Commercial Imperialism? Political Influence and Trade During the Cold War

Replication and visualization of Berger, Easterly, Nunn & Satyanath (2013, American Economic Review)

Does political power shape international trade? This question sits at the intersection of international relations and economics, touching on debates about imperialism, globalization, and the nature of the liberal economic order.

The empirical challenge is formidable: political influence operates through backroom deals, diplomatic pressure, and covert operations—leaving few traces in official statistics. How can we identify the causal effect of political power on trade when influence is, by design, invisible?

The Natural Experiment

The authors exploit a remarkable source of variation: declassified CIA documents revealing American covert interventions during the Cold War (1947–1989). These operations—ranging from coups d’état to sustained covert support for friendly regimes—provide plausibly exogenous variation in US political influence over foreign governments.

Why is this variation useful for causal inference? Because interventions were motivated primarily by ideology (containing Communism), not commerce. The CIA overthrew Mossadegh in Iran (1953) to prevent Soviet expansion, not to sell more Chevrolets. This separation of motive from outcome helps isolate the causal effect.

Preview of Findings

The results paint a striking picture of “commercial imperialism.” CIA interventions increased US imports from affected countries by approximately 34%—yet total imports remained unchanged. This reveals trade diversion, not trade creation: American firms displaced other suppliers rather than expanding the overall market. More tellingly, the effect concentrated in sectors where the US had a comparative disadvantage—precisely the opposite of what trade liberalization would predict, and exactly what we would expect from politically-driven favoritism toward uncompetitive American industries.

Part I: Historical Context and Data

1. The History of Power and Trade

The relationship between political power and commerce is ancient. The British Empire’s “Imperialism of Free Trade” used naval supremacy and colonial administration to open markets on favorable terms. China’s “unequal treaties” after the Opium Wars (1842, 1860) granted Western powers privileged trading access. Commodore Perry’s “opening” of Japan (1854) was backed by warships, not competitive prices.

But does political influence shape trade in the modern era of sovereign states, international institutions, and (ostensibly) rules-based commerce?

The Cold War as Laboratory

The Cold War (1947–1991) provides an ideal setting to study this question. The superpowers competed for influence across the developing world, often through covert means. Beginning in the 1990s, CIA documents were declassified, revealing the scope of American intervention:

51 successful interventions across the globe, including:

Year	Country	Operation
1953	Iran	Overthrow of Mossadegh (Operation AJAX)
1954	Guatemala	Overthrow of Árbenz (Operation PBSUCCESS)
1961	Congo	Support for Mobutu after Lumumba assassination
1964	Brazil	Support for military coup against Goulart
1973	Chile	Overthrow of Allende, installation of Pinochet

The Proposed Mechanism

How might CIA interventions affect trade? The authors propose a specific channel:

Lobbying: US firms in uncompetitive industries lobby Washington for overseas assistance
Installation: CIA operations install or sustain leaders sympathetic to American interests
Reciprocity: Installed regimes—owing their positions to Washington—direct government purchases toward American suppliers
Outcome: US imports rise, particularly in sectors where American firms are least competitive

2. Data Sources

The analysis combines several datasets:

Data	Source	Coverage
CIA interventions	Declassified documents; coded as “successful” if CIA installed or sustained a leader	1947–1989
Bilateral trade flows	Correlates of War (COW) Project, originally from IMF Direction of Trade Statistics	1947–1989
Gravity variables	CEPII; Head, Mayer & Ries (2010)	Distance, language, borders, colonial ties
Trade agreements	WTO, various sources	GATT membership, RTAs
Country characteristics	Penn World Tables, Polity IV	GDP, income per capita, democracy

The sample covers 166 countries over 43 years (1947–1989), yielding a rich panel for within-country analysis.

Part II: Theoretical Framework

3. The Gravity Model of International Trade

Why Gravity?

The gravity model is the workhorse of empirical trade analysis. Named for its resemblance to Newton’s law of gravitation, it predicts that bilateral trade is proportional to economic “mass” (GDP) and inversely related to “distance” (trade costs). Despite its simplicity, gravity typically explains a remarkable share of bilateral trade flows.

The authors follow Anderson & van Wincoop (2003), which emphasizes “multilateral resistance”—a crucial concept for proper specification.

Deriving Gravity from Microfoundations

Consumer preferences: Representative consumers in each country have CES (constant elasticity of substitution) preferences over differentiated varieties:

$$U_j = \left( \sum_i n_i \, q_{ij}^{\frac{\sigma-1}{\sigma}} \right)^{\frac{\sigma}{\sigma-1}}$$

where:

$q_{ij}$ = consumption in country $j$ of varieties from country $i$
$n_i$ = number of varieties produced by country $i$
$\sigma > 1$ = elasticity of substitution between varieties

Trade costs: Shipping goods is costly. We model “iceberg” trade costs: to deliver one unit from $i$ to $j$, you must ship $\tau_{ij} \geq 1$ units (the rest “melts” en route). If the factory-gate price in $i$ is $p_i$, the delivered price in $j$ is $p_i \tau_{ij}$.

