Growth Theory and Output Gaps

Every question on this spec using the words 'economic growth' is really asking about one of two different things, and the single biggest source of avoidable marks lost on this topic is not noticing which one. Actual growth is a rise in real GDP — however it happens, over whatever period. Potential growth is a rise in the economy's underlying productive capacity — how much it could produce if every resource were fully and efficiently employed, whether or not it's actually producing that much right now. These aren't two competing definitions of the same idea; they're answers to two different questions ('how much did the economy produce?' versus 'how much could it produce?'), and once that distinction is explicit, almost every cause, benefit and cost the spec lists sorts cleanly into one column or the other.

Actual growth happens whenever real GDP rises — which, on an AD/AS diagram, means a rightward shift of AD, a rightward shift of SRAS, or the economy simply moving closer to output it was already capable of producing but wasn't (closing an existing negative output gap). The spec names two specific routes into this (2.3.5.1b–c): growth from the AD components generally (a rise in C, I, G or net trade pulling real GDP up) and export-led growth specifically (a rise in X pulling real GDP up via the same AD = C+I+G+(X−M) identity — called out on its own because a stronger export sector is also, separately, a competitiveness and current-account story, not only a growth one). None of these routes change what the economy is *capable* of producing in the long run — they only change how much of that existing capability gets used right now.

Potential growth is different in kind, not just in size: it's an outward shift of the economy's LRAS curve (equivalently, its PPF) — the ceiling itself moving, not output moving toward an unchanged ceiling. The spec names four sources under item (2.3.5.1d): domestic investment and FDI (more or better capital per worker), innovation (new technology raising what a given quantity of inputs can produce), labour force growth including net migration (more workers means a bigger LRAS at every wage rate), and the degree of competition (weak competitive pressure lets firms sit above the lowest cost curve they could technically achieve, so stronger competition pushes actual output-per-input up toward the technically possible maximum) — plus a fifth source, productivity, under its own separate item (2.3.5.1e), covered in its own dedicated paragraph below. (This is spec point 2.3.5.1d-e's own named list of causes of potential growth — a different spec point, asking a different question, from 2.3.3.3's six named LRAS shifters, covered in full in the dedicated Aggregate Supply lesson: technology, productivity, education and skills, government regulation and tax, demography and net migration, and competition policy. The two lists don't map one-to-one — technology, education/skills and regulation/tax don't get their own named entry here, and investment/FDI and innovation don't get their own named entry there — because 'what causes potential growth' and 'what shifts LRAS' are the same underlying economics described from two different spec points, not two competing attempts at one list that happen to disagree.)

Labour force growth and capital investment both face a constraint that productivity growth doesn't: an economy can only add workers or machines so fast, and — the same diminishing-returns logic a single firm faces when it keeps adding workers to a fixed factory floor — adding more of one factor without limit, holding technology fixed, eventually adds less extra output per extra unit than the one before it. Productivity growth means more output from the *same* quantity of labour and capital; it isn't subject to that ceiling, because it's precisely what shifts the ceiling that labour-force and capital growth eventually run into. That's why the spec singles productivity out for its own dedicated clause (2.3.5.1e) rather than folding it in as a sixth item next to investment, innovation, labour force growth and competition — it's less a peer of those four causes than the shared channel most of them ultimately work through.

Benefits of growth (2.3.5.2a) follow from a rise in real GDP in general — higher living standards (more output means more goods and services available, on average, per person), lower unemployment (firms producing more typically need more workers to do it, which also reduces government spending on unemployment and welfare benefits — a distinct public-finances gain alongside, not instead of, higher tax revenue), higher firm profits (more output sold, more revenue), higher investment (profitable, growing firms retain and reinvest more, and rising confidence about future demand justifies new capital spending), higher tax revenue (a bigger tax base: more income tax from more employment and higher wages, more corporation tax from higher profits, more VAT from higher spending), and improved public services (assuming that extra tax revenue is actually spent that way, rather than banked or used to cut tax rates instead). Every one of these six is a real, examinable consequence of a rise in real GDP — but whether it's a *sustainable* consequence worth building policy around, or a temporary side-effect of an unsustainable AD boom, is exactly the actual/potential question the mechanism block below makes explicit. That same tax-revenue mechanism cuts two ways evaluatively: as well as funding public services, higher tax revenue from growth can narrow the government's fiscal/budget deficit, and — where the government chooses to spend it redistributively rather than broadly — can partly offset the inequality cost described below, rather than the benefit and cost sitting in unconnected paragraphs.

