Data Visualization by Jona Wilke
Between 1947 and 1989, the CIA intervened in 51 countries. Research shows these interventions increased American imports by a third—not by opening markets, but by installing friendly governments.
About the paper
Political influence operates through back channels—covert operations, diplomatic pressure, secret funding. How do you measure the trade effects of something designed to be invisible? Berger, Easterly, Nunn & Satyanath (2013, American Economic Review) found a way: declassified CIA documents that reveal exactly when and where the US intervened during the Cold War.
The identification strategy rests on a crucial insight: CIA interventions were motivated by Cold War ideology—containing communism—not by commercial interests. This separation of political motive from economic outcome creates a natural experiment. The authors embed intervention status in a gravity model of trade, the workhorse of empirical trade research, controlling for country size, distance, and time-invariant characteristics with two-way fixed effects.
The key test is an event study. If the US simply intervened in countries that were already becoming trading partners, we would see rising trade before intervention. The data show flat pre-trends—no differential patterns before CIA involvement—followed by a sharp, persistent increase in US imports afterward. Three additional tests sharpen the result: Soviet interventions show the opposite effect, the gains concentrate in sectors where the US is least competitive, and the entire effect operates through government procurement.
This is what American influence looked like.
The map shows every country where the CIA conducted covert operations during the Cold War. Color intensity reflects the share of years each country spent under US-backed leadership.
These weren't subtle diplomatic efforts. They included coups, election manipulation, and long-term support for authoritarian regimes. The data comes from declassified CIA documents and academic reconstructions of Cold War operations.
The study uses an “installation” variable: a country-year is coded as under US influence if the leader was installed or strongly supported by American covert action. This operational definition captures both the initial intervention and the duration of its political effects.
Latin America: The laboratory for regime change.
Guatemala, Chile, Brazil, Nicaragua—the Americas saw the most intensive intervention. Operation PBSUCCESS in Guatemala (1954) became the template for Cold War regime change: install a military leader, provide ongoing support, and maintain influence through economic and military aid.
The pattern repeated across the hemisphere. When Jacobo Árbenz threatened United Fruit Company's land holdings, he was overthrown. When Salvador Allende nationalized copper mines, he was removed. The economic interests were never far from the political calculations.
Guatemala spent 74% of the Cold War era under US-installed or US-supported governments—the highest intensity in the dataset.
The Middle East: Where oil met ideology.
Iran saw the longest continuous US influence—35 years from the 1953 coup against Mossadegh until the 1979 revolution. The Shah's regime didn't just open Iranian markets to American oil interests; it became a cornerstone of Cold War strategy in the region.
The 1953 coup (Operation AJAX) set a precedent: when nationalist leaders threatened Western commercial interests—in this case, the Anglo-Iranian Oil Company—regime change followed. The Cold War provided ideological cover, but the economic stakes were unmistakable.
Jordan, Saudi Arabia, and Lebanon served as pro-Western anchors, receiving substantial American support in exchange for political alignment and favorable treatment of US business interests.
Southeast Asia: The domino theory in practice.
From the Philippines to Indonesia to Vietnam, the US built a network of anti-communist allies. The “Secret War” in Laos became the largest covert operation in CIA history, with more bombs dropped on this small nation than on Germany in all of World War II.
Indonesia's 1965-66 transition—which left between 500,000 and one million dead— shifted the country firmly into the Western camp. The new regime immediately reversed nationalization policies and opened the economy to foreign investment.
When these allies purchased goods, did political loyalty translate into commercial advantage for American firms? Or was Cold War intervention purely ideological? The data tells a clear story.
This is what American influence looked like.
The map shows every country where the CIA conducted covert operations during the Cold War. Color intensity reflects the share of years each country spent under US-backed leadership.
These weren't subtle diplomatic efforts. They included coups, election manipulation, and long-term support for authoritarian regimes. The data comes from declassified CIA documents and academic reconstructions of Cold War operations.
The study uses an “installation” variable: a country-year is coded as under US influence if the leader was installed or strongly supported by American covert action. This operational definition captures both the initial intervention and the duration of its political effects.
Latin America: The laboratory for regime change.
Guatemala, Chile, Brazil, Nicaragua—the Americas saw the most intensive intervention. Operation PBSUCCESS in Guatemala (1954) became the template for Cold War regime change: install a military leader, provide ongoing support, and maintain influence through economic and military aid.
The pattern repeated across the hemisphere. When Jacobo Árbenz threatened United Fruit Company's land holdings, he was overthrown. When Salvador Allende nationalized copper mines, he was removed. The economic interests were never far from the political calculations.
Guatemala spent 74% of the Cold War era under US-installed or US-supported governments—the highest intensity in the dataset.
