Trade Policy Under Perfect Competition

ECON 171 · Spring 2026 · Week 5

Sasha Petrov

Today’s Agenda

  1. Tariffs: specific and ad valorem
  2. Welfare effects of a tariff in a small country
  3. Terms-of-trade effect in a large country

What We Aim to Learn

Motivation: why study tariffs now?

  • After four decades of falling tariffs under the WTO, US policy reversed sharply in 2018.
  • 2018–19 US-China trade war: 25% duties on ~$350 B of bilateral trade.
  • 2018 Section 232: 25% on steel, 10% on aluminum.
  • 2025 “reciprocal” tariffs: Yale Budget Lab estimates the US average effective tariff rate is at its highest level since the 1930s.
  • This week’s question: who actually bears the cost of a tariff, and by how much?

Weighted-mean applied tariff rate, 1988–2022. Source: World Bank WDI via Our World in Data. 2023+ US surge not yet in the series.

What are tariffs?

Specific vs. ad valorem tariffs

  • Specific: rate \(s\) in $/unit (e.g., $0.063/L on still wine). Duty \(= s \cdot M\).
  • Ad valorem: rate \(t\) on declared value (e.g., 25% on light trucks). Duty \(= t \cdot p^{*} \cdot M\).
  • Compound: both at once (common for dairy and other agriculture).
  • Each line is filed in ACE on CBP Form 7501 with HTS code, value, and rate.

Excerpt of a real CBP Form 7501 entry summary, with boxes 32 (Entered Value), 33 (HTSUS Rate), and 34 (Duty) highlighted.

25% \(\times\) $700 \(=\) $175    ⟸ ad valorem \(t \cdot p^{*} \cdot M\)

Real 2015 entry at the Port of Tacoma. HTSUS heading 8704.31 (small light trucks) carries the 25% “chicken tax” — LBJ’s 1964 retaliation against EU duties on US frozen chicken. Source: drift-and-drive.com. For a specific duty, Box 33A would read e.g. $0.063/L and would multiply Box 31 (quantity) instead.

What’s on a complete Form 7501?

Fully completed CBP Form 7501 Entry Summary, page 1, with every block filled in for a sample import.

  • Header (1–24): filer/entry codes, port codes, country of origin, manufacturer ID, dates.
  • Parties (25–26): ultimate consignee (where goods go) and importer of record (legally liable for duty).
  • Line items (27–34): HTS code, value, rate, duty — one row per commodity.
  • Totals (35–40): $3,870 entered value × 25% = $967.50 duty, plus $31.67 MPF = $999.17 owed.
  • Filer (41–43): a licensed broker signs and certifies on the importer’s behalf in ACE.

Training sample (filer Awesome Brokers LLC, consignee Awesome Corp, Austin TX): transmission shaft, HTSUS 8483.10.5000, UK origin, American Airlines airfreight. Source: importsacademy.com.

A real-world Form 7501

A real, partly-redacted CBP Form 7501 filed by a US small-business importer of playing cards from China in September 2021.

  • Real entry (Sept 2021): 3,333 packs of playing cards from China (HTSUS 9504.40.0000), ocean freight on CMA CGM Loire.
  • Redacted in public: importer (25–26), entered value (32A), duty (34), totals — privacy realities of 7501s shared on the open web.
  • Free MFN duty for HTS 9504.40 (0%), but MPF (0.3464%) + HMF (0.125%) still apply — user fees on every formal entry.
  • Filer code DP6 = customs broker filing for the importer; box 26 names the legally liable importer of record.

Real entry by Benebell Wen, a US author who self-publishes tarot decks; she released the redacted form publicly while documenting her import process. Source: benebellwen.com.

How does a tariff affect a small country?

Price and quantity effects of a tariff

\(a\) = PS gain · \(b, d\) = DWL · \(c\) = tariff revenue · \(a+b+c+d\) = CS loss

  • Tariff \(t\): Foreign exporters now receive only (domestic price \(-\,t\)) per unit sold to Home
  • At the old world price \(p^{*}\), Foreign would net \(p^{*}\!-\!t < p^{*}\). Too low — Foreign refuses; \(X_F\) collapses to zero
  • A shortage opens at \(p^{*}\): the full free-trade import volume goes unfilled
  • Price bids up to clear the shortage. Domestic producers expand along \(S_H\); consumers contract along \(D_H\)
  • At \(p^{*}+t\) Foreign re-enters (nets \(p^{*}\) again). Imports settle at the smaller \(Q_D^{t}-Q_S^{t}\)

Welfare effects of a tariff

\(a\) = PS gain · \(b, d\) = DWL · \(c\) = tariff revenue · \(a+b+c+d\) = CS loss

  • Tariff arrives. \(X_F\) shifts to \(p^*+t\); production rises to \(Q_S^t\), consumption falls to \(Q_D^t\).
  • CS loss \(= a + b + c + d\) (full blue trapezoid): consumers pay \(t\) extra on every unit they still buy.
  • PS gain \(a\) (maroon) + tariff revenue \(c = t \cdot M^t\) (gold): transfers, rebated to consumers.
  • DWL \(= b + d\) (red triangles): production distortion + consumption distortion.

Net welfare: \(\Delta W = -(b+d) < 0\). SOE tariff is purely deadweight — no TOT effect to offset.

Welfare on the PPF

Tip — scroll on the chart to zoom, drag to pan, double-click to reset.

