Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Trade Liberalization and Regional Dynamics∗
Rafael Dix-Carneiro†
Duke Universityand BREAD
Brian K. Kovak‡
Carnegie Mellon Universityand NBER
August 2016
Abstract
We study the evolution of trade liberalization’s effects on local labor markets, followingBrazil’s early 1990s trade liberalization. Regions that initially specialized in industries fac-ing larger tariff cuts experienced prolonged declines in formal sector employment and earningsrelative to other regions. The impact of tariff changes on regional earnings 20 years after lib-eralization was three times the size of the effect 10 years after liberalization. These findingsare robust to a variety of alternative specifications and to controlling for a wide array of post-liberalization shocks. The pattern of increasing effects on regional earnings is not consistent withconventional spatial equilibrium models, which predict that effect magnitudes decline over timedue to spatial arbitrage. We investigate potential mechanisms, finding empirical support fora mechanism involving imperfect interregional labor mobility and dynamics in labor demand,driven by slow capital adjustment and agglomeration economies. This mechanism graduallyamplifies the initial labor demand shock resulting from liberalization. We show that the mech-anism explains the slow adjustment path of regional earnings and quantitatively accounts forthe magnitude of the long-run effects.
∗This project was supported by an Early Career Research Grant from the W.E. Upjohn Institute for EmploymentResearch. The authors would like to thank Peter Arcidiacono, Penny Goldberg, Gustavo Gonzaga, Guilherme Hirata,Joe Hotz, Joan Monras, Enrico Moretti, Nina Pavcnik, Mine Senses, Lowell Taylor, Gabriel Ulyssea, Eric Verhoogen,and participants at various conferences and seminars for helpful comments. Ekaterina Roshchina provided excellentresearch assistance. Dix-Carneiro thanks Daniel Lederman and the Office of the Chief Economist for Latin Americaand the Caribbean at the World Bank for warmly hosting him while part of the paper was written. Remaining errorsare our own.†[email protected]‡[email protected]
1
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
1 Introduction
Prominent theories of international trade typically focus on long-run equilibria in which the re-
allocation of resources across economic activities is achieved without frictions. These models have
traditionally given little attention to the adjustment process in transitioning from one equilibrium
to another, creating a tension between academic economists advocating trade liberalization and
policy makers concerned with the labor market outcomes of workers employed in contracting sec-
tors or firms (Salem and Benedetto 2013, Hollweg, Lederman, Rojas and Ruppert Bulmer 2014).
While theory tends to focus on long-run outcomes, empirical studies of the labor market effects
of trade liberalization typically emphasize short- or medium-run effects. Frequently changing de-
signs of cross-sectional household surveys forced researchers to focus on relatively short intervals
to guarantee consistency over the periods analyzed (Goldberg and Pavcnik 2007). Thus, although
many countries underwent major trade liberalization episodes throughout the 1980s and 1990s (e.g.
Brazil, Mexico, and India, among others), we still know very little about the evolution of the effects
of these policy reforms on labor markets.
We fill this gap in the literature by using 25 years of administrative employment data from Brazil
to study the dynamics of local labor market adjustment following the country’s trade liberalization
in the early 1990s. We exploit variation in the degree of tariff declines across industries and
variation in the industry mix of local employment across Brazilian regions to measure changes in
local labor demand induced by liberalization. We then compare formal employment and earnings
growth between regions facing larger and smaller tariff declines, while controlling for pre-existing
trends in these outcomes.1 This approach allows us to observe the ensuing regional labor market
dynamics for 20 years following the beginning of liberalization.
The results are striking. We find large and slowly increasing effects on regional earnings and
employment. Regions facing larger tariff declines experience steadily deteriorating formal labor
market outcomes compared to other regions. These effects grow for more than a decade before
beginning to level off in the late 2000s. This pattern is robust to a wide variety of alternative
measurement strategies, weighting schemes, and controls for pre-existing trends across multiple
decades. The growing effects are not driven by post-liberalization shocks such as later tariff changes,
exchange rate movements, privatization, or the commodity price boom of the 2000s. We conclude
that liberalization’s effects on regional earnings and employment grew substantially over time.
This pattern challenges the conventional wisdom that labor mobility gradually arbitrages away
spatial differences in local labor market outcomes (Blanchard and Katz 1992, Bound and Holzer
2000). If that were the case, one would observe declining regional effects of liberalization on
1In this paper, we focus on formal labor market outcomes, covering workers with a signed work card providingaccess to the benefits and labor protections afforded by the legal employment system. See Dix-Carneiro and Kovak(2015b) for an analysis including the informal labor market, which includes the self-employed and employees withoutsigned work cards.
2
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
earnings, such that the short- and medium-run estimates of trade exposure in prior work would be
an upper bound on the long-run effects.2 Instead, we document increasing effects of liberalization;
the effect on regional earnings 20 years after the start of liberalization is more than 3 times larger
than the effect after 10 years. Liberalization’s long-run effects on regional labor market outcomes
are therefore much larger than initially supposed.
This surprising finding leads us to evaluate a variety of alternative mechanisms that might
account for the growth in liberalization’s effects on regional earnings. The evidence rules out mech-
anisms based on slow urban decline (as in Glaeser and Gyourko 2005), changing worker composition
(based on observable or unobservable characteristics), and slow responses of trade quantities to tariff
changes. Instead, we find strong evidence for a mechanism involving imperfect interregional labor
mobility and dynamics in labor demand, driven by a combination of slow regional capital adjust-
ment and agglomeration economies. Intuitively, as capital slowly reallocates away from harder-hit
regions, workers’ marginal products steadily fall. Similarly, with agglomeration economies, a nega-
tive local labor demand shock decreases local economic activity, reducing regional productivity, and
further decreasing the marginal product of labor. We find minimal responses of regional working-
age population to regional tariff declines, suggesting imperfect worker mobility across regions. In
this setting, dynamic labor demand, driven by slow capital adjustment or agglomeration economies,
can rationalize the steady relative decline in wages in regions facing larger tariff declines.
We present a wide array of evidence in support of this mechanism. Regions facing larger tariff
reductions experience steady declines in the number of formal establishments and declining average
establishment size, suggesting that capital stocks slowly reallocate away from negatively affected
regions. Capital investment shifts away from these regions on impact, with immediate declines in
establishment entry and job creation. In contrast, establishment exit and job destruction increase
slowly over time, consistent with firm owners waiting for installed capital to depreciate before
contracting or closing down regional establishments. Supporting the presence of agglomeration
economies, we show that employment in a given industry × region pair falls more when other
industries in the region face larger tariff cuts. Regional labor market equilibrium would suggest
the opposite in the absence of agglomeration economies (Helm 2016). Finally, we extend the
specific-factors model of regional economies in Kovak (2013) to incorporate slow factor adjustment
and agglomeration economies. Within this framework, we show that a proxy for regional capital
adjustment quantitatively accounts for a substantial portion of the long-run earnings effects that
we observe. Standard magnitude agglomeration economies and perfect long-run capital mobility
quantitatively account for all of the long-run earnings effects. In contrast to the other alternative
mechanisms that we considered, this dynamic labor demand mechanism is both qualitatively and
2Papers documenting short- and medium-run regional effects of trade exposure include Autor, Dorn and Hanson(2013), Costa, Garred and Pessoa (2016), Edmonds, Pavcnik and Topalova (2010), Hakobyan and McLaren (forth-coming), Hasan, Mitra and Ural (2006), Hasan, Mitra, Ranjan and Ahsan (2012), Kondo (2014), Kovak (2013),McCaig (2011), Topalova (2010), and many others.