Optimization: Consumers maximize utility subject to their budget constraint. The first-order conditions yield demand functions, which aggregate to bilateral trade flows:

$$X_{ij} = \frac{Y_i \, Y_j}{Y^W} \left( \frac{\tau_{ij}}{P_i \, P_j} \right)^{1-\sigma}$$

This is the structural gravity equation. Bilateral trade depends on:

Economic mass ($Y_i Y_j / Y^W$): Larger economies trade more
Bilateral trade costs ($\tau_{ij}$): Higher costs reduce trade
Multilateral resistance ($P_i, P_j$): Each country’s average trade costs with all partners

The Multilateral Resistance Problem

The price indices $P_i$ and $P_j$—termed “multilateral resistance” (MR)—capture that bilateral trade depends not just on bilateral costs, but on all alternative trading opportunities.

The econometric challenge: The MR terms are implicitly defined by a system of nonlinear equations and are unobservable. The solution: include exporter×year and importer×year fixed effects, which absorb $P_i$ and $P_j$ completely.

4. The Estimating Equation

From Theory to Regression

Taking logs of the gravity equation for country $c$'s imports from the US in year $t$:

$$\ln \left( \frac{m_{ct}^{US}}{Y_{ct}} \right) = \alpha_c + \alpha_t + \beta \cdot \text{USinfluence}_{ct} + \mathbf{X}_{ct}'\boldsymbol{\gamma} + \varepsilon_{ct}$$

Components:

Term	Interpretation
$\alpha_c$	Country fixed effects (absorb geography, history, permanent MR)
$\alpha_t$	Year fixed effects (absorb US GDP share, global shocks)
$\text{USinfluence}_{ct}$	=1 if CIA installed/supported leader in country $c$, year $t$
$\mathbf{X}_{ct}$	Controls: income per capita, democracy, leader tenure

The parameter of interest is $\beta$: the effect of US political influence on log imports.

5. Identification Strategy

What Identifies $\beta$?

With two-way fixed effects, $\beta$ is identified from within-country variation over time in US influence.

Example: Compare Chile’s imports from the US in:

1971 (Allende in power, USinfluence = 0)
1974 (Pinochet in power, USinfluence = 1)

The country fixed effect absorbs everything permanent about Chile. The year effects absorb global shocks. What remains is the within-Chile change associated with the political transition.

Threats and Tests

Pre-trends test: Event study shows no differential trends before intervention.

Placebo tests:

Soviet interventions reduce US imports (opposite sign)
US interventions don’t affect exports to target countries
No effect on imports from other Western allies

Part III: Empirical Results

6. Main Results

The baseline specification yields:

Specification	USinfluence	SE	Interpretation
FE + B-B (baseline)	0.293	0.109	≈34% increase in US imports

The coefficient of 0.293 implies that CIA interventions increased imports from the US by approximately 34% (= $e^{0.293} - 1$).

Trade Diversion, Not Creation

A critical finding: US imports rise by ~34%, but total imports are unchanged. This is trade diversion—purchases shift from other suppliers to American firms—not trade creation.

Export asymmetry: No effect on US exports to intervened countries. Governments control purchases (imports), not sales (exports).

Part IV: Mechanisms

7. The Comparative Advantage Test

This is the most powerful test in the paper. It distinguishes “commercial imperialism” from trade liberalization.

The Logic

If intervention caused trade liberalization: imports should rise in sectors where the US has comparative advantage.

If intervention caused political favoritism: imports should rise in sectors where the US has comparative disadvantage—sectors where US firms cannot compete on price.

Results

Using Revealed Comparative Advantage (RCA):

$$\text{Effect of intervention in sector } k = 0.525 - 1.202 \times RCA_k$$

US Comparative Position	RCA	Effect
Strong disadvantage	0.1	+50% increase
Neutral	0.44	≈ 0
Strong advantage	0.9	−56% (large negative)

This is the opposite of trade liberalization. The effect is entirely concentrated in sectors where the US is least competitive.

8. The Government Procurement Channel

The entire effect operates through government procurement:

Government Share	Effect of Intervention
0%	No effect
17% (median)	+23%
28% (75th percentile)	+38%

The direct effect is zero. Installed regimes repay Washington by steering government contracts to American firms. Private sector trade is unaffected.

Part V: Conclusion

What We Learned

This paper provides compelling evidence that political power shapes international trade:

Trade was diverted, not created: Total imports unchanged; American firms displaced other suppliers
Effect concentrated in US disadvantage sectors: The opposite of liberalization
Operates through government procurement: No effect in the private sector
US-specific: No spillover to other Western allies
Robust: Stable across specifications; supported by event study and placebo tests

Was It “Commercial Imperialism”?

Yes, in outcome: American commercial interests benefited substantially. US firms in uncompetitive sectors gained market access they could not have won on price.

Qualified, in intent: Interventions were primarily motivated by Cold War ideology, not commerce. But American businesses were significant beneficiaries even if not primary motivators.

References

Main Paper:

Berger, D., Easterly, W., Nunn, N., & Satyanath, S. (2013). Commercial Imperialism? Political Influence and Trade During the Cold War. American Economic Review, 103(2), 863-896.

Theoretical Foundations:

Anderson, J. E., & van Wincoop, E. (2003). Gravity with Gravitas: A Solution to the Border Puzzle. American Economic Review, 93(1), 170-192.
Baier, S. L., & Bergstrand, J. H. (2009). Bonus Vetus OLS: A Simple Method for Approximating International Trade-Cost Effects Using the Gravity Equation. Journal of International Economics, 77(1), 77-85.

Historical Sources:

Weiner, T. (2007). Legacy of Ashes: The History of the CIA. Doubleday.
Kinzer, S. (2006). Overthrow: America’s Century of Regime Change from Hawaii to Iraq. Times Books.