Costs of growth (2.3.5.3a) split the same way. Opportunity cost is the clearest potential-growth cost: the investment that raises LRAS has to come from somewhere, and in an economy already near full employment, more capital goods now means fewer consumer goods now — a straight PPF trade-off, not a free lunch. Environmental costs (resource depletion, pollution, emissions) attach to growth generally, however it's achieved, unless the extra output is specifically decoupled from resource use — and a real costs-of-growth mark scheme develops this as two separate mechanisms, not one vague 'environment' bullet: depleting non-renewable resources to produce today's output leaves less available to produce with in future, an intergenerational version of the same opportunity-cost logic just derived for potential growth ('The growth today could deplete natural resources reducing possible production in future'); and pollution is a cost that feeds back into the economy itself, not merely an externality sitting outside it — it can worsen health, living standards, wellbeing and worker productivity, each a real channel back into future output ('Pollution may affect health/living standards/wellbeing/productivity' — both January 2020 MS, Q13 indicative content). The mark scheme's own evaluative counterpoints to this cost turn on the same measurement-difficulty point the output gap uses: how much of a given rise in pollution is actually caused by growth (versus other factors) and how to monetise the economic damage it does are both estimates, not directly observed figures — while growth's own higher tax revenue can, in turn, fund the resources used to mitigate the pollution it causes. A widening trade deficit is a common actual-growth cost: as real incomes rise, households typically buy more imports too, so consumption-led actual growth tends to pull in imports faster than export-led or productivity-led growth does — and that cost cuts both ways, not just via imports: the same positive-output-gap inflation the worked chain below derives (SRAS steepening as scarce resources are bid up, pushing the price level higher) also makes exports less price-competitive abroad, so rapid actual growth can widen a current-account deficit from the export side as well as the import side (the same mechanism, developed more fully in growth-and-inflation.ts, where domestic inflation running ahead of trading partners' inflation is shown damaging competitiveness and the current account directly). Inequality is a distributional cost that can attach to either type, but is a particular risk of investment- or innovation-led potential growth specifically, since the returns to new capital and new technology often accrue disproportionately to whoever already owns the capital or holds the skills the new technology complements. And inflation — the cost most tightly derived from the actual/potential distinction itself — is specifically a risk of *actual* growth running ahead of *potential* growth, which is exactly what an output gap measures, and exactly what the worked chain below derives rather than asserts.

An output gap (2.3.5.4) is the difference between actual real GDP and potential (trend) real GDP, usually expressed as a percentage of potential output: a positive gap means actual output is temporarily above what the economy can sustainably produce; a negative gap means it's below. The genuinely hard part, spec-named explicitly as its own sub-point ('difficulties of measurement', 2.3.5.4d), is that potential output is never directly observed — nobody can read the economy's true capacity off a dial the way a thermometer reads temperature. It has to be *estimated*, usually from the historical trend in output over many years, and different estimation methods (and simply more data arriving later) produce different trend estimates for the very same period — which is exactly why real published output-gap figures get revised, sometimes substantially, well after the period they describe.

Everything derived above is provable on the time-series diagram already shown — actual real GDP plotted against a trend line over time. But the same output gap has a second, equally examinable picture: the standard AD/SRAS/LRAS diagram this course already uses for every other short-run-versus-long-run question. The trap-taxonomy entry above names this as one of the single worst-performed diagram questions found across the whole archive, so it earns its own dedicated diagram block below rather than staying as a single sentence inside the time-series one.

The translation is exact, not approximate: the short-run equilibrium — where AD intersects SRAS — plays the role the 'actual' line played on the time-series version, and LRAS, drawn as its own vertical (classical) or part-vertical (Keynesian) curve at potential output, plays the role the 'trend' line played there. A short-run equilibrium sitting to the RIGHT of LRAS is a positive output gap — actual output beyond what the economy can sustainably supply, the same inflation-risk condition the worked chain above derives, now read off as a horizontal gap between two curves instead of two lines over time. A short-run equilibrium to the LEFT of LRAS is a negative output gap — the same spare-capacity condition, read the same way.

The reason this version is harder to draw correctly isn't the economics — it's that it asks for three curves and two separate labelled points (the short-run equilibrium AND full-employment output on LRAS) in one diagram, where the time-series version only ever asks for two lines. Missing LRAS entirely, drawing it in the wrong position relative to the short-run equilibrium, or drawing it correctly but labelling it SRAS are three separate, independently confirmed ways to lose the same marks — see the diagram's own commonError below for exactly what a real examiner report found.