3
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
quantitatively consistent with the observed earnings responses.
Only recently have researchers begun measuring reallocation costs and the dynamics of labor
market adjustment following trade policy reforms. The papers in this literature calibrate or esti-
mate small open economy models in order to study their quantitative implications for welfare and
the implied transitional dynamics when facing hypothetical changes in trade policy.3 We contribute
to this literature by describing empirical transitional dynamics in response to a real-world trade
liberalization. We document the importance of dynamic labor demand in the evolution of liberal-
ization’s effects on labor markets and suggest that incorporating this mechanism into quantitative
models is an important task for future work.
A growing empirical literature finds substantial differences in the effects of trade exposure across
local labor markets with different industry structures.4 Each of these papers measures the effects of
trade shocks over a fixed time window of 7 to 10 years. We contribute to this literature by placing
the single-year estimates from prior work into a dynamic context, documenting the evolution of
trade liberalization’s regional effects over time. This exercise is possible because our data provide
complete yearly coverage of the formal labor market, even at fine geographic levels, and because
Brazilian liberalization represents a discrete shock occurring during a well-defined time period. A
similar analysis would be much more challenging when studying shocks that continually evolve over
time, such as Chinese export growth, because it is difficult to separate the influence of dynamics
from the effects of newly arriving shocks.5
Our paper proceeds as follows. Section 2 describes the history and institutional context of
Brazil’s early 1990s trade liberalization. Section 3 describes the data sources, local labor market
definition, and empirical approach. Section 4 presents i) our main results for liberalization’s effects
on regional earnings and employment, ii) a wide array of robustness tests, and iii) analyses ruling
out the influence of post-liberalization shocks. Section 5 evaluates potential mechanisms that could
account for the growing earnings effects of liberalization. Section 6 concludes.
2 Trade Liberalization in Brazil
Brazil’s trade liberalization in the early 1990s provides an excellent setting in which to study the
labor market effects of changes in trade policy. The unilateral trade liberalization involved large
declines in average trade barriers and featured substantial variation in tariff cuts across industries.
Many papers have examined the labor market effects of trade liberalization in the Brazilian context
3Examples include Artuc, Chaudhuri and McLaren (2010), Caliendo, Dvorkin and Parro (2015), Cosar (2013),Dix-Carneiro (2014), Kambourov (2009), Traiberman (2016), and many others.
4See footnote 2 for citations.5Autor, Dorn, Hanson and Song (2014) discuss this point in their study of the effects of Chinese export growth
across U.S. industries.
4
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
to take advantage of this variation.6
In the late 1980s and early 1990s, Brazil ended nearly one hundred years of extremely high
trade barriers imposed as part of an import substituting industrialization policy.7 In 1987, nominal
tariffs were high, but the degree of protection actually experienced by a given industry often
deviated substantially from the nominal tariff rate due to i) a variety of non-tariff barriers such
as suspended import licenses for many goods and ii) a system of “special customs regimes” that
lowered or removed tariffs for many transactions (Kume, Piani and de Souza 2003).8 In 1988 and
1989, in an effort to increase transparency in trade policy, the government reduced tariff redundancy
by cutting nominal tariffs and eliminating certain special regimes and trade-related taxes, but there
was no effect on the level of protection faced by Brazilian producers (Kume 1990).
Liberalization effectively began in March 1990, when the newly elected administration of Presi-
dent Collor suddenly and unexpectedly abolished the list of suspended import licenses and removed
nearly all of the remaining special customs regimes (Kume et al. 2003). These policies were replaced
by a set of import tariffs providing the same protective structure, as measured by the gap between
prices internal and external to Brazil, in a process known as tariffication (tarificacao) (de Carvalho,
Jr. 1992). In some industries, this process required modest tariff increases to account for the lost
protection from abolishing import bans.9 Although these changes did not substantially affect the
protective structure, they left tariffs as the main instrument of trade policy, such that tariff levels
in 1990 and later provide an accurate measure of protection.
The main phase of trade liberalization occurred between 1990 and 1995, with a gradual reduction
in import tariffs culminating with the introduction of Mercosur. Tariffs fell from an average of 30.5
percent to 12.8 percent, and remained relatively stable thereafter.10 Along with this large average
decline came substantial heterogeneity in tariff cuts across industries, with some industries such as
agriculture and mining facing small tariff changes, and others such as apparel and rubber facing
declines of more than 30 percentage points. We measure liberalization using long-differences in the
log of one plus the tariff rate from 1990 to 1995, shown in Figure 1. During this time period, tariffs
accurately measure the degree of protection faced by Brazilian producers, and tariff changes from
6Examples include Arbache, Dickerson and Green (2004), Goldberg and Pavcnik (2003), Gonzaga, Filho andTerra (2006), Kovak (2013), Krishna, Poole and Senses (2014), Menezes-Filho and Muendler (2011), Pavcnik, Blom,Goldberg and Schady (2004), Paz (2014), Schor (2004), and Soares and Hirata (2016) among many others.
7Although Brazil was a founding signatory of the General Agreement on Tariffs and Trade (GATT) in 1947, itmaintained high trade barriers through an exemption in Article XVIII Section B, granted to developing countriesfacing balance of payments problems (Abreu 2004). Hence, trade policy changes during the period under study wereunilateral.
8These policies were imposed quite extensively. In January 1987, 38 percent of individual tariff lines were subjectto suspended import licenses, which effectively banned imports of the goods in question (Authors’ calculations fromBulletin International des Douanes no.6 v.11 supplement 2). In 1987, 74 percent of imports were subject to a specialcustoms regime (de Carvalho, Jr. 1992).
9Appendix Figure A1 shows the time series of tariffs. Note the tariff increases in 1990 for the auto and electronicequipment industries.
10Simple averages of tariff rates across Nıvel 50 industries, as reported in Kume et al. (2003). See Appendix A.1for details on tariff data.
5
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
1990 to 1995 reflect the full extent of liberalization faced by each industry. We do not rely on the
timing of tariff cuts between 1990 and 1995, because this timing was chosen to maintain support
for the liberalization plan, cutting tariffs on intermediate inputs earlier and consumer goods later
(Kume et al. 2003).