The Middle East: Where oil met ideology.
Iran saw the longest continuous US influence—35 years from the 1953 coup against Mossadegh until the 1979 revolution. The Shah's regime didn't just open Iranian markets to American oil interests; it became a cornerstone of Cold War strategy in the region.
The 1953 coup (Operation AJAX) set a precedent: when nationalist leaders threatened Western commercial interests—in this case, the Anglo-Iranian Oil Company—regime change followed. The Cold War provided ideological cover, but the economic stakes were unmistakable.
Jordan, Saudi Arabia, and Lebanon served as pro-Western anchors, receiving substantial American support in exchange for political alignment and favorable treatment of US business interests.
Southeast Asia: The domino theory in practice.
From the Philippines to Indonesia to Vietnam, the US built a network of anti-communist allies. The “Secret War” in Laos became the largest covert operation in CIA history, with more bombs dropped on this small nation than on Germany in all of World War II.
Indonesia's 1965-66 transition—which left between 500,000 and one million dead— shifted the country firmly into the Western camp. The new regime immediately reversed nationalization policies and opened the economy to foreign investment.
When these allies purchased goods, did political loyalty translate into commercial advantage for American firms? Or was Cold War intervention purely ideological? The data tells a clear story.
But total imports remained unchanged. This was trade diversion, not trade creation. American firms displaced other suppliers—they didn't expand the market.
Correlation isn't causation. How do we know the interventions caused this?
The gravity model: how economists measure trade.
International trade follows a remarkably consistent pattern: countries trade more with large, nearby partners. This “gravity model” of trade—named for its similarity to Newton's law—is the workhorse of empirical trade research.
The researchers estimate: how much did US imports from a country change after CIA intervention, controlling for the country's size, distance from America, and time-invariant characteristics? The coefficient captures the “extra” trade attributable to political influence, beyond what gravity predicts.
But correlation isn't causation. Perhaps America simply intervened in countries that were already becoming trading partners. The event study addresses this by examining trade patterns before and after intervention.
Before intervention: no differential trends.
The pre-intervention period is the crucial test. Look at years t-3 to t-1. The coefficients cluster around zero, and the confidence intervals all cross the zero line. Countries that would later experience CIA intervention showed no unusual trade patterns beforehand.
This is what economists call the “parallel trends assumption.” Before intervention, future-intervened countries behaved like control countries. There's no evidence the US was simply targeting countries already predisposed to trade with America.
The flat pre-trends rule out reverse causation. The US wasn't simply intervening in countries that were already buying American goods. Something changed after intervention.
After intervention: a persistent trade shift.
Starting around t+2, US imports increase significantly. The confidence intervals move above zero and stay there. By t+5, imports are roughly 28% higher than the counterfactual—what they would have been without intervention.
The timing is consistent with the political economy story. Regime change takes time to translate into economic policy. New leaders must consolidate power, restructure government procurement, and build relationships with American firms. The lag suggests a causal mechanism, not mere statistical noise.
Placebo test: Soviet interventions show the opposite effect— reduced US imports. This is US-specific, not a general consequence of superpower involvement or political instability.
The gravity model: how economists measure trade.
International trade follows a remarkably consistent pattern: countries trade more with large, nearby partners. This “gravity model” of trade—named for its similarity to Newton's law—is the workhorse of empirical trade research.
The researchers estimate: how much did US imports from a country change after CIA intervention, controlling for the country's size, distance from America, and time-invariant characteristics? The coefficient captures the “extra” trade attributable to political influence, beyond what gravity predicts.
But correlation isn't causation. Perhaps America simply intervened in countries that were already becoming trading partners. The event study addresses this by examining trade patterns before and after intervention.
Before intervention: no differential trends.
The pre-intervention period is the crucial test. Look at years t-3 to t-1. The coefficients cluster around zero, and the confidence intervals all cross the zero line. Countries that would later experience CIA intervention showed no unusual trade patterns beforehand.
This is what economists call the “parallel trends assumption.” Before intervention, future-intervened countries behaved like control countries. There's no evidence the US was simply targeting countries already predisposed to trade with America.
The flat pre-trends rule out reverse causation. The US wasn't simply intervening in countries that were already buying American goods. Something changed after intervention.
After intervention: a persistent trade shift.
Starting around t+2, US imports increase significantly. The confidence intervals move above zero and stay there. By t+5, imports are roughly 28% higher than the counterfactual—what they would have been without intervention.
The timing is consistent with the political economy story. Regime change takes time to translate into economic policy. New leaders must consolidate power, restructure government procurement, and build relationships with American firms. The lag suggests a causal mechanism, not mere statistical noise.
