THE DYNAMIC AND DISTRIBUTIONAL ASPECTS OF IMPORT TARIFFS
| Author | Wolfgang Lechthaler,Mariya Mileva |
| DOI | http://doi.org/10.1111/iere.12484 |
| Published date | 01 February 2021 |
| Date | 01 February 2021 |
INTERNATIONALECONOMIC REVIEW
Vol. 62, No. 1, February 2021 DOI: 10.1111/iere.12484
THE DYNAMIC AND DISTRIBUTIONAL ASPECTS OF IMPORT TARIFFS∗
By Wolfgang Lechthaler1and Mariya Mileva
Kiel Institute for the World Economy, Germany ; California State University, U.S.A.
(E-mail: mariya.mileva@csulb.edu)
We use a dynamic trade model with two sectors and two types of workers to analyze the optimal setting of
income-generating tariffs. Westudy dynamic and distributional aspects focusing on the time horizon of policy-
makers and workers. The level of tariffs preferred by workers depends on the sector where they are employed
as well as their skill class, with the relative weight of both aspects determined by the time horizon of the work-
ers. Unskilled workers in the unskilled-intensive sector are the ones most in favor of protectionism and might
even benefit from a trade war.
1. introduction
Motivated by the surge in protectionist tendencies on campaign trails all over developed
countries, we take a fresh look at the optimal setting of income-generating tariffs. To do so,
we use a dynamic model with two countries, two factors of production, two sectors, endoge-
nous firm entry, and firm heterogeneity. This allows us to draw a rich picture of the dynamic
distributional aspects of import tariffs.2We find that the level of tariffs preferred by a spe-
cific worker depends on the sector where the worker is employed as well as her skill class,
with the relative weight of both aspects determined by the time perspective of the worker.
For a worker who is more concerned about the immediate future, the sector of employment
is more important, for a worker who is more concerned about long-run outcomes, the skill-
class is more relevant. The workers most in favor of tariffs are the unskilled workers in the
unskilled-intensive sector.
The analysis of the setting of optimal tariffs has a long tradition in the trade literature. One
shortcoming of this literature is that it is typically based on static models. This is a shortcom-
ing for at least two reasons. On the one hand, adjustment dynamics are ignored. On the other
hand, tariffs are typically set by elected politicians who tend to care more about the next cou-
ple of years than the infinite future (the new steady state). So, focusing on purely static mod-
els might yield implausible policy conclusions.
Two recent papers (Larch and Lechthaler 2013; Lechthaler 2017) try to close this gap by
analyzing the setting of tariffs in a dynamic version of the Melitz (2003) model. Both papers
consistently find that a shorter time horizon implies lower optimal tariffs because the short-
run effects of higher tariffs are worse than the long-run effects.3However, in both papers,
distributional aspects are missing because they rely on the representative agent framework
∗Manuscript received February 2018; revised February 2020.
1Please address correspondence to: Wolfgang Lechthaler, Kiel Institute for the World Economy, Keil, Germany.
E-mail: wolfgang.lechthaler@ifw-kiel.de.
2Previous papers (Costinot et al., 2016; Demidova and Rodriguez-Clare 2009; Felbermayr et al., 2013; Lechthaler
2017) have shown that including firm heterogeneity is important for understanding the effects of import tariffs. None
of these papers, however, has looked at the dynamic distributional aspects of tariffs that are the focus of this study.
3For a more general discussion of the short-run and long-run effects of structural reforms and related political
economy issues, see Rodrik (1996).
199
© 2020 The Authors. International Economic Review published by Wiley Periodicals LLC on behalf of the Economics
Department of the University of Pennsylvania and the Osaka University Institute of Social and Economic Research
Association
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs
License, which permits use and distribution in any medium, provided the original work is properly cited, the use is
non-commercial and no modifications or adaptations are made.
200 lechthaler and mileva
of the Melitz model. Thus there is only one sector and only one factor of production and
each worker is affected equally by changing tariffs. This is at odds with recent evidence show-
ing that workers employed in import-competing sectors were especially adversely affected by
trade liberalization with China (see, e.g., Autor et al., 2013; Dauth et al., 2014; Ebenstein et al.,
2009; Pierce and Schott 2016).
To be able to capture this differential exposure of workers to international trade as well as
the sluggish adjustment of an economy after a trade shock and the potential short-sightedness
of policymakers, we use the dynamic version of the model of Bernard et al. (2007) developed
in Lechthaler and Mileva (2019), augmented by income-generating tariffs. The model features
two factors of production, skilled and unskilled workers, two sectors that use both factors with
different intensities, and two countries with different endowments of skilled versus unskilled
workers. The model is calibrated to broadly match important features of the two economies of
the United States and China.
Following Ossa (2014), we first analyze unilaterally optimal tariffs, that is, the optimal U.S.
tariff under the assumption that the Chinese tariff does not change.4In response to an in-
crease in the U.S. tariff, aggregate U.S.-consumption increases permanently, but due to en-
hanced firm investment, the increase is very sluggish. Thus, as in Larch and Lechthaler (2013)
and Lechthaler (2017), a policymaker who is interested in maximizing aggregate consumption
sets a lower tariff the shorter his time horizon.
However, workers are very differently affected by the tariff increase. In the short run,
workers in the skill-intensive sector lose because an increase in tariffs partially unwinds the
economy’s specialization in its comparative advantage sector while workers in the unskilled-
intensive sector gain (the United States has a comparative advantage in skill-intensive pro-
duction because it has relatively more skilled workers than China). In the long run, skilled
workers lose while unskilled workers gain, because the unwinding of specialization reduces
the relative demand for skilled workers.
