Berger et al. (2013b)

• Abstract: We provide evidence that increased political influence, aris-ing from CIA interventions during the Cold War, was used to create alarger foreign market for American products. Following CIA interven-tions, imports from the US increased dramatically, while total exportsto the US were unaffected. The surge in imports was concentrated inindustries in which the US had a comparative disadvantage, not a com-parative advantage. Our analysis is able to rule out decreased trade costs,changing political ideology, and an increase in US loans and grants asalternative explanations. We provide evidence that the increased importsarose through direct purchases of American products by foreign govern-ments.

1

Background• The general question of whether power is an important determinant of

international trade, particularly in the more recent past, is difficult toexamine empirically because shifts in power relations between govern-ments are often the result of decisions that are made behind the veil ofgovernment secrecy.

– History provides us with many examples of the use of political powerto promote trade and other national interests, the starkest being theunequal treaties imposed by Western powers on China and Japanduring the nineteenth and early twentieth centuries.

• We surmount this problem by relying on the use of recently declassifiedCIA documents to generate a country- and year-specific measure of theinfluence of the US government over foreign countries.

– CIA documents older than 25 years fall under the Freedom of Infor-mation Act.

2

Related Studies: CIA during Cold War

• Dube et al. (2011) examine the stock prices of US companies in Iran,Guatemala, Cuba, and Chile before and after the CIA authorized plansfor covert coups. They find that the stock returns of companies that wereboth connected to the CIA and stood to gain from the coups increasedimmediately after the authorizations.

• Berger et al. (2013a) examine the effect of interventions on democracy,and find that CIA and KGB interventions have a negative effect on sub-sequent democracy.

– They use lower frequency data at five year intervals.

3

Related Studies: Other International Settings

• Dreher and Jensen (2007) show that IMF conditionality is correlatedwith whether countries vote in-line with the US in the UN General As-sembly.

• Kilby (2009) shows that the World Bank’s structural adjustment condi-tions are less stringent for countries whose voting in the UN is morealigned with the US.

• Kuziemko and Werker (2006) show that when countries have a seat onthe UN security council they receive more foreign aid from the US.

4

(Successful) CIA Intervation Measure• USinfluencet,c is an indicator variable that equals 1, in a country c and

year t, if the CIA either installed a foreign leader or provided covertsupport for the regime once in power.

– The activities used by the CIA to install and help maintain the powerof specific regimes included:∗ The creation and dissemination of (often false) propaganda, usu-

ally through radio, television, newspapers, and pamphlets.∗ Covert political operations, which typically consisted of the pro-

vision of funds and expertise for political campaigns.∗ The destruction of physical infrastructure and capital, as well as

covert paramilitary operations, that included the supply of armsand military equipment, direct involvement in insurgency and coun-terinsurgency operations, and the coordination of coups and assas-sinations.

5

– There are many instances in which the CIA set out to remove anexisting leader and install a new leader in power.

∗ The CIA-organized coups in Iran in 1953, Guatemala in 1954, andChile in 1973 are the most well-known examples of such cases.

– In other cases, the CIA did not engage in activities to install theleader into power, but once in power, at some point, began to en-gage in activities to help maintain the power of the regime.

∗ Typically, these were covert counterinsurgency operations under-taken by the CIA.

• We restrict our analysis to the Cold War period, 1947–1989.

6

Illustration: CIA in Chile

• CIA involvement in Chile first occurred in the 1964 Presidential elec-tion, when the CIA provided covert funding and support for the Chris-tian Democratic Party candidate Eduardo Frei Montalva. Eduardo Freiwon the election and continued to receive CIA support while he was inpower. In the 1970 election, Salvador Allende (he was not installed orsupported by the CIA), a candidate from a coalition of leftist parties, waselected, and remained in power until the famous CIA orchestrated coupof 1973. After the coup, Augusto Pinochet took power and was backedby the CIA until 1988.

• USinfluencet,c equals 1 from 1964 to 1970, 0 in 1971 and 1972, and then1 from 1973 to 1988.