  • Tariff arrives. Budget kinks at \(A\) (import side \(-(p^*+t)\), export side \(-p^*\)). Utility drops to \(U_{1'}\).
  • Producers respond. Kink follows \(A \to B\) along the PPF. Utility rises to \(U_2\) (PS gain \(a\)).
  • Rebate. Kink shifts up to \(B' = (Q_C^B,\, Q_W^B + tM)\). Utility rises to \(U_3\) (revenue \(c\)).
  • DWL bracket at \(C^{FT}\): \(U_1 - U_3 = b + d\) (production + consumption distortion remain).

Same DWL as the PE picture — visualised as a utility gap rather than triangle areas.

What changes for a large country?

Tariff effect on the terms of trade

Tip — scroll on the chart to zoom, drag to pan, double-click to reset.

  • Tariff arrives. Wedge \(t\) between \(p^{*}+t\) (Home) and \(p^{*}\) (Foreign). Country curves don’t shift — agents move along them at their effective prices.
  • World \(RS\) on the \(p^{*}\) axis appears to shift right; world \(RD\) appears to shift left (Home’s contributions evaluated at \(p^{*}+t\)).
  • New equilibrium: \(p^{*}\) falls. Home buys imports cheaper — the TOT effect.
  • Beggar-thy-neighbor: Home’s gain is Foreign’s loss. \(\lambda \to \infty\) recovers SOE.

Foreign-side arbitrage

Foreign consumers buy in Foreign’s own market — they don’t observe Home’s border. So where does the \(p^{*} - t\) they “see” come from?

  • Two channels for Foreign producers. Sell domestically at \(p^F\), or export — gross border price \(p^{*}\), net \(p^{*} - t\) after remitting the tariff to Home’s customs.
  • Arbitrage. In equilibrium Foreign does both, so producers must be indifferent between channels: \(p^F = p^{*} - t\).
  • Self-correcting. If \(p^F > p^{*} - t\): no exports → Home unfilled → \(p^{*}\) rises. If \(p^F < p^{*} - t\): all output exported → Foreign shortage → \(p^F\) rises. Either way back to equality.
  • Foreign consumers buy from these producers at the domestic price \(p^F\). By the arbitrage, \(p^F = p^{*} - t\). Their response to \(p^{*} - t\) is mediated through producers’ export option — not direct sensitivity to Home’s border.

On the world \(RS/RD\) chart under foreigner-pays, all Foreign behaviour (production + consumption) is plotted at \(p^{*} - t\) — through internal arbitrage in Foreign’s market, not Home-side sensitivity.

Who absorbs the tariff?

Tip — scroll on the chart to zoom, drag to pan, double-click to reset.

  • Foreigner-pays. Foreign nets \(p - t\), so \(X_F\) contracts at every \(p\). Combined supply \(S_H + X_F\) shifts left; \(D_H\) stays put.
  • New equilibrium at higher Home price \(p^{H}_{t}\). \(S_H\) expands; Foreign exports shrink.
  • Wedge \(t\) between \(p^{H}_{t}\) (Home pays) and \(p^{F}_{net} = p^{H}_{t} - t\) (Foreign nets).
  • Incidence: top band = Home’s price rise (loss); bottom band = Foreign’s price drop (= Home’s TOT gain).

The optimal tariff \(t^{*}\)

  • \(\Delta W(t) = \Delta PS + G - \Delta CS\) from Home’s perspective
  • For small \(t\): TOT gain (linear in \(t\)) beats DWL (quadratic). \(\Delta W\) rises
  • For large \(t\): DWL dominates and imports collapse. \(\Delta W\) falls
  • Interior maximum at \(t^{*}\) — the optimal tariff for a large country
  • \(\lambda \to \infty\): curve flattens to the SOE shape — strictly negative for any \(t > 0\)

But: Foreign loses more than Home gains. Beggar-thy-neighbor — and Foreign retaliates ⇒ trade war.

Wrap-up

Key takeaways from today

  • A tariff drives a wedge between domestic and border prices: \(p_H = p_F + t\).
  • Small open economy: Foreign supply is perfectly elastic — Home absorbs the full wedge. Welfare change \(\Delta W = -(b + d) < 0\), purely deadweight.
  • Large open economy: Home’s tariff drags down the world price, transferring rents from Foreign (the TOT effect). The optimal tariff is positive — but it’s beggar-thy-neighbor.
  • Strategic logic: if Foreign retaliates, both lose (Pareto-dominated by free trade). WTO/GATT exists to discipline unilateral tariff use.

Tariffs aren’t always welfare-reducing — but the cases where they help one country are the cases where they hurt another more.

Skills you’ve added to your toolkit

  • Decompose welfare into four regions: PS gain \(a\), production DWL \(b\), tariff revenue \(c\), consumption DWL \(d\). CS loss \(= a + b + c + d\).
  • Translate between the partial-eq picture (Home market with \(S_H\) and foreign export supply \(X_F\)) and the GE picture (PPF + relative price line).
  • Pick the right assumption: SOE when Home’s market share is small; LOE when Home is big enough to move the world price.
  • Movement along ≠ shift: the tariff changes the effective price each country faces — agents move along their unchanged curves. Only the world RS on the border-price axis appears to shift.
  • Incidence is about elasticities: who legally writes the check (importer vs. exporter) does not change the equilibrium — the wedge is \(t\) either way.