Calvin Jones reflects on the economic potential of automation and the implications for Wales
Debate on the socio-economic impact of artificial intelligence, robotisation and consequent automation is hard to ignore, with a number of international academic-scientific, policy and economic agencies weighing in on the extent, importance and timescale of this transformation. In the UK it even jostles, if only briefly, with Brexit as the catastrophe de jour. This debate is notable and important for many reasons, not least its focus on the implications of automation for different occupations, infrastructures and social classes, and because it is explicitly linked to wider social transformations, such as the implementation of a Universal Basic Income. What is also notable though is its (almost) complete a-spatiality; for example the economic debate is almost wholly about firstly, whether automation will decrease (massively, significantly or slightly) or perhaps increase the number of jobs that are viable for human undertaking; and secondly, about which occupations, industries and capacities will be rendered obsolete or more valuable.
There is little (but some and increasing) consideration of where such impacts will hit greatest, or where the ‘new’ jobs will arise, meaning that interested and concerned politicians can only outline the general challenge and opportunity, and with related policy relegated to little more than general, unmoored aspiration. This is vitally important, because previous economic transformations – for example the move to the service and property-based economy in the UK, or more fundamentally the original English industrial revolution – have resulted in wholesale shifts of economic opportunity across space, with often problematic consequences for places (both benefitting and left behind).
This stuff then likely matters for Wales; Future Advocacy have just made a decent stab at assessing how this will play out across the UK although I would contend their assumptions are somewhat questionable – not least that my own industry, education, is at low risk. So – what is my judgement about where Wales specifically stands in an automated world? We can draw some inference from existing studies and forecasts and the economic landscape of Wales, albeit always remembering the dangers in appearing too precise and knowledgeable in what is a very fluid transformation.
1. Process and routine activities, and related jobs, are most vulnerable to automation.
Whilst studies of automation vary hugely in their forecasts of job replacement, and consequently in their alarmism, there is a general agreement that processes that are repetitive and/or rules-based are easier for machines, computers and algorithms to do better, quicker, cheaper and more reliably than humans. There is I think a reasonable argument that Wales lost many of these jobs (in manufacturing) decades ago and that (some of) what is left is ‘high value’, requiring autonomy and complex-problem solving. Additionally, we must distinguish processes from jobs – which often include multiple processes that may be individually replicable, but not as a ‘bundle’. Nonetheless, the incursion of automation into process-based service activities is worrisome for Wales. Back-office customer service and financial/insurance occupations, the sort that Wales has been quite successful in attracting, would seem to be vulnerable to algorithmic replacement and the rise of the bots. And it does not end there; if I were a youngish audit accountant (or low-ranking contract lawyer) with a fat mortgage I would be frankly terrified.
2. Industrial Geography Matters
At the end of a decade of anaemic growth and continued uncertainty, Wales – and regions like it – are still at the wrong end of a number of potentially calamitous economic upheavals. The talk is of Ellesmere Port versus Sochaux or Bridgend versus Cologne, not Ellesmere Port versus Paris and Bridgend versus Dearborn. If it were ever Broughton versus Toulouse, I know where my fiver would go. A branch-plant, FDI orientation is limiting enough in ‘normal’ times, but likely to be more so when technology and related cost-changes drive the transformation of productive and distributive approaches – and indeed favour some consumers over others. Companies tend to keep R&D and innovative activities close to home, or if not, in globally-strong existing creative-industrial clusters. These activities, and top-level management more generally, are perhaps at the very end of the robotic firing line. When frostbite hits, it is the fingers and toes that go first.
3. Ownership Matters
Back in the middle of the 19th Century a bloke with a big beard (and his mucker, Engels) suggested that over time the incursion of machines into the production process would concentrate income and wealth into an ever-smaller segment of society – the owners of those machines – whilst making miserable the bulk of humans who didn’t own those machines, and whose income and prospects were negatively shaped by them. Marx’ general obstreperousness and radicalism has, for 150 years, largely overshadowed the basic correctness of his analysis. It is ever more apposite; if humans are pushed from the productive process ever more quickly then, in the absence of a social transformation on an arguably similar scale to that forecast by Marx, returns to capital rather than labour will become centrally important in supporting human welfare. The lack of large Wales-based corporates is not a great omen. Whilst regional figures on indirect ownership – for example via shares or pension funds – are difficult to obtain and interpret, it is likely that the poorer, Welsh population is a priori worse off than in other parts of the UK certainly. It is notable that for South Eastern regions of the UK, over half of GVA is already attributable to non-wage elements, whereas for the North and Wales, over half is waged returns to employment (Scottish is almost exactly 50:50).
4. Brexit, and the sort of Brexit, Matters
The interrelationship between Brexit and the incursion of automation is very hard to judge – it will likely be mediated on a company by company basis, especially in Wales. Of course, a ‘soft’, single-market Brexit would leave economic relationships largely unchanged, for better or worse, although with likely very negative impacts on science and R&D in Wales (such as it is). A ‘hard’ Brexit may accelerate processes of disinvestment in Wales that are automation-related or, somewhat analogous to the Trumpist rhetoric in the US, trade issues may get (or be given) the blame for what are mostly technological impacts. A slightly more positive reading would be that the UK (and its corporations), in the absence of easy free trade deals and (perhaps) lower international-capital flows might in a relatively isolated state be slower to adopt automated processes. This does seem rather an optimistic interpretation however, and would of course imply a set of UK firms that were less internationally competitive, with all the implications that has for tax receipts, welfare and employment more generally.
In summary, then, Wales is perhaps unlikely to enter any new industrial revolution at the (b)leading edge, or even, arguably, as one of the generic regional pack. Previous rounds of economic transformation should warn us of the dangers of being under prepared, under skilled and insufficiently autonomous at such times. This is not necessarily doom and gloom however; times of deep structural change give rise to opportunity as well as threat, for those that can identify and grasp it. Automation is no different. The shape and implications of a positive, transformative and progressive reaction to the ‘rise of the robots’ are, however deep, wide and complex. So perhaps best the subject of a future blog or for other, brighter folk.
All articles published on Click on Wales are subject to IWA’s disclaimer.