7

8

Trade Data: COW Dataset

• The Correlates of War (COW) Trade Dataset reports annual aggregatebilateral trade flows (measured in millions of nominal US dollars).

– For the post WWII period, the data are originally from the Interna-tional Monetary Fund’s Direction of Trade Statistics.

– The COW dataset imputes missing flows by using, for example, theexporter’s trade statistics if data on imports are missing from theimporter’s accounts.

– Because importing countries typically keep more precise records ofshipments (because of the existence of tariffs) than exporting coun-tries, the dataset uses importing country accounts when both sourcesexist.

9

Estimated Relationship

lnmUS

t,c

Yt,c= αt +αc +βUSinfluencet,c +ϕ ln

τUSt,c

PUSt Pt,c

+Xt,cΓ+ εt,c.

• mUSt,c is the imports into c from the US normalized by c’s total GDP.

• Xt,c includes ln per capita income, an indicator for Soviet/KGB interven-tions, an indicator variable that equals 1 if there is a change in leadershipas well as a measure of the tenure of the current leader,

– Recent studies show that leaders matter (e.g., Jones and Olken (2009)).

and an indicator variable that equals 1 if an observation is democracy.

– Berger et al. (2013a) shows that successful CIA interventions ad-versely impacted democracy, as defined by Cheibub et al. (2010).

10

• For the bilateral trade costs τt,c,e when shipping goods from c to e,

lnτt,c,e = µ1 lndistc,e + µ2Ilangc,e + µ3Iborder

c,e + µ4Igattt,c,e + µt Irta

t,c,e,

where distc,e is the distance between the two countries, Ilangc,e is an indica-

tor variable that equals 1 if the two countries share a common language,Iborderc,e is an indicator for them sharing a border, Igatt

t,c,e is an indicator thatquals 1 if both are GATT participants in year t, and Irta

t,c,e equals 1 if bothcountries belong to a regional trade agreement in t.

• As in Baier and Bergstrand (2009),

lnPt,c + lnPt,e =N

∑i=1

θt,i lnτt,c,i +N

∑j=1

θt, j lnτt, j,e −N

∑k=1

N

∑m=1

θt,kθt,m lnτt,k,m,

where θt,i ≡ Yt,i/YWt .

– Yt,e is total GDP of e in t, and YWt is world GDP in t.

11

12

13

Underlying Mechanisms: Government Purchase• As a share of GDP, government purchases have typically been around 20

percent for industrialized nations and 15 percent for developing nations.Removing compensation to employees and focusing only on purchasesof goods, the figures become 10.3 and 8.8 percent, respectively.

– In addition, it is well-known that government purchases are highlydiscriminatory, with suppliers typically based on criteria other thanlowest costs, and that influence, power, and connections are impor-tant factors that affect governments’ choice of suppliers.

– The share of government expenditures in GDP is taken from the PennWorld Tables 6.3.

• The industry level data are from the United Nations’ Comtrade Database.

– Only begin in 1962, and therefore the sample only includes yearsbetween 1962 and 1989.

14

– 2-digit Standard International Trade Classification (SITC) industries.

∗ Industry fixed effects are also included.

– Utilizing a South Korean 413-industry Input-Output (I-O) table from2000, we construct measures of the proportion of total output fromeach industry that is sold to the government and the share of eachindustry’s imports purchased by the government, and divides indus-tries into those with above median levels of government purchases(or imports) and those with below median levels.

15

Interaction

• y = a+(b+ cz)x⇐⇒ y = a+bx+ cxz

–

y x z

15 2 1

10 1 2

22 3 1

→

y x xz

15 2 2

10 1 2

22 3 3

→ y = 1+(5+2z)x

16

• The effect of CIA interventions on observations between the tenth andnineteenth percentiles range from 0.099 to 0.373.

– The government expenditure shares for observations at the tenth andnineteenth percentiles are 0.077 and 0.277.