As discussed below, along with regional differences in industry mix, the cross-industry variation
in tariff cuts provides the identifying variation in our analysis. Following the argument in Goldberg
and Pavcnik (2005), we note that the tariff cuts were nearly perfectly correlated with the pre-
liberalization tariff levels (correlation coefficient = -0.90). These initial tariff levels reflected a
protective structure initially imposed in 1957 (Kume et al. 2003), decades before liberalization. This
feature left little scope for political economy concerns that might otherwise have driven systematic
endogeneity of tariff cuts to counterfactual industry performance.
To check for any remaining spurious correlation between tariff cuts and other steadily evolv-
ing industry factors, we regress pre-liberalization (1980-1991) changes in industry employment and
wage premia on the 1990-1995 tariff reductions, with detailed results reported in Appendix B.1.
We attempted a variety of alternative specifications and emphasize that the results should be in-
terpreted with care, as they include only 20 tradable industry observations. Most specifications
exhibit no statistically significant relationship, but heteroskedasticity-weighted specifications place
heavy weight on agriculture and find a positive relationship. Agriculture was initially the least pro-
tected industry, and it experienced approximately no tariff reduction. It also had declining wages
and employment before liberalization, driving the positive relationship with tariff reductions. Con-
sistent with earlier work, when omitting agriculture, tariff cuts are unrelated to pre-liberalization
earnings trends (Krishna, Poole and Senses 2011). Given these varying results, we include controls
for pre-liberalization outcome trends in all of the analyses presented below to account for any po-
tential spurious correlation. Consistent with the notion that the tariff changes were exogenous in
practice, these pre-trend controls have little influence on the vast majority of our results.
3 Data and Empirical Approach
3.1 Data
Our main data source for regional labor market outcomes is the Relacao Anual de Informacoes
Sociais (RAIS), spanning the period from 1986 to 2010. This is an administrative dataset assem-
bled yearly by the Brazilian Ministry of Labor, providing a high quality census of the Brazilian
formal labor market (De Negri, de Castro, de Souza and Arbache 2001, Saboia and Tolipan 1985).
Accurate information in RAIS is required for workers to receive payments from several government
benefits programs, and firms face fines for failure to report, so both agents have an incentive to
provide accurate information. RAIS includes nearly all formally employed workers, meaning those
with a signed work card providing them access to the benefits and labor protections afforded by
6
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
the legal employment system. It omits interns, domestic workers, and other minor employment
categories, along with those without signed work cards, including the self-employed (see Appendix
B.2 for summary statistics on the informal sector, and Dix-Carneiro and Kovak (2015b) for anal-
yses covering the informal labor market). These data have recently been used by Dix-Carneiro
(2014), Helpman, Itskhoki, Muendler and Redding (forthcoming), Krishna et al. (2014), Lopes de
Melo (2013), and Menezes-Filho and Muendler (2011), though these papers utilize shorter panels.
The data consist of job records including worker and establishment identifiers, allowing us to track
workers and establishments over time. We utilize the establishment’s geographic location (munici-
pality) and industry, and worker-level information including gender, age, education (9 categories),
and December earnings.11
These data have various advantages relative to previous work on the effects of trade on local
labor markets. First, relative to Kovak (2013) and Autor et al. (2013), we can analyze the dynamics
of adjustment to the trade liberalization shock, as RAIS data are available every year. Second, RAIS
is a census rather than a sample, so it is representative at fine geographic levels.12 Third, a rich set
of labor market outcomes can be analyzed with such data, including how liberalization affected job
creation and job destruction rates, the number of active establishments, and the establishment size
distribution. Fourth, the ability to follow workers over time allows us to control for both observable
and unobservable worker characteristics.
As is typically the case in administrative employment datasets, the limitation of RAIS is a lack
of information on workers who are not formally employed, making it impossible to tell whether
a worker is out of the labor force, unemployed, informally employed, or self-employed. This is
important in the Brazilian context, with informality rates often exceeding 50 percent of all employed
workers during our sample period.13 When we need information on individuals who are not formally
employed, or information before 1986, we supplement the analysis using the decennial Brazilian
Demographic Census, covering 1970-2010. While these data provide much smaller samples and do
not permit following individuals over time, they cover the entire population, including the informally
employed, unemployed, and those outside the labor force.14 When possible, we corroborate results
from RAIS using the Demographic Census, finding very similar results across datasets.
Throughout the analysis, we limit our sample to include working-age individuals, aged 18-64.
When studying employed individuals, we omit those working in public administration and those
11RAIS reports earnings for December and for the entire year. We use December earnings to ensure that our resultsare not influenced by seasonal variation or month-to-month inflation. See Appendix Section A.2 for more detail onthe RAIS database.
12The National Household Survey (Pesquisa Nacional por Amostra de Domicılios - PNAD) would be a naturalalternative data source for a yearly analysis, but it only provides geographic information at the state level, does notallow one to follow individual workers over time, and provides a much smaller sample.
13Authors’ calculations using Brazilian Demographic Census.14See Appendix A.3 for more detail on the Demographic Census data and Dix-Carneiro and Kovak (2015b) for
analyses covering the informal labor market.
7
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
without valid information on industry of employment.15 Appendix A provides additional detail,
including descriptions of auxiliary data sources and variable construction.
To analyze outcomes by local labor market, we must define the boundaries of each market. We
use the “microregion” definition of the Brazilian Statistical Agency (IBGE), which groups together
economically integrated contiguous municipalities (counties) with similar geographic and productive
characteristics (IBGE 2002), closely paralleling an intuitive notion of a local labor market. When
necessary, we combine microregions whose boundaries changed during our sample period to ensure
that we consistently define local labor markets over time. This process leads to a set of 475
consistently identifiable local labor markets for analyses falling within 1986-2010 and 405 markets
for analyses using data from 1980 and earlier.16
3.2 Empirical Approach
Our empirical analysis follows the literature on the regional effects of trade by comparing the evo-
lution of labor market outcomes in regions facing large tariff declines to those in regions facing
smaller tariff declines. Intuitively, regions experience larger declines in labor demand when their
most important industries face larger liberalization-induced price declines (Topalova 2007). Kovak
(2013) presents a specific-factors model of regional economies that captures this intuition (a gen-
eralization of this setup appears below in Section 5.4.1). In this model, the regional labor demand
shock resulting from liberalization is
∑i
βriPi, where βri ≡λri
1ϕi∑
j λrj1ϕj
, (1)
hats represent proportional changes, r indexes regions, i indexes industries, ϕi is the cost share of
non-labor factors, and λri is the share of regional labor initially allocated to tradable industry i. Pi
is the liberalization-induced price change facing industry i, and (1) is a weighted average of these
price changes across tradable industries, with more weight on industries capturing larger shares of
15We exclude public administration because the labor market in this field operates quite differently from the restof the market. This choice has no substantive effect on any of our results.
