Knowing the model is the easy part.

A depreciation raises the domestic-currency price of anything bought from abroad. That is the first missing step: imported goods and, crucially, imported inputs become more expensive in pounds.

The second step splits in two, and strong answers separate them. Imported finished goods enter the consumer price index directly, so their prices feed through to measured inflation almost immediately. Imported inputs raise firms’ production costs, shifting short-run aggregate supply left, which raises the price level of domestically produced goods too.

So the chain is: depreciation → higher sterling price of imports → (a) direct effect on the price of imported consumption, and (b) higher input costs → leftward shift in short-run aggregate supply → higher price level.

Having built it, the condition becomes visible: the size depends on the economy’s import intensity, and on whether firms can absorb costs in margins rather than passing them on — which itself depends on how competitive their markets are.

A demand-side explanation — higher incomes — shifts demand right. Price rises and the quantity transacted or built rises with it.

A supply-side explanation — tighter planning — shifts supply left. Price rises and quantity falls.

So the two hypotheses are distinguishable, and the distinguishing variable is quantity, not price. Rising prices alongside rising construction points to demand; rising prices alongside falling construction points to supply.

In practice both are usually happening at once, which is why the honest answer is about relative magnitude: which shift is larger, and what would the quantity data have to look like for each story to be the dominant one?

The reasoning is sound given perfect competition: with normal profit only, a firm that absorbed the tax would make a loss and exit, so price must rise by the full amount in the long run.

With four firms the assumptions fail in a specific way that matters. These firms are likely to earn supernormal profit, so absorbing part of a tax is survivable rather than fatal. They also face a downward-sloping demand curve for their own output, so raising price loses customers to rivals.

That changes the prediction: pass-through is likely to be partial, and how partial depends on the elasticity of demand facing each firm and on how rivals respond — a strategic question the competitive model does not contain.

The conclusion therefore does not survive in its strong form. "Fully passed on" becomes "partly passed on, with the share depending on elasticity and on the competitive response".

Note the direction of the error: the student overstated pass-through. Saying which way an assumption bends the answer is worth more than saying it is unrealistic.

The first sentence needs a mechanism. A subsidy paid to producers lowers their marginal cost, shifting supply right; the new equilibrium price is lower, but by less than the subsidy unless demand is perfectly elastic.

The second needs a condition. How much purchasing rises depends on the price elasticity of demand for electric cars, which is likely to be higher where a close substitute exists — and the closeness of petrol cars as a substitute depends on charging infrastructure, which is itself a policy variable.

The "however" needs to do something. Fiscal cost is only an argument if it is weighed: against the value of the externality being corrected, and against what else the money could buy. As written it is a fact, not a counterargument.

The conclusion needs to choose. For example: the subsidy is more likely to be justified where the marginal external damage from petrol vehicles is high and where charging infrastructure makes demand responsive — and least justified where it mainly transfers money to buyers who would have switched anyway.

That last clause is the strongest thing in the paragraph, because it identifies the case where the policy fails on its own terms.

A demand-side story cannot do it. A leftward shift in aggregate demand lowers output and lowers the price level, so unemployment rises and inflation falls. The facts do not match.

A leftward shift in short-run aggregate supply does exactly what is described: the price level rises and equilibrium output falls, so employment falls with it. Oil is an input across most of the economy, so a price rise is a genuine economy-wide cost shock rather than a change in one market.

So the aggregate demand and supply diagram is the right frame, and the analytical content is the direction of the shift and the reason for it.

The labour market diagram is not wrong, but it is downstream — it shows the consequence for employment rather than the cause, and using it alone leaves the mechanism unexplained. The production possibility frontier is the weakest choice: it describes capacity rather than short-run equilibrium, and nothing about a price change moves the frontier itself.

The strongest answers use one diagram properly rather than three loosely.

With substantial spare capacity, the short-run aggregate supply curve is relatively flat. Extra demand is met by employing idle resources, so output rises substantially and the price level rises only a little. The multiplier operates close to its full size, because the newly employed spend a large share of new income.

Near full employment, aggregate supply is steep. The same demand increase runs into resource constraints, so most of it appears as a higher price level rather than higher output. The real multiplier is much smaller.

A second channel differs too. Near capacity, government borrowing is more likely to raise interest rates and displace private investment — crowding out — whereas with idle resources and slack credit demand that effect is weaker.

The judgement therefore has to be conditional: the policy is considerably more effective at raising real output when spare capacity exists, and closer to purely inflationary when it does not.

