For most of industrial history, mechanisation displaced workers from specific tasks while simultaneously generating new categories of employment that absorbed the displaced. The power loom eliminated the handloom weaver, yet the Industrial Revolution created the mill operative, the railway engineer, and the factory supervisor. Economists have long invoked this historical pattern — sometimes associated with the "lump of labour fallacy" — to dismiss fears of technological unemployment as misguided: technology destroys jobs at the task level while expanding the economic pie sufficiently to produce new occupations that previous generations could not have imagined. The implicit assumption is that human adaptability, supported by education and labour market flexibility, will always find new roles for displaced workers.

The contemporary wave of automation challenges this comforting historical analogy in several important respects. Earlier mechanisation overwhelmingly substituted machines for physical, routine labour — the cotton picker, the assembly line worker, the typesetter. The new wave, driven by artificial intelligence, machine learning, and advanced robotics, penetrates cognitive labour: legal research, medical diagnostics, financial analysis, translation, and creative content generation are all being performed with varying degrees of competence by algorithmic systems. When automation moves up the skills ladder and begins displacing knowledge workers — historically the beneficiaries of technological change — the presumption that education alone constitutes a reliable hedge against displacement becomes difficult to sustain.

Labour economists have documented a phenomenon known as labour market polarisation: the hollowing out of middle-skill, middle-wage occupations — bookkeepers, data entry clerks, paralegals, radiographers — that are precisely codifiable and thus susceptible to algorithmic replacement, while employment at both the high-skill, high-wage end (software engineers, surgeons, strategic consultants) and the low-skill, low-wage end (cleaners, carers, delivery drivers requiring physical dexterity and social judgement) has expanded. The resulting "hourglass" labour market creates distributional consequences that aggregate productivity figures systematically obscure: the median worker may face stagnant wages and precarious employment even as GDP and corporate profits reach historic peaks.

The policy response to automation has been conspicuously inadequate. Active labour market policies — retraining programmes, wage insurance, portable benefits — remain chronically underfunded in most advanced economies relative to the scale of displacement anticipated by credible forecasting models. More structurally, the tax system in most OECD nations was designed for an economy of stable, full-time employment: corporate tax revenues, payroll taxes, and social contributions all presuppose wage employment as the primary vehicle of value capture and redistribution. As automation substitutes capital for labour, the tax base erodes precisely when the social demands on the state — for retraining, income support, healthcare — intensify. Proposals to tax robots, expand earned income credits, or introduce a universal basic income have circulated in policy discourse for years but have faced fierce resistance from business interests and fiscal conservatives alike.

What the current conjuncture demands is not a choice between technological progress and human welfare but a serious reimagining of the social contract. The productivity dividends of automation are real: when machines perform cognitive drudgery, they theoretically free human attention for richer, more creative, and more fulfilling work. Realising that potential, however, requires deliberate institutional design — not the laissez-faire assumption that markets will organically distribute the gains. Historically, the benefits of transformative technology have accrued disproportionately to capital owners in the short to medium term, and only reached labour through sustained collective bargaining and political pressure. There is little reason to believe the current transition will be different in its distributional dynamics — unless institutional frameworks are deliberately redesigned to ensure that the gains of the automated horizon are broadly shared.

1. According to Paragraph 1, what has historically been the economists' standard argument against fears of technological unemployment?
   • A. That technology destroys jobs at the task level while creating sufficient new occupations to absorb displaced workers. ✓
   • B. That automation primarily affects unskilled workers who can be easily retrained for new roles.
   • C. That governments have historically intervened effectively to fund retraining and prevent mass displacement.
   • D. That labour unions have successfully negotiated transition agreements during every major industrial shift.

   Paragraph 1 explicitly states economists invoke the pattern that 'technology destroys jobs at the task level while expanding the economic pie sufficiently to produce new occupations.' The passage also notes the 'lump of labour fallacy' framing used to dismiss displacement fears. Option A is a direct paraphrase. Options B, C, and D introduce claims that are not made in Paragraph 1.

2. The passage defines 'labour market polarisation' as which specific phenomenon?
   • A. The global migration of manufacturing jobs from developed to developing economies.
   • B. The widening wage gap between unionised and non-unionised workers in advanced economies.
   • C. The hollowing out of middle-skill, middle-wage occupations susceptible to algorithmic replacement. ✓
   • D. The concentration of employment in technology-sector hubs such as Silicon Valley.

   Paragraph 3 explicitly defines labour market polarisation as 'the hollowing out of middle-skill, middle-wage occupations — bookkeepers, data entry clerks, paralegals, radiographers — that are precisely codifiable and thus susceptible to algorithmic replacement.' Option C is a direct paraphrase. Options A, B, and D describe different phenomena not discussed in the passage.

3. According to Paragraph 4, why does automation structurally undermine existing OECD social contribution systems?
   • A. Because automated production generates profits in offshore tax havens beyond the reach of national authorities.
   • B. Because automation displaces workers into the informal economy where social contributions are not collected.
   • C. Because multinational corporations successfully lobby against automation taxes in most legislatures.
   • D. Because payroll-tax-based fiscal systems presuppose wage employment, which automation systematically erodes. ✓

   Paragraph 4 states: 'the tax system in most OECD nations was designed for an economy of stable, full-time employment: corporate tax revenues, payroll taxes, and social contributions all presuppose wage employment as the primary vehicle of value capture.' As automation substitutes capital for labour, 'the tax base erodes precisely when the social demands on the state intensify.' Option D accurately captures this structural mismatch. Options A, B, and C introduce claims not made in the passage.

4. What does the author imply by stating that 'aggregate productivity figures systematically obscure' the distributional consequences of labour market polarisation?
   • A. That official statistics are deliberately manipulated to conceal the social costs of automation.
   • B. That GDP growth can coexist with, and thereby mask, median wage stagnation and precarious employment. ✓
   • C. That economists rely too heavily on quantitative data and should incorporate qualitative research methods.
   • D. That labour market statistics are inherently more reliable than GDP figures for assessing economic health.

   Paragraph 3 states 'the median worker may face stagnant wages and precarious employment even as GDP and corporate profits reach historic peaks.' The author implies that aggregate measures like GDP capture gains at the top without revealing the stagnation experienced by the median worker — making B the correct inference. Option A alleges deliberate manipulation, which the passage does not assert; C and D introduce methodological claims outside the passage's scope.

5. What can be inferred about the author's view of the historical precedent argument used to dismiss automation fears?
   • A. The author accepts it fully, arguing AI will ultimately generate more occupations than it destroys.
   • B. The author considers it entirely irrelevant given that contemporary automation is categorically unprecedented.
   • C. The author finds it partially valid but insufficient, given that AI now penetrates cognitive rather than only physical labour. ✓
   • D. The author rejects it on methodological grounds, arguing historical analogies cannot predict future labour markets.

   The author presents the historical analogy fairly in Paragraph 1, then opens Paragraph 2 with 'The contemporary wave of automation challenges this comforting historical analogy in several important respects' — specifically because AI now displaces cognitive labour, not only physical tasks. This is not a total rejection (ruling out B and D) nor an acceptance (ruling out A), but a qualified challenge: valid in its historical context, insufficient for the present. Option C is the accurate characterisation.