16This geographic classification is a slightly aggregated version of the one in Kovak (2013), accounting for additionalboundary changes during the longer sample period. Related papers define local markets based on commuting patterns(e.g. Autor et al. (2013)). Our local market definition performs well based on this standard as well – only 3.4 and4.6 percent of individuals lived and worked in different markets in 2000 and 2010, respectively. The main regionaldefinition is shown in Figure 2. The analysis omits 11 microregions, shown with a cross-hatched pattern the figure.These include i) Manaus, which was part of a Free Trade Area and hence not subject to tariff cuts during liberalization;ii) the microregions that constitute the state of Tocantins, which was created in 1988 and hence not consistentlyidentifiable throughout our sample period; and iii) a few other municipalities that are omitted from RAIS in the1980s. The inclusion or exclusion of these regions when possible has no substantive effect on the results. We alsoimplemented the main analyses using a more aggregate local labor market definition, “mesoregions” defined by IBGE,and results are nearly identical.
8
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
initial regional employment.17 Thus, although all regions face the same vector of liberalization-
induced price changes, differences in the regional industry mix generate regional variation in labor
demand shocks.
We operationalize this shock measure by defining the “regional tariff reduction” (RTR), which
utilizes only liberalization-induced variation in prices, replacing Pi with the change in log of one
plus the tariff rate.
RTRr = −∑i
βrid ln(1 + τi) (2)
τi is the tariff rate in industry i, and d represents the long difference from 1990-1995, the period of
Brazilian trade liberalization. We calculate tariff changes using data from Kume et al. (2003), λri
using the 1991 Census, and ϕi using 1990 National Accounts data from IBGE.18 Together, these
allow us to calculate the weights, βri. Note that RTRr is more positive in regions facing larger
tariff reductions, which simplifies the interpretation of our results, since nearly all regions faced
tariff declines during liberalization.
Figure 2 maps the spatial variation in RTRr. Regions facing larger tariff reductions are pre-
sented as lighter and yellower, while regions facing smaller cuts are shown as darker and bluer.
The region at the 10th percentile faced a tariff reduction of 0.2 percentage points, while the region
at the 90th percentile faced a 10.7 percentage point decline. Hence, in interpreting the regression
estimates below, we compare regions whose values of RTRr differ by 10 percentage points, closely
approximating the 90-10 gap of 10.5 percentage points. Note that there is substantial variation
in the tariff shocks even among local labor markets within the same state. As we include state
fixed effects in our analyses, these within-state differences provide the identifying variation in our
study.19
We use the following specification to compare the evolution of labor market outcomes in regions
facing large tariff reductions to those in regions facing smaller tariff declines. For each year t ∈[1992, 2010], we estimate an equation of the following form:
yrt − yr,1991 = θtRTRr + αst + γt(yr,1990 − yr,1986) + εrt, (3)
where yrt is the value of a regional outcome such as earnings or employment, θt is the effect of
17Following Kovak (2013), we drop the nontradable sector, based on the assumption that nontradable prices movewith tradable prices. We confirm this assumption by calculating a measure of local nontradables prices in Section4.1.
18See Appendix A.4 for more detail on the construction of (2). We use the Census to calculate λri because theCensus allows for a less aggregate industry definition than what is available in RAIS (see Appendix A.1) and becausethe Census allows us to calculate weights that are representative of overall employment, rather than just formalemployment. That said, shocks using formal employment weights yield very similar results (Appendix Table B5,Panel D).
19A regression of RTRr on state fixed effects yields an R2 of 0.36; i.e. 64% of the variation in RTRr is not explainedby state effects. Our main conclusions are unaffected by the inclusion or exclusion of state fixed effects.
9
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
liberalization on outcomes by year t, αst are state fixed effects, and (yr,1990 − yr,1986) is a pre-
liberalization trend in the outcome variable. While the change in outcome varies with the year
t under consideration, the liberalization shock, RTRr, does not. Instead, it always reflects the
regional measure of tariff reductions during liberalization, from 1990 to 1995.20 Using this strat-
egy, each year’s θt represents one point on the empirical impulse response function describing the
cumulative local effects of liberalization as of each post-liberalization year. This methodology cap-
tures only relative effects across regions, as does the rest of the literature examining the regional
or sectoral effects of trade.
We use 1991 as the base year for outcome changes, and include state fixed effects to account for
any state-specific policies that might commonly affect outcomes for all regions in the same state,
such as state-specific minimum wages, introduced in 2002 (Neri and Moura 2006).21 We control for
pre-liberalization changes in outcomes (yr,1990 − yr,1986) to address the possibility of confounding
pre-existing trends, and consider longer pre-liberalization trends as a robustness test. For our main
outcomes, we present results with and without state fixed effects and pre-trends, with little effect
on the coefficients of interest. Since many of our dependent variables are themselves estimates, we
weight regressions based on the inverse of their standard error to account for heteroskedasticity.
We also cluster standard errors at the mesoregion level to account for potential spatial correlation
in outcomes across neighboring regions.
To consistently estimate θt, εrt must be uncorrelated with RTRr, conditional on the state fixed
effects and outcome pre-trend. For this identification assumption to be violated, there would need
to be an omitted variable that i) drives wage or employment growth across regions within a state
and ii) is correlated with RTRr but iii) is not captured by pre-liberalization outcome trends. While
such a feature is unlikely to exist, in Section 4.2 we confirm that our results are robust to a wide
variety of potential confounders and alternative specifications.
Our empirical approach is similar to prior studies examining the local effects of trade liberal-
ization, but we make two important contributions to that literature. First, the RAIS data allow us
to calculate changes in regional outcomes in each year following liberalization. We trace out the
dynamic regional response to liberalization as it evolves over time, rather than observing liberal-
ization’s local effect in only one post-shock period, as in the prior literature (e.g. Topalova (2007),
Autor et al. (2013), or Kovak (2013)). The RAIS data also allow us to control for pre-liberalization
trends that might otherwise confound the analysis. Second, we study a discrete, well-defined trade
policy shock that was complete by 1995. This contrasts with Autor et al. (2014), who use U.S.
panel data to study the effects of growing trade with China. They emphasize that the continuously
evolving nature of Chinese trade confounds their ability to study the dynamic response to a trade
20Recall from Section 2 that tariffs declined between 1990 and 1995, after which they remained relatively stable.21Using 1991 as the base year allows us to take advantage of more detailed industry information in the 1991 Census
when calculating the industry distribution of regional employment (λri), and makes our results comparable withKovak (2013).
10
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
shock at any given point in time.
4 Results
4.1 Main Findings
We begin by examining the effects of liberalization on formal sector earnings and employment
in local labor markets. First, we calculate “regional earnings premia,” which reflect average log
monthly earnings for workers in a given region, controlling for the composition of the regional
workforce. For each year t, we regress log December earnings for worker j on flexible controls for
age, sex, and education (Xjt); industry fixed effects (φit); and region fixed effects (µrt).22
ln(earnjrit) = XjtΓt + φit + µrt + ejrit (4)
The region fixed effect estimates from this regression, µrt, represent the regional log earnings premia
for the relevant year. By estimating these regressions separately in each year, we allow for changes
in the regional composition of workers (X) and changes in the returns to worker characteristics
(Γ) over time.23 This approach ensures that our earnings estimates are not driven by changes in
observable worker composition, changing discrimination, changes in the returns to schooling, or any
other changes in the returns to observable characteristics that operate at the national level. Our
dependent variable when studying earnings is then the change in regional log earnings premium
from 1991 to each subsequent year, 1992 to 2010.
