The New Trade Model

ECES905205 Meeting 5

I Made Krisna Gupta

2022-09-01

Latest iteration

  • Gains from trade so far:

    • different technology.
    • different factor of endowment.

  • Today:

    • product variety.
    • productivity improvement from selection.

Internal economies of scale

  • We relax perfect market a bit: non-trivial fixed cost.

  • Consequently, average cost goes down with scale.

  • Meaning, a market can only host a limited number of firms.

  • In reality, fixed cost come in many firms: R&D, network, machines, etc.

Internal economies of scale

  • In reality, there are some industries where the global market only shared by small number of MNCs.

    • semiconductor, aeroplane, software, soft drink.

  • In this context, trade brings more gain:

    • the most efficient firms will dominate market and push cost down.

    • inefficient firms exit.

example: mobile phone

Source: StatCounter Global Stats - Device Vendor Market Share

Monopolistic competition

  • In monopolistic competition, firms have some ability to control price.

  • Condition which enable this:

    • Few number of players.

    • DIfferentiated products.

  • Product differentiation may come from various aspect:

    • Technology: smartphones, OS, recipe.

    • Network effect: social media, apps store, Gojek/Grab.

Monopoly: a brief review

  • Assume an industry face demand \(Q=a-bP\)
  • \(P\) is not exogen (why?), thus rearrange:
    • \(P=\frac{a}{b}-\frac{Q}{b}\)
  • Revenue in the industry \(R=Q\times P\), subs \(P\):
    • \(R=Q \times \left(\frac{a}{b}-\frac{Q}{b}\right)\)
    • \(MR=\frac{a}{b}-2\frac{Q}{b}=P-\frac{Q}{b}\)

Monopoly: a brief review

  • Total cost: \(C=F+cQ\)
  • Marginal cost: \(c\)
  • Average cost: \(\frac{F}{Q}+c\)
  • as \(Q \rightarrow \infty\), \(\frac{F}{Q} \rightarrow 0\)
  • The larger F, the larger Q is required.

The cost curve

AC2 has a higher fixed cost, thus requires more Q to reach “flat-ish” part of the cost curve.

Monopoly Q

  • monopolist akan memaksimalkan profit di MR=MC
  • pastinya akan ngecharge P>AC
  • profit \(=(P-AC)Q\)

Monopolistic competition

  • Characteristics of monopolistic competition industries:

    • Firms can differentiate.
    • Rival’s price is exogenous.
    • free entry.

  • We also introduce high fixed cost!

    • while entry is free, a firm must pay some fixed cost.

  • Can you mention some examples?

Karakteristik monopolistic

\[ Q=S\times\left[1/n-b(P-\bar{P})\right] \] - Q = total sales of individual firm - S = total sales of the industry - n = number of firms - b = elasticity of supply - P = firms’ price - \(\bar{P}\) = rival’s price index

Karakteristik monopolistic

\[ Q=S\times\left[1/n-b(P-\bar{P})\right] \]

  • A firm’s sales go up if total sales of the industry goes up and/or rivals’ price go up.
  • A firm’s sales goes down if the number of firms go up and its own price go up.
  • All firms are identic: \(P=\bar{P}\), all firms share the market equally \(S/n\)
  • \(AC=\frac{F}{Q}+c \Rightarrow AC=n\frac{F}{S}+c\)

Monopolistic characteristics

\(AC=n\frac{F}{S}+c\)

  • Average cost goes up as numbers of firm go up. This is because firms’ market share goes down. berkurang.
  • As the overall pie goes up, average cost go down: the fixed cost can be shared among more sales.
  • With a linear demand \(Q=a-bP\), firms produce at MR=MC
  • \(P-\frac{Q}{Sb}=c \rightarrow \frac{Q}{Sb}=P-c \rightarrow P-c=\frac{1}{nb}\)

Monopolistic Characteristics

\(P-c=\frac{1}{nb}\)

  • As the number of firms go up, \(P-c\) (markups) goes down.
    • Since c is exogenous, reduction of markups is driven by \(P \Downarrow\)
    • More firms, less market share, P must go up to avoid loss.
  • Firms will keep coming in as long as there’s profit.
  • In the long-run, number of firms settles at \(P=AC\)

Monopolistic

  • Any higher than \(n^*\), firms cost is too much.
  • Any lower than \(n^*\), mark-up will go up, incentives for new entrant to join.

