.77 in the case of deliveries by drone. The latter are well below the midpoint of the scale.
Perceived ability of government to manage labour market effect of new technologies
In the face of appreciable concern over potential threat posed by the 4IR to jobs in future, how confident is the public that the government can intervene successfully to minimise adverse impacts of new technologies on the labour market? As Figure 4 shows, a little more than a third (37%) were very or fairly confident that government could effectively put in place strategies to ensure that new technologies do not result in job losses, whereas 57% were doubtful. Poorer South Africans tend to be more sceptical than better off South Africans (based on a subjective poverty measure). While the survey results do not address the role to be played by other actors (especially market actors), it does nonetheless provide a sense of views on state policy to address any adverse labour impact that automation may have in coming years.
Conclusion
This high-level overview of attitudes towards the fourth industrial revolution in South Africa suggests that South Africans are generally more optimistic than circumspect about the impact of the newest digital technologies on society, the economy and on their personal quality of life. However, knowledge of these technologies still remains quite circumscribed, and a sizeable minority share of the adult public feels it does not have the necessary level of skills to take advantage of the these digital technologies in their daily lives or in their jobs. Importantly, there is recognition that automation is going to affect the labour market and there are signs of appreciable concern over the threat this poses to one’s own employment.
There is also a limited level of comfort with robots performing a range of tasks. In particular, less than a fifth would be comfortable with a robot performing a medical procedure, driving them in a driverless car or taxi in traffic, or having a drone deliver goods. This points to quite low levels of acceptance of the application of robots in specific situations.
If technological change creates further polarisation and inequality in the labour market and sustained reductions in human employment, then carefully planned social and labour market policies will be required to address low pay, precarious employment, and expanded, long-term unemployment. At this stage though, the public is fairly pessimistic about the ability of government to minimise the human costs of the fourth industrial revolution. Ongoing multi-sectoral dialogue will clearly be required to promote new insights and develop actions and responses to fourth industrial revolution. This will need to be backed up with effective communication campaigns to inform the public about technological change and the planned response for the country.