Table 1 shows the results of estimating (3) for regional formal sector log earnings premia and
formal log employment. All estimates for the coefficient on RTRr are negative, indicating that re-
gions facing larger tariff reductions experience relative declines in earnings or employment. Consider
Panel A, which presents liberalization’s effect on regional earnings. Columns (1) to (3) examine
changes in earnings from 1991 to 2000, while columns (4) to (6) examine changes from 1991 to
2010, such that the effects cumulate over time. Columns (2) and (4) add state fixed effects, and
columns (3) and (6) add pre-trend controls for the change in the regional outcome from 1986 to
1990. The coefficient estimate of -0.529 in column (3) indicates that a region facing a 10 percentage
point larger tariff reduction (approximately the 90-10 gap in RTRr) experienced a 5.29 percentage
point larger proportional decline (or smaller increase) in formal earnings from 1991 to 2000. This
magnitude is similar in size to the corresponding estimate in Kovak (2013) (-0.439), which used a
different data source (Census of Population) and covers all workers rather than restricting attention
22We use monthly earnings rather than hourly wages because RAIS only provides hours from 1994 onward. Censusresults using hourly wages are similar.
23Appendix B.3 presents the coefficient estimates from (4) for 1991, 2000, and 2010. In Section 5.2, we control forobservable and unobservable worker heterogeneity by pooling across years and including individual fixed effects. Theresults are very similar.
11
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
to the formally employed. The estimate of -1.594 in column (6) indicates that the gap in earnings
growth expanded to 15.94 percentage points by 2010.
This increase in liberalization’s effect on earnings from 2000 to 2010 is a striking feature of
Table 1. It indicates that the divergent earnings growth in regions facing different tariff reductions
continued well beyond the liberalization period. Figure 3 confirms this pattern by plotting the
coefficients on RTRr (θt) for each year. The points for 2000 and 2010 correspond to the RTRr
coefficients in columns (3) and (6) of Table 1. The vertical lines indicate that liberalization began in
1991 and was complete by 1995. We present coefficient estimates for 1992-94, but these should be
interpreted with care, as liberalization was still ongoing.24 The local earnings effects of liberalization
appear just after liberalization and steadily grow for more than a decade, before leveling off in the
late 2000s, a pattern that is very robust to details of the specification.25 Figure 3 also shows
pre-liberalization coefficients, in which the dependent variable is the change in regional earnings
premium from 1986 to the year listed on the x-axis, and the independent variable is RTRr. If
anything, the relative earnings declines in regions facing larger tariff reductions represent a reversal
of the pre-liberalization trend. Recall that all post-liberalization results control for pre-liberalization
trends, as shown in (3).
It is likely that the prices of local nontradable goods change in response to the regional shocks
to the prices of traded goods (Kovak 2013, Monte 2016). If this is the case, the relative decline in
nominal earnings in regions facing larger tariff reductions may be partly offset by declines in the
local price index. To empirically evaluate this possibility, we construct local price indexes using
housing rents information in the Census, following the approach of Moretti (2013).26 Only the 1991
and 2010 Censuses included rent questions, so we can only calculate the change in rental prices
for 1991-2010. Our local price index uses consumption weights from the Brazilian Consumer Price
Index system (IPC) and accounts for the fact that the prices of non-housing nontradables tend
to move with housing prices. See Appendix A.5 for details on constructing the index. We then
calculate the change in log real earnings as the change in log nominal earnings minus the change in
log local price level. Panel B of Table 1 shows the effect of regional tariff reductions on the change
in real regional earnings for 1991-2010. As expected, the effect on real earnings in column (6) is
smaller than the effect on nominal earnings by about 21 percent. This difference confirms that
regional nontradable prices move with tradable prices, falling more in places facing larger tariff
reductions. However, the long-run effects of liberalization on real regional earnings are still large
and statistically significant.
24However, the tariff cuts were almost fully implemented by 1993, so these early coefficients are still informativeregarding liberalization’s short-run effects. When regressing RTRr on an alternate version measuring tariff changesfrom 1990-93, the R2 is 0.93.
25See Section 4.2 for a variety of robustness tests. Appendix B.3 shows the underlying scatterplots, confirming ourchoice of linear estimating equation and showing that the results are not driven by outliers. Appendix Table B.5shows that the same pattern appears when estimating formal earnings or formal hourly wages using Census data.
26As in the U.S., the Brazilian government does not produce local price indexes outside a few large cities.
12
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table 1 Panel C and Figure 4 both examine liberalization’s effects on regional log formal em-
ployment. The year 2000 estimate of -3.533 shows that a region facing a 10 percentage point larger
tariff reduction experienced a 35.3 percentage point larger proportional decline (smaller increase)
in formal employment from 1991 to 2000. As with earnings, the employment effect grew substan-
tially from 2000 to 2010, indicating that employment growth continued to diverge for regions facing
different regional tariff changes. Most of this divergence was complete by 2004, after which the
estimates level off.27
Note that since Table 1 and Figure 4 examine formal employment, there are two channels
through which formal employment might decline in regions facing more negative shocks. Formally
employed workers may migrate away from negatively affected places to more favorably affected
places, or existing residents of the region may shift into or out of formal employment. Table 2 rules
out the interregional migration mechanism, showing that a region’s working-age population did not
respond to RTRr.28 We measure working-age population using Census data, so we can observe
individuals outside formal employment, and control for 1980-1991 outcome pre-trends. These pre-
trends may be subject to mechanical endogeneity, since both the pre-trends and the dependent
variables include the 1991 outcome level, so we instrument for the 1980-1991 pre-trends using the
1980 dependent variable level (columns (2) and (5)) or the 1970-1980 change (columns (3) and (6)).