Trade & monopolistic

\[ AC=n\frac{F}{S}+c \\ P=\frac{1}{nb}+c \]

  • Opening to trade increases the market size.
    • \(S \uparrow\) increase AC (less share per firm), but pricing is unaffected.
    • New equilibrium is lower price and more firms in the market.
  • More firms->more varieties.

Nummeric examples

  • Let an automotive industries as follows:

\[ Q=S\times[(1/n)-(1/30000)\times(P-\bar{P})] \\ \]

  • with this cost structures:

\[ AC=750000000/Q+5000 \]

  • Let home sells 900k vehicles a year while foreign produces 1,600k.

Numeric examples

Contoh numerik

Home, no trade Foreign, no trade Integrated market
Output 900,000 1,600,000 2,500,000
n 6 8 10
Output per firm 150,000 200,000 250,000
AC=P 10,000 8,750 8,000

Highlighted feature

  • Even when both countries have the same technology (identical cost structure), trade brings positive impact.
  • Monopolistic competition is benefited from intra-industry trade.
    • Export car but also import car.
    • Think of Toyota and Ford.
  • more car variety: from 6 & 8 brands to 10 brands for both countries.

Highlighted feature

  • Note that intra-industry trade is more beneficial for Home
    • price reduces by 2k, foreign only by 750.
  • This is because home country have a smaller market.
    • The smaller the market, the less fixed cost is shared.
    • This is why small countries like Singapore, Malaysia, Vietnam and Thailand pursue FTAs vigorously.

Firms response to trade

  • From 14 brands of car, only 10 survives.
  • These 10 brands get more market share, hence able to reduce cost.
  • Since all firms are identical in our analysis, we do not care who out and who’re staying.
  • However, in the real world, firms are not identical: 4 firms who exit must be the worst performers.

Firms response to trade

  • Market integration forced firms with worse cost structure to exit.
  • If the fixed cost to enter the industry is the same, cost variation will come from variable cost.
  • Then there’s a threshold \(c^*\) where all firms with \(c>c^*\) exit.

More gain from trade

  • This, in turn, adds to the gain from trade:

    • worst firms will exit, only firms with \(c<c^*\) stays.
    • meaning, price index further even more from lower average variable cost.
    • If 5 worst students drop out, average GPA will go up.

  • Firms who exit will be absorbed by the stayers: either naturally or through acquisition/joint-venture.

Export cost

  • We can associate a fixed cost with export cost:
    • establish presence, market research, finding partners.
    • Only the most productive firms join the global market.
  • Opposite argument is also true:
    • Learning by doing: As the firm keeps on exporting, they learn from the global market and become much more productive.

Multinationals and FDI

  • In general, FDI is considered more costly than ekspor.
  • Greenfield FDI: Brand new investment.
  • Brownfield FDI: purchasing someone else’s asset.
  • Horizontal FDI: Replicating domestic success.
  • Vertical FDI: global value chain.

Multinationals and FDI

  • Horizontal FDI is done under high trade cost.
    • the goal is market access. Usually will not export to other countries.
  • Vertical FDI: the goal is to supress cost.
    • Low-tech countries can export high-tech goods.
    • Requires low trade cost, and possible large intermediate imports.
    • leads to intrafirm trade.

Intrafirm Trade

  • Analyzing between-country trade is getting less relevant as the importance of intrafirm trade improves.
  • Half of total imports of US are intrafirm. However, around 10% of US trade with Indonesia is intrafirm.
  • The difference between exporters and non-exporters is mostly driven by firms that both import and export. This is true also in Indonesia.
  • Amiti and Konings (2007) show that a reduction of import tariff by half improves productivity of an importing firm by 12%, which is larger than the reduction of output tariff by destination countries.
  • Channels: foreign technology embeded in the inputs, variety of inputs, cheaper and higher quality inputs.
  • Others try to find more evidence for Indonesia (e,g., Pane and Patunru 2019, Gupta 2020), but data availability in firm level remains a huge challenge.