Placebo test: Soviet interventions show the opposite effect— reduced US imports. This is US-specific, not a general consequence of superpower involvement or political instability.
Liberalization or favoritism?
We've established that CIA interventions increased US imports. But through what mechanism? The Cold War narrative emphasized free markets and economic liberalization as alternatives to communism. If interventions simply opened previously closed economies, we'd expect trade to follow comparative advantage.
Comparative advantage is the foundation of trade theory: countries should export what they're relatively good at producing. If the US has comparative advantage in aircraft and computers, liberalization should increase US exports of those goods.
The chart tests this hypothesis. RCA (Revealed Comparative Advantage) measures where the US is actually competitive in world markets. An RCA above 1 means the US exports more of that good than the world average. Below 1 means the US is at a disadvantage—other countries produce it more efficiently.
The pattern reveals political favoritism, not market opening.
The results directly contradict the liberalization story. The trade effect is concentrated in sectors where the US has comparative disadvantage— exactly where American firms couldn't compete on merit against foreign suppliers.
In sectors where the US should win in a free market (high RCA), the intervention effect is close to zero. But in sectors where the US is uncompetitive (low RCA), American imports surge after intervention. Friendly regimes weren't opening markets; they were steering purchases toward American suppliers despite better alternatives.
This is the smoking gun. Market liberalization would show the opposite pattern— gains where the US is competitive. Instead, we see gains precisely where the US shouldn't win. This is political favoritism, not free trade.
The channel: government procurement as patronage.
How did installed regimes deliver commercial benefits to American interests? The mechanism is government procurement. When the researchers split the data by who is doing the buying, the effect operates entirely through government purchases. The direct effect on private-sector trade is statistically indistinguishable from zero.
This makes institutional sense. A newly installed leader controls what the government buys. They don't control what private citizens and businesses purchase. Military equipment, infrastructure projects, public works—these contracts could be steered to American firms as repayment for US support.
The effect scales with government size. Countries with larger public sectors show stronger trade effects. At the median government share, intervention increased US imports by 23%. At the 75th percentile, the effect was 38%. More government purchasing power meant more opportunities for political favoritism.
The picture is clear: installed regimes used state purchasing power to reward American commercial interests. This was commercial imperialism operating through political channels—not markets, but power.
Liberalization or favoritism?
We've established that CIA interventions increased US imports. But through what mechanism? The Cold War narrative emphasized free markets and economic liberalization as alternatives to communism. If interventions simply opened previously closed economies, we'd expect trade to follow comparative advantage.
Comparative advantage is the foundation of trade theory: countries should export what they're relatively good at producing. If the US has comparative advantage in aircraft and computers, liberalization should increase US exports of those goods.
The chart tests this hypothesis. RCA (Revealed Comparative Advantage) measures where the US is actually competitive in world markets. An RCA above 1 means the US exports more of that good than the world average. Below 1 means the US is at a disadvantage—other countries produce it more efficiently.
The pattern reveals political favoritism, not market opening.
The results directly contradict the liberalization story. The trade effect is concentrated in sectors where the US has comparative disadvantage— exactly where American firms couldn't compete on merit against foreign suppliers.
In sectors where the US should win in a free market (high RCA), the intervention effect is close to zero. But in sectors where the US is uncompetitive (low RCA), American imports surge after intervention. Friendly regimes weren't opening markets; they were steering purchases toward American suppliers despite better alternatives.
This is the smoking gun. Market liberalization would show the opposite pattern— gains where the US is competitive. Instead, we see gains precisely where the US shouldn't win. This is political favoritism, not free trade.
The channel: government procurement as patronage.
How did installed regimes deliver commercial benefits to American interests? The mechanism is government procurement. When the researchers split the data by who is doing the buying, the effect operates entirely through government purchases. The direct effect on private-sector trade is statistically indistinguishable from zero.
This makes institutional sense. A newly installed leader controls what the government buys. They don't control what private citizens and businesses purchase. Military equipment, infrastructure projects, public works—these contracts could be steered to American firms as repayment for US support.
The effect scales with government size. Countries with larger public sectors show stronger trade effects. At the median government share, intervention increased US imports by 23%. At the 75th percentile, the effect was 38%. More government purchasing power meant more opportunities for political favoritism.
The picture is clear: installed regimes used state purchasing power to reward American commercial interests. This was commercial imperialism operating through political channels—not markets, but power.
CIA interventions during the Cold War systematically redirected trade toward American firms. Not through market liberalization, but through political favoritism channeled through government procurement.
Berger, Daniel, William Easterly, Nathan Nunn, and Shanker Satyanath. “Commercial Imperialism? Political Influence and Trade During the Cold War.” American Economic Review 103, no. 2 (2013): 863-96.
Visualization by Jona Wilke