All of the population estimates are insignificantly different from zero, and the IV point estimates
are quantitatively small, indicating that workers losing formal employment in harder hit regions
did not leave the region, but transitioned out of formal employment and into informal employment
or non-employment.29
Together, these results are quite surprising, particularly compared to the conventional wisdom
from the literature studying local labor demand shocks. The standard framework predicts initially
large wage effects of local labor demand shocks, as labor supply is approximately fixed in the
very short run, after which employment adjustment arbitrages away spatial wage differences, and
observed wage effects fall in magnitude (Blanchard and Katz 1992, Bound and Holzer 2000). This
mechanism is consistent with the steadily growing employment effects in Figure 4, but is at odds
with the growing earnings effects in Figure 3. It predicts large negative coefficients shortly after
liberalization, but then declining magnitude effects as arbitrage partly equalizes earnings growth
across regions. Even in the absence of equalizing migration, as shown in Table 2, one would expect
27To assess the scale of our long-run estimates, consider Dix-Carneiro (2014), which studies a very similar settingwith slow adjustment of labor across Brazilian industries rather than regions. After estimating the model’s parametersusing RAIS data, he simulates the economy’s response to a price shock when capital is mobile across industries (seehis Figures 4 and 6). The long-run wage elasticity in the adversely affected sector (High-Tech Manufacturing) is -1.56.This is exceedingly close to our 2010 earnings estimate of -1.594. The long-run employment elasticity in Dix-Carneiro(2014) is -3.2. Although this is somewhat smaller than our 2010 employment estimate of -4.663, the two effects aresimilar in magnitude, suggesting that our findings are reasonable in the context of this type of model.
28Similarly, Autor et al. (2013) find little evidence for population responses to trade shocks in the U.S.29In Dix-Carneiro and Kovak (2015b), we further document these shifts from formal employment into informal
employment and non-employment using Census data.
13
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
constant effects over time. Instead, we find continuing divergence in earnings growth for 14 years
following the end of liberalization, with earnings growth in regions facing larger tariff reductions
lagging further and further behind other regions. This pattern means that the local labor market
effects of trade estimated in prior work for a single post-liberalization year actually understate the
longer run effects. The remainder of the paper focuses on examining and explaining this surprising
result.
4.2 Robustness
We first establish that the steadily growing earnings effects are robust to alternative measurement
and specification choices and that they were not driven by confounding effects from other shocks
to Brazilian local labor markets. Detailed analyses appear in Appendix Sections B.6–B.8, and we
summarize the results here.
Appendix B.6 shows that the growing earnings estimates are robust to alternative pre-trend
controls, RTRr shock measures, earnings premium measures, and weighting. We use Census data to
construct longer pre-liberalization earnings trends, from 1970-1980 and from 1980-1991, and control
for these alongside the 1986-1990 RAIS pre-liberalization trends present in our main specification.30
We construct alternative RTRr measures, i) using industry weights, λri, reflecting only formal
employment, ii) using effective rates of protection, which account for the effects of tariffs on inputs
and outputs for each industry, and iii) including a zero price change for the nontradable sector. We
also construct alternative earnings premium measures. The first is calculated without controlling
for industry fixed effects, maintaining national industry-level earnings variation in the regional
earnings premia.31 The second measure simply uses mean log earnings, without controlling for
any worker characteristics. Finally, we present results weighting regions equally or weighting by
the region’s 1991 formal employment. In all cases, our main results are confirmed, finding steady
growth in liberalization’s effects on regional earnings. The employment effects are similarly robust
to these alternatives.
Although our findings are robust to these specification and measurement changes, the effects
of liberalization could appear to grow over time because of correlated shocks occurring after trade
liberalization. To explain the smooth growth of the effects in Figures 3 and 4, such confounders
would need to affect industries or regions similarly to liberalization and would need to grow steadily
over time or occur quite regularly. Although these circumstances are unlikely, in Appendix B.7 we
30Because 1991 is the base year for our post-liberalization earnings growth outcome, 1980-1991 pre-liberalizationtrends are subject to mechanical endogeneity. We resolve this problem by calculating an alternative earnings growthmeasure with 1992 as the base year. See Panel C of Table B6.
31By omitting the industry fixed effects, these regional earnings measures include both direct industry effects andlocal general equilibrium effects. As shown in Appendix B.6, the associated estimates are only a bit larger than themain results, indicating that local equilibrium effects account for the majority of the overall effects of liberalizationon regional earnings.
14
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
construct controls for a wide variety of salient economic shocks in the post-liberalization period,
demonstrating that they cannot account for the growing earnings effects.
If tariff changes after 1995 were correlated with those occurring during liberalization (1990-
95), they might drive the apparently increasing effects of liberalization, although this is unlikely
since post-1995 tariff changes were very modest. We calculate post-liberalization regional tariff
reductions as in (2), but use tariff reductions between 1995 and year t > 1995, and include these
post-liberalization tariff reductions as additional controls alongside RTRr. Other potential con-
founders are the Brazilian Real devaluations that occurred in 1999 and 2002. If these exchange rate
movements affected industries differently, they might have been correlated with tariff changes dur-
ing liberalization. We construct industry-specific real exchange rates as import- or export-weighted
averages of real exchange rates between Brazil and its trading partners. We then take the change
in log real exchange rate from 1990 to year t, and calculate regional shocks using weighted averages
as in (2). There was also a substantial wave of privatization during our sample period. We address
privatization by controlling flexibly for the 1995 share of regional employment at state-owned firms
or the change in this share from 1995 to t. Controlling for each of these post-liberalization shocks
has little effect on the earnings results, which continue to exhibit substantial post-liberalization
growth in all cases.
The global commodity price boom of the late 2000s is another potential post-liberalization
confounder that might explain our growing earnings results, particularly since agricultural products
faced the most positive tariff change during liberalization. In Appendix B.7.4, we provide extensive
evidence ruling out this possibility. First, the timing of the commodity price boom does not
correspond to the timing of our effects. Commodity prices were flat or declining between 1991 and
2003, during which our earnings and employment results grew substantially. Commodity prices
then grew sharply after 2004, when our results began to level off.32 We show that the substantial
growth in earnings effects remains when i) dropping regions most exposed to commodity price
growth, by restricting the region sample to include only those with below-median or bottom-
quartile employment shares in agriculture and mining or ii) when restricting our regional earnings
measure only to workers in manufacturing. Finally, we use two approaches to directly control for
the regional effects of the commodity price boom. We control for the share of workers in agriculture
and mining and for changes in regional commodity prices using the measure introduced by Adao
(2015). We also control for China’s effects on commodity markets using the import and export
quantity measures and instruments from Costa et al. (2016).33 Both sets of controls have little
influence on the observed earnings effects of RTRr.
32A similar argument applies to Bustos, Caprettini and Ponticelli (2016), who study the effects of geneticallymodified crops in Brazil. Genetically modified crops were outlawed before 2003 and only permanently authorized in2005, so this channel cannot explain the substantial growth in earnings effects before 2005.
33Special thanks to Rodrigo Adao for providing commodity price data and code, and to Francisco Costa forproviding the shock and instrument measures from Costa et al. (2016).
15
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
As a final set of robustness tests, Appendix B.8 presents results when splitting the sample
by tradable and nontradable sector and by skill. We find growing earnings effects in all of these
subsamples. This pattern is particularly noteworthy for the nontradable sector, as it confirms that
regional labor market equilibrium transmits the effects of liberalization from the tradable sector
to the nontradable sector, as predicted in the model of Kovak (2013), which is the basis for the
RTRr shock. The earnings effects for more skilled workers are a bit larger than those for less skilled
workers, while the employment effects are larger for less skilled workers. However, these results
should be interpreted with care, as the RTRr shocks are derived from a model with a single type
of labor.34
Together, the results in this section demonstrate the robustness of our main findings to alterna-
tive measures and estimation approaches and rule out a wide variety of salient post-liberalization
shocks as potential confounders. We conclude that the earnings and employment profiles shown in
Figures 3 and 4 reflect growing causal effects of liberalization over time. In the next section, we
consider a variety of potential mechanisms that could drive this growth in liberalization’s effects
on local labor market outcomes.