17

Testing Alternative Explanations: Trade Integration

• Balassa (1965)’s measure of revealed comparative advantage:

RCAt,c,i =xt,c,i

∑c xt,c,i

/∑i xt,c,i

∑i ∑c xt,c,i,

where xt,c,i denotes the aggregate exports of country c in a 2, 3, or 4-digitSITC industry i in year t.

• The market size of less developed countries (LDCs) is much smallerthan of DCs.

– An alternative measure of RCA is also calculated using only theshare of exports to LDCs, rather than the share of exports globally.

18

• ImporterRCAt,c,i and USInfluencet,c×ImporterRCAt,c,i are also included.

19

Political Ideology• Existing evidence that countries with more similar political ideologies

trade more.

• We estimate a regression that examines each country’s imports from allexporters (with country-pair fixed effects included), not just the US.

• Using voting data from the UN General Assembly,

VUSt,c ≡ 1−

dt,c

maxc′ dt,c′.

– dt,c is the sum of the vote distance between country c and the US forall votes in year t, where a vote in opposition to the US is given adistance of 1, and a vote with the US is given a distance of 0.

• We also measure VUSt,e using indicator variables that equal 1 if: (i) ex-

porter e was a NATO member, (ii) e was among the original OECDmembers (from 1961), or (iii) e is from Western Europe (or is the USA).

20

21

US Loans and Grants

• We examine the value of US economic aid (which includes grants andconcessional loans) and military aid (which includes grants, conces-sional loans, and training) received by each country.

– Much of military aid is spent on goods exported overseas to supportUS troops and other personnel, and these goods are not included inthe IMF’s Direction of Trade statistics.

• We also examine loans given by the US Export-Import Bank (Ex-ImBank), an institution with a mandate to provide loans to foreign firmsthat want but are having trouble obtaining financing from private lenders.

22

23

References

Baier, Scott L. and Jeffrey H. Bergstrand (2009) “Bonus vetus OLS: A simplemethod for approximating international trade-cost effects using the gravityequation,” Journal of International Economics, Vol. 77, No. 1, pp. 77–85.

Balassa, Bela (1965) “Trade Liberalisation and “Revealed” Comparative Ad-vantage,” The Manchester School, Vol. 33, No. 2, pp. 99–123.

Berger, Daniel, Alejandro Corvalan, William Easterly, and ShankerSatyanath (2013a) “Do superpower interventions have short and long termconsequences for democracy?” Journal of Comparative Economics, Vol.41, No. 1, pp. 22–34. Symposium in Honor of Thrainn Eggertson.

Berger, Daniel, William Easterly, Nathan Nunn, and Shanker Satyanath(2013b) “Commercial Imperialism? Political Influence and Trade duringthe Cold War,” American Economic Review, Vol. 103, No. 2, pp. 863–96.

24

Cheibub, Jose Antonio, Jennifer Gandhi, and James Raymond Vreeland(2010) “Democracy and dictatorship revisited,” Public Choice, Vol. 143,No. 1-2, pp. 67–101.

Dreher, Axel and Nathan M. Jensen (2007) “Independent Actor or Agent?An Empirical Analysis of the Impact of U.S. Interests on InternationalMonetary Fund Conditions,” Journal of Law and Economics, Vol. 50, No.1, pp. 105–124.

Dube, Arindrajit, Ethan Kaplan, and Suresh Naidu (2011) “Coups, Corpora-tions, and Classified Information,” The Quarterly Journal of Economics,Vol. 126, No. 3, pp. 1375–1409.

Jones, Benjamin F. and Benjamin A. Olken (2009) “Hit or Miss? The Effectof Assassinations on Institutions and War,” American Economic Journal:Macroeconomics, Vol. 1, No. 2, pp. 55–87.

25

Kilby, Christopher (2009) “The political economy of conditionality: An em-pirical analysis of World Bank loan disbursements,” Journal of Develop-ment Economics, Vol. 89, No. 1, pp. 51–61.

Kuziemko, Ilyana and Eric Werker (2006) “How Much Is a Seat on the Se-curity Council Worth? Foreign Aid and Bribery at the United Nations,”Journal of Political Economy, Vol. 114, No. 5, pp. 905–930.

26