This chapter draws together the findings to give an overview of the potential scale of change and opportunity facing different industrial sectors.
Table 8 summarises the potential for disruption and mitigation in different industrial sectors in London. Orange indicates high potential for disruption, yellow medium and green low. For mitigation, orange indicates low potential for mitigation, yellow medium and green high. In summary, it shows:
- The potential for automation, based on Frey and Osborne’s occupational analysis, translated to the London workforce.
- The level of reliance on EU labour, which may be disrupted by the impact of Brexit.
- The exposure of low-waged sectors to changes in regulation or market conditions.
- The potential for growth in demand for particular occupations unlocked by automation and other factors.
- The existing skills levels, and the flexibility they represent.
- The strength of clusters in London, and the resilience to economic change that these embody.
There are a number of sectors which seem to face particularly intense disruption: construction, retail, accommodation and food, and administrative and support services are all relatively poorly paid and highly dependent on a migrant workforce — as well as offering high potential for automation. As Brexit approaches and pressure on pay mounts, it is relatively likely that automation will become more widespread in these areas — at least for those tasks where no “engineering bottlenecks” (the need for advanced dexterity, creative intelligence and social intelligence) are present. On occasion, this may require redefinition of tasks — for example, separating the manufacture of standardised housing units or components from their assembly on complex and cluttered building sites. In construction, the issue may become acute more quickly, as the workforce is relatively aged: across the UK, around 20 per cent of construction workers are aged over 55. 22
That said, each of these sectors exhibits considerable variety within London, which cannot always be picked up through use of standardised statistics. While basic retail and food service functions are already being automated — as automatic tills and Internet shopping grow — London has a considerable concentration of specialist, often high-end traders and restaurants, where serving staff and shop assistants will not be as easily replaced as in fast-food chains, supermarkets or high street outlets.
In London’s shops and restaurants, relatively high qualification levels may be attributable not only to graduates “trading down”, but also the nature of work within the capital.
In other sectors, such as arts, entertainment and recreation, and financial, professional and technical services, the immediate potential for automation seems lower, while the potential for job growth seems higher and agglomeration effects seem more pronounced. However, there may be significant change within these sectors as more routine tasks are automated, enabling higher productivity and a shift of workforce to higher value-adding activities.
A special note of caution should be sounded with the financial and insurance sector. While the overall automation potential for people working in this sector is medium, some underlying occupations (especially associate professional roles) might be much more exposed, as future demand shifts to tech occupations rather than traditional finance professional roles. In 2000, more than 600 cash equity traders worked at Goldman Sachs’s New York headquarters. These have since been replaced by automated programs supported by 200 computer engineers — and two equity traders. This may turn out to affect other activities in investment banking too: Goldman has identified 146 steps taken in any initial public offering of shares, and “many are begging to be automated.” 23
Many transport occupations are associated with a particularly high risk of automation. At the same time, many jobs in this sector are relatively well paid and open to those who did not pursue further or higher education having left school. A sharp decrease in demand for these workers — if the speed of automation is as great as many expect it to be — may create a challenge on a similar scale to the decline of manufacturing employment in previous decades.
It is also worth bearing in mind that these disruption and mitigation factors have interdependencies, not only within but also between sectors. The interdependencies become clear when looking at the relationship between immigration, wages and automation. For example, higher barriers to immigration will reduce labour supply (particularly in low-wage occupations), potentially leading to higher wages (despite mixed results in research on this topic). 24 This might in turn encourage investment in capital to automate these jobs, but displaced workers moving to other sectors might then drive down wages there.