5 Mechanisms
As mentioned above, the conventional model of local labor markets predicts large effects of liberal-
ization just after the tariff change and smaller effects as labor reallocation arbitrages away spatial
differences in earnings growth. Our findings contradict this prediction, instead exhibiting increas-
ing differences in earnings growth for 15 years after liberalization between regions facing larger
and smaller tariff reductions. In this section, we consider a variety of potential mechanisms that
might explain these growing earnings effects, finding strong empirical support for mechanisms in-
volving imperfect interregional labor mobility and dynamic labor demand, particularly slow capital
adjustment and agglomeration economies.
5.1 Urban Decline
Glaeser and Gyourko (2005) and Notowidigdo (2013) present models of urban decline in which the
slow depreciation of housing stocks drives slow adjustment in local labor markets facing permanent
negative labor demand shocks. In their models, the price of housing falls sharply in depressed
markets, incentivizing individuals to remain in the city in spite of nominal earnings losses follow-
ing the demand decline. As housing slowly depreciates, this incentive dissipates, so population
and therefore employment steadily decline. This mechanism could therefore rationalize the slowly
growing employment effects we document in Figure 4.
34For a more general model with two skill types, see Dix-Carneiro and Kovak (2015a).
16
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
However, as in the conventional model of local labor markets, this mechanism predicts the
opposite of what we find for earnings in Figure 3. Although wages fall on impact in regions
facing negative shocks, they recover slowly over time as workers leave the market due to housing
depreciation.35 Moreover, the mechanism depends on declining population in cities facing negative
shocks. In Brazil, overall population growth was large enough during our sample period that out of
475 local labor markets, only 11 experienced population decline between 1991 and 2000, and only 6
did so between 1991 and 2010.36 Table 2 also finds no response of local working-age population to
RTRr. Thus, while the slow housing depreciation mechanism is quite relevant for rust-belt cities
in the U.S., it does not appear to apply in the Brazilian context.
5.2 Changing Composition of Worker Unobservables
Liberalization might cause earnings to slowly decline in regions facing larger tariff reductions rela-
tive to other regions because of worker selection. Higher-earning workers may be more likely to leave
the formal labor market in harder-hit regions, and this selective worker reallocation may increase
over time. Although we flexibly control for detailed worker characteristics including age, sex, and
education when calculating regional earnings premia in our main specifications, worker composition
may also adjust along unobservable dimensions. To examine this possibility, we calculate alternative
earnings premia, pooling the RAIS data across years and controlling for worker-level fixed effects,
which capture time-invariant worker characteristics including unobservables. We implement this
procedure in two ways (See Appendix B.9 for details). First, we restrict the coefficients on the indi-
vidual fixed effects to be constant over time, assuming that the returns to worker characteristics do
not change. Panel B of Table 3 presents results using these regional earnings premia. Second, we
allow the coefficients on the individual fixed effects to vary across years, allowing for changes in the
returns to worker characteristics over time. Given that the returns to observable worker character-
istics change substantially over time (see Appendix B.3), this is an important generalization. Panel
C of Table 3 presents results using these earnings premia. In both cases, the growth in earnings
effects remains, and the results from the more flexible earnings premium specification in Panel C
are quantitatively very close to the results from the main specification in Panel A. These findings
rule out worker selection as a mechanism driving the observed growth in the earnings effects of
liberalization.
35Note that Glaeser and Gyourko (2005) do not model a production side and instead directly shock wages oramenities. However, a simple extension of their model to include labor market equilibrium would have the featurescited here, as in Notowidigdo (2013). Glaeser and Gyourko (2005) also argue that local average wages will decline overtime in negatively shocked markets because the most productive workers have the strongest incentive to leave. Asshown in Section 5.2, since we control for worker characteristics when calculating regional earnings premia, selectioneffects of this kind are not driving our results.
36Authors’ calculations using Census data.
17
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
5.3 Slow Response of Imports or Exports
Although trade liberalization was complete by 1995, it is possible that trade quantities were slow to
respond to the sharp change in trade policy, perhaps because of difficulty in forming new trade links
with firms abroad. Prices faced by Brazilian producers may evolve slowly in response to tariff cuts
if import quantities respond slowly to liberalization. If so, the slow evolution of imports in response
to the tariff cuts could potentially explain the slow growth in the effects of liberalization over time.
To address this possibility, we follow Autor et al. (2013) by constructing changes in imports and
exports per worker for each industry from 1991 to each subsequent year, using Comtrade data.37
We then form regional weighted averages of these changes in trade flows, weighting by the industry’s
initial share of regional employment. See Appendix A.6 for details on the construction of these
measures.
We first examine the effect of regional tariff reductions on these regional measures of import,
export, and net export growth, looking for evidence of slow growth in trade quantities that might
drive the slow growth in earnings effects. We do so using the trade growth measures as dependent
variables in (3). Figure 5 plots the effects of RTRr on each trade flow measure.38 First, consider
the effects on regional imports (blue circles). As expected, regions facing larger tariff reductions
experienced larger increases in the regional import measure. These import increases occurred
immediately after liberalization, with large positive coefficients already present in 1995. Because
we measure trade flows in $100,000 units, the 1995 coefficient of 0.144 implies that a region facing
a 10 percentage point larger tariff reduction experienced a $1,440 larger increase in imports per
worker. These import effects actually decrease on average until 2003 (coefficient estimate = 0.070),
in sharp contrast to the earnings effects, which grew to more than two-thirds of their long-run level
during the same time period. After 2003, the import effects increase, but this coincides with a
leveling-off in the earnings and employment effects. This timimg is inconsistent with slow import
growth driving our results.
The sign of the export effects (red triangles) is positive, indicating that industries experiencing
larger export increases were on average located in the same regions as industries facing larger
tariff reductions.39 This effect works against the hypothesis that slow trade quantity growth drove
relative earnings declines in these regions. After 2003, both the import and export effects grow
quite substantially, following the overall trends in Brazilian imports and exports. Note, however,
that the relationship between RTRr and net exports (green diamonds) falls from 2005 to 2010,
again a time period with substantial growth in the earnings effects. Overall, the time paths of
37Appendix B.10 shows results for an ad-hoc alternative functional form using the change in log trade, yielding thesame conclusions.
38In Figure 5 we do not have pre-liberalization trends for trade flows because Comtrade data for Brazil begin in1989.
39The positive sign for the export effect is not driven by any particular industry or industries and is robust todropping agriculture and/or natural-resource industries.
