Page 5 - Azerbaijan State University of Economics

M.R. Jamilov, R.M. Jamilov: Factor-Augmented J-Curve

management has remained a focal point of interest for empirical trade scholars during
the past three decades. The theoretical foundations originate in Magee (1973) and
Dornbusch and Krugman (1976), who postulated that a cheaper currency should carry a
positive net impact on the balance of trade. The empirical complexity arises in the
dynamic of simultaneous reactions of both exports and imports to an exogenous
currency devaluation shock. It is generally expected that in the short run a devalued
currency will be more flexible and trigger a decline in the value of exports and a rise in
imports, due to the so-called “price effect”. However, in the long run the selling power
of exporters (because of the cheaper currency) increases, exportation expands, and
eventually overpowers the rise in imports via the “quantity effect”. If the volume effect
dominates the price effect, or in other words – the long-run elasticity of the trade
balance in response to the exchange rate shock is larger than unity – then we observe the
so-called Marshall-Lerner condition. If plotted over time, the dynamic of the balance of
trade will resemble the letter “J”, leading to the now famous J-curve effect.
Bahmani-Oskooee (1985) provides a pioneering empirical investigation of the J-
curve phenomenon using aggregated data [Some other examples of the aggregated approach include
Narayan (2004), Halicioglu (2007), and Hsing (2008)]. The presumably infective aggregation bias,
present in all aggregated approaches to the question, is solved in Rose and Yellen (1989),
who proposed to treat the matter with disaggregated, country-specific bilateral trade
balance data [Some of the papers belonging to the bilateral approach are Bahmani-Oskooee and Brooks
(1999), Bahmani-Oskooee et al. (2006), Halicioglu (2008), Bahmani-Oskooee and Kutan (2009), Perera
(2011), and Jamilov (2013)] . Starting from Ardalani and Bahmani-Oskooee (2007), however,
literature has glided towards further disaggregation, now to the level of industry-specific
balance of trade parameters [The industrial approach includes such titles as Bahmani-Oskooee and
Wang (2008), Bahmani-Oskooee and Hajilee (2009), Bahmani-Oskooee and Hegerty (2009), Bahmani-
Oskooee and Mitra (2009), Soleymani and Saboori (2012), Bahmani-Oskooee et al. (2013)]. Despite the
plethora of empirical attacks at the J-curve question, results are still substantially
heterogeneous across regions, time periods, and individual industries. A more thorough
review of the J-curve literature is provided by Bahmani-Oskooee and Ratha (2004).
In spite of some mechanical precision of working with highly disaggregated data
series, the industry-level J-curve studies are still not entirely intuitive, are often very
spacious, and the results are rarely illustrative for policy purposes. First, contemporary
industry-level studies fail to explicitly account for mutual commonality of the industries
examined, neglecting the confounding factor of mutual dependency and

5