These effects, of course, perfectly resemble the well-known Stolper–Samuelson and
Ricardo–Viner theorems. However, we combine both aspects in a unified framework and
bring them together with preferences for trade protection and a worker’s time perspective.
Thus it is neither solely the skill class of a worker nor the sector where she is employed that
determines her preference for tariffs, but rather a combination of both, with the time perspec-
tive of the worker determining their respective weights.
Thus a skilled worker in the unskilled-intensive sector might prefer free trade if she has
a long-term perspective but favor tariffs if she has a sufficiently short-term perspective. In
contrast, an unskilled worker in the skill-intensive sector favors lower tariffs when she has a
short-term perspective but a higher tariff when she has a long-term perspective. As expected,
unskilled workers in the unskilled-intensive sector are the ones most in favor of tariffs and
even more so if they have a short-term perspective.
It is often argued that the fear of retaliation prevents policymakers from raising tariffs be-
cause in a trade war in which both countries raise their tariff both countries would suffer
lower welfare. Although from an aggregate perspective this is still true in our model, our anal-
ysis adds a nuance to this outcome. In the Nash-equilibrium of a noncooperative game, aggre-
gate consumption and the consumption of most worker groups are lower than in the status
quo but this is not true for the unskilled workers in the unskilled-intensive sector. Although
their consumption in the new steady state with higher tariffs is basically the same as in the
old steady state, during the transition period they enjoy substantial, albeit temporary gains in
consumption. Thus even though the economy as whole would suffer from a trade war, some
workers still gain and are thus willing to support an increase in tariffs even in the face of po-
tential retaliation.
4For most of our analysis,we assume that the tariff is the same for both sectors but in Section 6 we allow for differ-
ent tariffs across sectors.
tariffs: dynamics and distribution 201
Our article connects to three different strands of the literature. First, the large literature
on the setting of Nash-equilibrium and optimal tariffs, see, for example, Johnson (1953), Gros
(1987), Krugman (1991), Bond and Syropoulos (1996), Bagwell and Staiger (1999), Yi (2000),
Ornelas (2005), Demidova and Rodriguez-Clare (2009), Felbermayr et al. (2013), Ossa (2014),
or Nicita et al. (2018). Second, the literature that analyzes the effects of trade with China, for
example, Autor et al. (2013), Dauth et al. (2014), Ebenstein et al. (2009), Pierce and Schott
(2016), Amiti et al. (2019), or Fajgelbaum et al. (2019). And finally, the relatively young but
growing literature on the dynamic adjustment to trade shocks, see, for example, Alessan-
dria and Choi (2014), Melitz and Burstein (2011), Cacciatore (2014), Cacciatore and Ghironi
(2020), Co¸sar (2013), Dix-Carneiro (2014) or Kambourov (2009). None of these papers, how-
ever, has considered the joint analysis of tariffs in a dynamic setting with inter-industry trade.
The rest of the article is structured as follows. Section 2 describes the model. Section 3 dis-
cusses the dynamic adjustment of aggregate variables in response to an increase in tariffs, and
the role of the policymaker’s planning horizon for the optimal unilateral tariff. Section 4 ana-
lyzes the dynamic adjustment of worker-specific variables and worker-specific preferences for
tariffs. Section 5 discusses Nash-equilibrium tariffs (“trade wars”), Section 6 discusses sector-
specific tariffs, and Section 7 concludes.
2. theoretical model
Our model economy consists of two countries, Home (H) and Foreign (F). Each country
produces two goods, good Sand good U. The production of each good requires two inputs,
skilled and unskilled labor. The sector that produces good Sis skill-intensive, that is, the pro-
duction of good Srequires relatively more skilled labor than the production of good U. Coun-
try H has a comparative advantage in producing good Sbecause it has a higher relative en-
dowment of skilled labor. Similarly, F has a comparative advantage in sector Ubecause it has
a higher relative endowment of unskilled labor. We assume that unskilled labor is more abun-
dant than skilled labor in both countries in order to generate a positive skill premium.5We
also assume that incumbent firms and workers are immobile across sectors, that is, they have
to stick to the sector where they have entered. Reallocation across sectors is modeled via re-
tiring firms and workers that are replaced by new entrants. These new entrants can choose
their sector upon entry.6In the following, we describe all the decision problems in H; equiv-
alent equations hold for F.
2.1. Households. In our model, workers are organized within households, where house-
holds are defined by the skill class and the sector of employment of its members, that is, each
member of a household belongs to the same skill class and works in the same sector. Due to
the structure of our model, this implies that in each country there are four types of house-
holds, one consisting of skilled workers employed in sector S, one consisting of skilled workers
employed in sector U, and likewise, one consisting of unskilled workers employed in sector S
and another one consisting of unskilled workers employed in sector U.
Workers are retiring at an exogenous rate sand get replaced by newly entering workers.
These newly entering workers inherit their skill class but are free to choose their sector of em-
ployment based on labor demand (Section 2.2), and stay in that sector for the rest of their
working life. By choosing a sector of employment, they automatically get assigned to the cor-
responding household. The entry decision of newly entering workers thus endogenously de-
termines the size of each household. We assume that consumption and saving decisions are
done by the household not by individual workers, but letting workers decide would yield
5What matters for comparative advantage are relative endowments, so skilled labor can be scarce in both
countries.
6In the Appendix, we consider the case of perfect labor mobility across sectors as well as the case of endogenous
skill formation.
To continue reading
Request your trial