VJC 2024 H1EC Prelim Answers and Mark Scheme (20 Sep) Final

1 Answers and Marks Scheme to 2024 H1EC Prelim Exam Question 1 (a) (i) Describe the trend in the average selling price of graphics processing units (GPUs) from Q1 2019 to Q2 2022. [General trend] The a verage selling price of GPUs experienced an overall increase from Q1 2019 to Q1 2022 [1]. [Refinement] However, there was a reversal of trend after Q3 2021, where the average selling price of GPUs fell from Q3 2021 to Q2 2022. [1] [2] (ii) Using Extract 1 and a diagram, explain the observed trend in the average selling price of GPUs. The rise in average selling price of GPUs is due to a fall in supply and rise in demand for GPUs. There was a rise in demand for GPUs due to the change in tastes and preferences of consumers. During COVID -19, consumers staying home bought more GPUs for gaming and crypto -mining, leading to a rise in willingness to purchase GPUs at all prices and hence increase in the demand for GPUs (rightward shift of the demand curve from D0 to D1). [1] There was a fall in supply of GPUs due to supply chain disruptions and shipping delays, causing a reduction in the number of units of GPUs producers were able to offer for sale at every price level. This resulted in a fall in the supply of GPUs (leftward shift of the supply curve from S0 to S1). [1] The rise in demand and fall in supply results in a shortage of (Qd-Qs) units at the initial equilibrium price of P0, resulting in an upward pressure on prices. [1] As price increases, quantity demanded will fall and quantity supplied will rise until equilibrium is attained at E1 where quantity demanded = quantity supplied at Q1, where equilibrium price has increased from P0 to P1. [1] [5]

2 Diagram [1] (b) With reference to Extract 1, explain why a graphic processing unit has the characteristic of excludability. According to Extract 1, a GPU commands a price. [1] This means one must pay to obtain the good. Non -payers are excluded from consuming this good. [1] [2] (c) Explain the likely impact of the introduction of software such as “cuLitho” on the total revenue of GPU producers, and comment on whether this is likely to hold in the future. Explain impact of “cuLitho” on TR of GPU producers [4] With the introduction of “cuLitho”, there would be a fall in the marginal cost of production of GPUs due to a speed up in computational lithography, leading to a rise in supply as GPU producers are now willing and able to produce more GPUs at every price level. [1] With a rise in supply, the resultant surplus would result in a fall in equilibrium price and a rise in equilibrium output. [1] The demand for GPUs is likely to be price inelastic due to a lack of close substitutes. As such, a fall in the price of GPUs would lead to less than proportionate rise in quantity demanded. [1] [6]

3 The fall in revenue from a fall in price thus outweighs the rise in revenue from a rise in quantity demanded, hence total revenue of GPU producers would have fallen. [1] Comment on whether this is likely to hold in the future [2] With the development of GPU substitutes such as SLIDE over time, the demand for GPUs will become less price inelastic or become price elastic instead. As such, we could see a smaller fall in total revenue or the resultant change in revenue might even be reversed and we could see a rise in revenue instead. [2] Accept any other factor that questions the ceteris paribus assumption e.g. entry of new firms over time → increase in SS. (d) (i) Explain how a rational firm uses the marginalist principle to decide how many units of semiconductors to produce. [Explain Marginalist Principle] A rational firm aims to maximise net total benefit (i.e. total profits) by making use of the marginalist principle, weighing the marginal benefit of producing semiconductors against its marginal cost. [1] [Explain MB and MC in context] The marginal benefit for the firm would be the revenue received from the sale of an additional unit of semiconductors, and the marginal cost would be the cost of the factors of production used in the production of an additional unit of semiconductors. [1] [2] (ii) With reference to Extract 4 and using a diagram, explain why the price mechanism could lead to an over-production of semiconductors. [Explain source of MF in context] In the production of semiconductors, negative externalities are generated. The production process involves the release of industrial wastewater into public waterways. As such, third parties who are not involved in the production of semiconductors could incur medical costs due to the water pollution. These negative externalities result in a divergence between the marginal private cost (MPC) and marginal social cost (MSC). [1] [Private vs social optimum] Profit -maximising firms only consider their marginal private benefit (MPB) and marginal private cost (MPC) in the production of semiconductors. They disregard MEC and supply according to MPC. The free-market equilibrium output Q0 is where DD (MPB) = SS (MPC). [1] The socially optimal output Q* is lower, where marginal social benefit (MSB) = MSC, which is where society’s welfare is maximised. There is thus an overproduction of (Q0-Q*) units of semiconductors. [1] [5]

4 [Welfare loss to society] From output Q* to Q0, the additional cost to society (area Q*bcQ0) is greater than the additional benefit to society (Q*bcQ0), hence resulting in a welfare loss of abc, meaning that there is allocative inefficiency. [1] Diagram [1] (iii) Discuss the appropriateness of the government policy of incentivising innovation in the semiconductor industry to improve societal welfare. Introduction From d(ii), there is a loss in societal welfare and hence allocative inefficiency in the semiconductor industry due to the negative externalities arising from semiconductor production. R1: Explain how government policy to incentivise innovation in the semiconductor industry can improve societal welfare [P] The government may decide to implement a policy to incentivise innovation in the semiconductor industry, which would lower the negative externalities generated and hence improve social welfare. [EE] The government may choose to provide research and development (R&D) grants to semiconductor firms to encourage their R&D efforts in the area of clean production. In doing so, this would lower the cost of conducting R&D for firms, which would incentivi se them to increase R&D efforts in the area of clean production, as the MC of doing R&D now becomes lower than the MB of doing R&D (i.e. the expected rate of return on the R&D efforts). [8]

5 [EE] If firms’ R&D efforts are successful, cleaner methods are invented and these methods are cheaper than the old methods, firms will be incentivised by the latter to switch to newly developed cleaner methods of production. With a reduction in MPC to MPC’, supply rises to SS’. This would red uce the MEC generated in the semiconductor industry (from MEC to MEC’), hence the MSC falls and shifts downwards from MSC to MSC’. As a result, the socially optimal quantity increases from Q* to Q*’ (where MSB = MSC’), leading to a smaller extent of overproduction of semiconductors and a smaller welfare loss of area xyz, [L] thus improving societal welfare. Figure 1: Market for semiconductors R2: Explain limitations of the above policy, in terms of appropriateness [P] However, there are limitations to such a policy that might affect its appropriateness in improving societal welfare. [EE] Firstly, the outcomes of innovation efforts are uncertain. It is possible that substantial resources are pumped into research on the development of cleaner methods of production in the semiconductor industry, but with none of the research coming to fr uition. Even if the researc