Export Dynamics in Colombia: Firm-Level
March 3, 2016
In 1995 President Bill Clinton released the ﬁrst version of the ”Clinton list.” In hopes of
curbing the laundering of Colombian drug money, the Clinton administration ﬁned U.S.
ﬁrms and U.S. Citizens (also permanent residents) who did business with companies and
individuals on this famed list. This will serve as our backdrop as Eaton et al∗ research
entry and exit in the Colombian export market.
In this paper, our authors have both investigated aggregate export growth and de-
composed export growth. Their decomposition, researching the decision to export at all
versus how much to export, reveals some patterns to note.
• Aggregate export ﬂuctuation can be explained solely by the behavior of ﬁrms who
have been exporting for at least one year.
• Firms who export in years t and t+1 grow rapidly after.
• There is a ”right” way to expand exports. Firms that export to Latin American mar-
kets ﬁrst (notably Ecuador and Venezuela) end up expanding into the United States
and other foreign markets. The inverse does not hold true.
∗Eaton, Eslava, Kugler, Tybout (2007)
2 Theoretical Comments
While this paper does not seek to expand our theoretical understanding of export
dynamics, it does attempt to apply external validation to the results by Eaton, Kortum
and Kramarz (2004) (henceforth EKK). EKK address Melitz (2003) theory and expand
upon it by researching the anatomy of trade among French ﬁrms. While EKK did not
reconstruct the model verbatim, their explanation of the model was succinct and made
the empirical results tractable within the framework of the model. If Eaton et al could
follow suit by adding a brief discussion concerning the pertinent economic theory in this
paper, it would prove helpful to the reader.
3 Empirical Comments
This paper addresses an important topic in both the ﬁelds of trade and economic de-
velopment. Policy makers would do well to take heed of their conclusions; if a ”winning”
strategy to exporting existed, Colombia’s private sector would beneﬁt from inducing all
ﬁrms to execute this strategy (much like what happened in Chile) and, less obviously, an
increase in foreign exchange reserves (serving as a macroeconomic buffer). However, the
research methodology gives me pause. I have the following concerns.
• Internal Validation
– It has been well documented in the growth literature that country effects mat-
ter (Barro 1989, Asiedu 2006). As prefaced in the summary, Colombia went
through decades of political instability which led to notable effects of uncer-
tainty. The drug war waged between the Colombian government and the king-
pins lead to terrorist acts across the country. Firms who are willing to under-
take the large sunk costs associated with exporting would only do so if they
believe their ﬁrm would remain intact the next day (this assurance could have
been guaranteed through protection bribes to the maﬁa or the military). If this
were to occur, even once, it would not be accurate to claim that all ﬁrms have
perfect, or equal, foresight and only those with ”better” vision, those who pur-
chased informal insurance, will export in time t and t + 1.
– Country instability in its most general form, may not go far enough to address
ﬁrm export behavior. From the 1980s until recently Colombia has struggled to
quell a booming drug business, fueled by international consumption of con-
trolled substances such as cocaine. President Clinton was the ﬁrst of many
presidents to ask the Treasury to monitor ﬁrms who are associated with the
illicit drug trade. Therefore, due to unique conditions in Colombia, ﬁrms faced
differentiated trade barriers associated with reputation that would explain the
empirical ﬁndings. This potential source of selection bias leaves me to question
the internal validity of these results.
– In 1999 Colombia experienced severe inﬂation as the Central Bank decided to
allow their currency to ﬂoat for the ﬁrst time since the institutionalization of the
Peso. It is possible that only ﬁrms who had assets denominated in other cur-
rencies, most likely the dollar, were the ones to weather the storm. As Kamil
writes in How Do Exchange Rate Regimes Affect Firms’ Incentives to Hedge Cur-
rency Risk in Emerging Markets? (2009) ﬁrms react to a ﬂoating exchange rate by
reducing foreign borrowing and hedging a higher share of their dollar liabili-
ties by increasing export revenues and assets denominated in foreign currency.
Since only large ﬁrms may have the resources to do so, this would call into
question the authors’ interpretation of the data.
– Having a beautiful mountain scenery comes with great cost to the Colom-
bian transportation infrastructure network. Only 20% of roads in Colombia
are paved, and for such a mountainous country that means one must choose
between ﬂying manufactured goods to the nearest port, the expensive option,
or sending the goods through a poorly maintained dirt road network, differ-
entiating trade costs given the location of your business in Colombia. In Who’s
Getting Globalized (Atkin and Donaldson 2012) within country trade costs for
Ethiopia and Nigeria can range between 7 and 15 times more than trade costs
between the US and Canada. Firms may face different trade costs based on
location within country. This could explain their differential export participa-
tion. Even further, different regions experienced varied intensities of the drug
war throughout the sample years. These two effects would call into question
the cross-sectional and the time pattern export participation results.
• External Validation
– As a matter of geography, Colombia is the only country in South America
which borders both the Paciﬁc and the Caribbean Ocean. This allows ﬁrms
to export to either ocean without incurring the fee associated with utilizing the
Panama canal. If this holds true for only Colombia. How could we generalize
this result to their South American neighbors?
– Many of these questions can be addressed via robustness checks of Colombia’s
South American neighbors. If these results hold in the Ecuadorian or Venezue-
lan export market, we can throw out geographical and historical constraints,
allowing us to expand our results to the rest of the continent.
To conclude these remarks, I believe that this paper needs to acknowledge the existence
of Colombia speciﬁc factors in the empirical analysis. Otherwise, one should question the
external and possibly the internal validity of the conclusions. Colombia, Venezuela and
Ecuador are all virtual neighbors, but by having such a distinct geography and history,
one could argue that these conclusions will not hold for Colombia’s neighbors.
Asiedu, Elizabeth. ”Foreign Direct Investment in Africa: The Role of Natural Re-
sources, Market Size, Government Policy, Institutions and Political Instability.” United
Nations UNiversity. United Nations University, 2006. Web. 25 Feb. 2016.
Atkin, David, and Dave Donaldson. ”Who’s Getting Globalized? The Size and Impli-
cations of Intra-national Trade Costs.” (2015): n. pag. Web.
Eaton, Johnathan, Samuel Kortum, and Francis Kramarz. ”An Anatomy of Interna-
tional Trade: Evidence From French Firms.” Econometrica 79.5 (2011): 1453-498. Web.
Kamil, Herman. ”How Do Exchange Rate Regimes Affect Firms’ Incentives to Hedge
Currency Risk? Micro Evidence for Latin America.” IMF Working Papers 12.69 (2012): 1.
Melitz, Marc J. ”The Impact of Trade on Intra-industry Reallocations and Aggregate
Industry Productivity.” Econometrica 71.6 (2003): 1695-1725.