A strong answer adds a qualification the standard version misses: infrastructure spending also raises productive capacity over time, shifting long-run aggregate supply right. So even the near-capacity case is not simply inflationary — it depends on the horizon you are judging over, and on whether the projects are actually productive.

Start with the structure of the comparison. The gain is concentrated in one industry; the cost is spread across every industry that uses steel as an input. So the first question is relative size: how large is steel production compared with steel-using manufacturing?

That comparison is the crux, and it is an empirical question rather than something theory settles — you would want employment figures for the two groups before committing. Note the shape of it, though: steel production is one industry, while steel-using manufacturing spans construction, vehicles, machinery, packaging and appliances. A single industry being outweighed by the many that buy from it is the pattern worth testing for.

But size alone is not decisive. The cost per steel-using firm is diluted across many firms and may be small relative to their other costs, while the benefit per steel producer is concentrated and large. That is why the politics runs the other way from the economics — concentrated benefits mobilise, diffuse costs do not.

Two further considerations sharpen the judgement. Retaliation by trading partners would impose additional costs on exporters, a third group not yet counted. And if the tariff is temporary and the domestic industry uses the protection to become competitive, the long-run picture differs from the short-run one — though the empirical record on that is not encouraging.

A defensible conclusion: the tariff is likely to reduce national output and employment, because the affected downstream sector is larger, with the strongest case against being the possibility of retaliation and the weakest case for resting on an infant-industry argument that requires the protection to actually expire.

Reverse causation. Countries with strong labour markets and low unemployment may be more willing to set a high minimum wage, precisely because they can absorb it. The economic strength causes the policy rather than the other way round.

A common cause. Institutional quality, education systems, or the strength of collective bargaining could plausibly raise both the minimum wage and employment independently. The correlation is then real and the causal link between the two is absent.

Selection and comparability. Countries differ in how unemployment is measured, in the size of the informal sector, and in what fraction of workers the minimum actually binds for. A high minimum that covers few workers is not the same policy as a lower one that covers many.

None of these shows the student’s conclusion is false. They show the evidence presented cannot establish it — which is a different and more careful claim.

What evidence would help? Comparisons that exploit variation within a country over time, or between neighbouring regions with different minimums but similar conditions, get closer to isolating the effect.

Separate efficiency from distribution before trying to judge. On efficiency, the charge internalises an externality: each driver imposes delay and pollution on others and does not pay for it, so the untaxed quantity of driving exceeds the efficient one. Falling traffic, better air and faster buses are all consistent with the policy working as intended.

The retail finding needs care. Lower takings inside the zone is not the same as lower total retail activity — some spending may have moved elsewhere, and some shoppers arriving by bus may have replaced those arriving by car. Whether this is a real cost depends on evidence the statement does not contain.

The hardship finding is different in kind. It is a genuine distributional cost, and it is not answered by the efficiency argument. A policy can raise total welfare while making an identifiable group worse off, and that is exactly what is happening here.

The design question then becomes the interesting one. The charge and the use of its revenue are separable decisions: revenue could fund public transport improvements, or exemptions could be targeted at those with no alternative. Judging the policy without asking what happens to the revenue is judging half of it.

A defensible judgement: the charge is likely to improve efficiency, with the strength of the case depending on how good the public transport alternative is; the distributional objection is real and is best addressed through the revenue rather than by abandoning the charge, since exempting the drivers who most need to travel would weaken exactly the behaviour it is trying to change.

Start with the structure. Supply is perfectly inelastic — the venue holds what it holds. With fixed supply, any increase in demand raises the equilibrium price and cannot raise quantity. That single observation frames everything else.

Now ask what a fixed below-market price does. If tickets are priced below the clearing level, quantity demanded exceeds supply and the tickets must be rationed some other way: queues, bots, luck in a ballot. This is the price-ceiling mechanism, arrived at from a different direction.

A resale market then appears, and the difference between the face price and the resale price accrues to resellers rather than to the artist or the venue. Dynamic pricing is, in effect, the seller capturing that surplus instead.

So the efficiency argument for dynamic pricing is real: it allocates tickets to those willing to pay most and removes the resale margin. But willingness to pay is not the same as intensity of desire — it is constrained by income, so the allocation tracks ability to pay, which is the substance of the fairness objection rather than a confusion.

The judgement should identify what is actually in dispute: not whether the price is "correct", but whether allocating a positional good by willingness to pay is acceptable when the alternative allocates it by luck or by time spent queueing. A defensible position is that dynamic pricing is more efficient and less equitable than a ballot, and that which matters more depends on what the good is for.