I teach economics at Stellenbosch. I don’t consider myself the best teacher there’s ever been (this is backed by empirical evidence on rating scores and faculty rewards), but also probably not the worst either (purely anecdotal and experiential evidence). Some of my teaching is at the postgraduate level – Economic History – but most of it is undergraduate, from a textbook, and mostly replicable by any university that offer second-year Macroeconomics.
And, increasingly, also by online universities that now offer MOOCs: ‘massive open online courses’. Instead of listening to me explain the IS-LM model, students now have the opportunity to sign up for online lectures from the best macroeconomists out there, and listen to these when they want, where they want. MOOCs have been billed as the ‘end of universities as we know it’. See here, here and here. Instead of every university offering mediocre, entry-level courses, why not have one superprofessor at the best university in the world broadcast his lectures to a global audience? Broader access, high quality and for a much lower price.
This is of course part of what I think will be the key question for future generations: the extent which human labour can and should be replaced by robots. In Europe and the US, convenience store staff are being replaced by automated payment points. Instead of call centre operators, I now get calls from recorded human voices selling products (which, incidentally, is much easier to switch off). But let’s remember that technology can be both labour-substituting and complementary: while a harvesting machine will probably substitute a farmer’s need for many workers, a tractor makes a farmer more productive and its acquisition might even require the farmer to employ more people to expand production. To take an example closer to home: technological advancement (like more advanced computers and better statistical packages) augments my ability to do research; speak to any economist older than 50 and they will tell you how they used to use punch cards for regression analysis. Quaint, but extremely time-consuming.
So technological change creates winners and losers. Capital investment (accumulation as Adam Smith called it) make labour more productive, increasing productivity, wages, and therefore living standards. But technological improvement also changes the labour requirement: it often substitutes lower-skilled jobs for higher-skilled jobs. Here is the best example: Tesla’s new manufacturing plant really only requires people for the first 18 months of its operation (to programme the circuit). Thereafter, as soon as production begins, everything (apart from oversight) will be done by computers and robotic arms. Is this the future of car manufacturing? Will the workers in our vehicle manufacturing plants become obsolete? Probably.
This is of course of great concern if you live in a country with an unemployment rate of 40%, a country where every political party fighting in this year’s election hopes to win votes by ‘creating jobs’. What is the solution, then? To ban technological innovation and change? Of course not. To do so would lead to stagnation in productivity growth, meaning a decline in our ability to remain internationally competitive, meaning lower profits for firms, lower incomes and wages, and lower living standards. (Incidentally to the labour economists out there, won’t labour subsidies have the same effect?) But to stick our collective heads in the sand is also not a solution: workers that lose their jobs is not only a political dilemma, but a moral one too. And a difficult one. And then there’s still the additional (and bigger) problem of the 5 million South Africans that are actively looking for work, and another 4 million that would like to work but has stopped looking.
So who is likely to lose to technology? Is my job threatened by the arrival of these MOOCs? Probably not. I would argue that any service job that requires some socialising to be effective – and learning, especially in the social sciences, is a social activity – will be complementary to technology rather than being substituted by it. A course in pure mathematics or machine learning can perhaps be usefully taught online. But any course where context becomes important – where the past experience of the student interacts with the teaching material to create a unique product for all the other students – is unlikely to be substitutable by an online, one-way interaction. A new NBER Working Paper by Acemoglu, Laibson and List supports this view. They investigate the concern that internet-based educational resources (MOOCs) will be ‘disequalising’, creating superstar teachers and a winner-take-all education system that, for example, will put me out of a job. Instead, they find quite the opposite:
We contend that a major impact of web-based educational technologies will be the democratisation of education: educational resources will be more equally distributed, and lower-skilled teachers will benefit. At the root of our results is the observation that skilled lecturers can only exploit their comparative advantage if other teachers complement those lectures with face-to-face instruction. This complementarity will increase the quantity and quality of face-to-face teaching services, potentially increasing the marginal product and wages of lower-skill teachers.
And, I would argue, this is not only true for university professors, but for any service industry. Simple, replicable tasks will be outsourced to robots, but any task that requires context and interaction, and especially ones that require an emotive response, will require humans, and lots of them. The Economist listed a number of jobs that are likely to be replaced by robots. (I wrote about it here.) The top four jobs most unlikely to be replaced are recreational therapists, dentists, athletic trainers and clergy. Context, interaction, emotion. The four jobs most likely to be replaced are telemarketers, accountants and auditors, retail salespeople and technical writers. Simple, replicable tasks.
Technological advancement creates far more winners than losers. Users benefit (hello, smart phones). The inventors benefit (hello, Apple and Samsung). It creates thousands of new job opportunities (hello, app builders) and fantastic incomes (like the 55 people that just sold WhatsApp to Facebook for $19 billion. Even if each employee only had a 0.01% share in the business, that equates to R160 million each). We all become more productive because of it (you might be reading this post on your smart phone on your way to work). Yet the proprietors and employees of the previous technology suffer (Nokia, the South African Postal Service, or the messenger pigeon).
Not only is my job safe, but I will probably benefit from the MOOCs. This dynamic is also true in all other industries: Certainly, some jobs will be replaced by technology, but many others (even jobs that don’t yet exist, or that require only basic skills) will benefit from it. If our high level of unemployment is to fall, then we need to 1) focus our training on those skills that are complementary to the new technology, and 2) ensure that government policies do not obstruct the impact of technological change.