18
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
regional imports and exports do not match those of regional earnings, suggesting that slow trade
quantity adjustment is not driving our results.
To confirm this point, we directly control for regional import and export growth when examining
the effect of RTRr on regional earnings premia. If the growing earnings effects remain in spite of
including these controls, we can be confident that a different mechanism is at play. We examine
the relationship between earnings growth and RTRr, as in (3), including controls for the regional
growth in imports (RegImprt) and exports (RegExprt) from 1990 to year t.
yrt − yr,1991 = θtRTRr + β1RegImprt + β2RegExprt + αst + γt(yr,1990 − yr,1986) + εrt (5)
The import and export coefficients, β1 and β2, are constant over time, allowing us to test whether
the slow evolution of trade flows explains the evolution of earnings growth (since RegImprt and
RegExprt change over time, unlike RTRr). Panel B of Table 4 shows that the effect of RTRr on
regional earnings still grows steadily over time when controlling for changes in regional imports and
exports, implying that slow trade quantity growth is not driving the effects.
A remaining concern is that if regional imports and exports are endogenous to regional earn-
ings growth, then the coefficients on RTRr will be biased along with the trade flow coefficients.
Panels C and D address this issue following the strategy of Autor et al. (2013), instrumenting for
Brazilian trade flows using trade flows for other countries.40 We consider instruments based on the
combination of Argentina, Chile, Colombia, Paraguay, Peru, and Uruguay (“Latin America”) and
on Colombia alone, which liberalized during the same time period as Brazil and imposed similar
tariff cuts across industries (Paz 2014). In each case, we measure imports and exports between
these countries and the rest of the world, excluding Brazil.41 Panels C and D show the results. In
both cases, the effects of RTRr continue to grow over time, with a similar magnitude to the main
results, shown in Panel A. These and the preceding results in this section rule out slow import or
export responses as the mechanism driving the slowly growing earnings effects.
5.4 Dynamic Labor Demand
A remaining potential mechanism driving the growing effects of liberalization on earnings and
employment involves dynamics in labor demand. If labor is imperfectly mobile across regions and
an initial labor demand shock is followed by a dynamic process that amplifies the shock’s effects
over time, one will observe the growing regional earnings and employment effects we document.
We consider two potential sources of these dynamics: agglomeration economies (e.g. Kline and
Moretti 2014) and slow adjustment of capital stocks (e.g. Dix-Carneiro 2014). As we will show,
both appear to play important roles in explaining our findings.
40We also include regional measures of commodity price growth from Adao (2015) in the set of instruments.41Due to Comtrade data availability, the changes in trade flows for Latin America are calculated from 1994 to each
subsequent year and those for Colombia alone are calculated from 1991 to each subsequent year.
19
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
5.4.1 Evidence on the Importance of Dynamic Labor Demand
To study these mechanisms and formalize our argument, we generalize the specific-factors model
in Kovak (2013) to include agglomeration economies and slow adjustment of labor and capital. We
focus on the formal economy, consisting of many regions, indexed by r, which may produce goods
in many industries, indexed by i. Production in each industry uses Cobb-Douglas technology with
constant returns to scale and three inputs: labor, a fixed factor, and capital. Formal labor, Lr, is
assumed to be perfectly mobile between industries within a region. The fixed factor, Tri, is usable
only in its respective region and industry and is fixed over time. This factor represents inputs such
as natural resources, land, or very slowly depreciating infrastructure and capital that are effectively
fixed over the time horizons we consider. Capital, Kri, is also usable only in its respective region
and industry but may change slowly over time through depreciation and investment decisions.42
Output of industry i in region r is
Yri = AriL1−ϕiri
(T ζiriK
1−ζiri
)ϕi(6)
where ϕi, ζi ∈ (0, 1). Goods and factor markets are perfectly competitive, and producers face
exogenous prices Pi, common across regions and fixed by world prices and tariffs. To allow for
the possibility of agglomeration economies, we allow productivity, Ari, to vary with the amount
of local economic activity. We also allow for factor adjustment by letting Lr and Kri change over
time. Recall that changes in Lr primarily reflect workers entering or leaving the formal workforce
rather than other channels such as interregional migration, as shown in Table 2. We assume that
changes in Kri reflect depreciation and firms’ investment decisions rather than physical mobility
via secondary markets for installed capital.
As shown in Appendix C, factor market clearing, zero profits, and cost minimization imply the
following equilibrium relationship, in which hats represent proportional changes.
wr =∑i
βriPi +∑i
βriAri − δr
(Lr −
∑i
λri(1− ζi)Kri
)(7)
where βri ≡λri
1ϕi∑
j λrj1ϕj
> 0 and δr ≡1∑
j λrj1ϕj
> 0.
wr is the proportional change in the regional wage, and λri is the initial share of regional employment
in industry i. This is an equilibrium relationship because the factor supplies and productivity levels
may respond endogenously to the liberalization shock reflected by Pi.
42We separate fixed factors and variable capital for two reasons. First, our research design is based on regionaldifferences in industry mix, which are driven by fixed factors. Second, including fixed factors in each region ensuresthat all regions maintain some economic activity even when faced with very negative shocks. Hence, this formulationis common in the literature on agglomeration economies (e.g. Helm (2016) and Kline and Moretti (2014)).
20
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
As a thought exercise, suppose we were to hold productivity and factor supplies constant (Ari =
Lr = Kri = 0). In that case, the wage change equals the simple weighted average price shock in
(1). In this restricted model, there is no scope for dynamic effects of liberalization, and one would
observe a substantial wage effect of liberalization on impact, with no changes thereafter. More
realistically, if productivity or factor supplies evolve over time in response to the liberalization-
induced price shocks, then the effects of liberalization on regional wages can change over time as
well.
First, we consider factor supply responses. Imagine that only regional labor supply responds
to liberalization, while maintaining Ari = Kri = 0. Immediately following liberalization, wages
decline more in regions facing larger tariff reductions, and formal employment falls more in these
regions, as in Figure 4. Equation (7) shows that this change in employment partly offsets the
wage losses experienced on impact, since δr > 0. If employment adjusts slowly, then the observed
wage effects of liberalization get smaller over time. In other words, with labor adjustment only,
the model reflects the conventional prediction that liberalization’s effects on local wages decline
over time. If we allow both regional employment and regional capital stocks to vary in response to
liberalization, complex patterns can emerge, depending on the relative speed of labor and capital
adjustment. For example, if regional labor is held fixed and capital stocks contract more in regions
facing larger tariff declines (as we show below), the marginal product of labor will fall, and relative
wages will decline even further in harder hit regions, as seen in Figure 3. More generally, the model
can qualitatively rationalize growing earnings effects of liberalization if the labor supply elasticity
is finite and capital adjusts more quickly than labor.
Now consider changes in productivity, Ari. We assume that these result from agglomeration
economies in which changes in the amount of local economic activity drive changes in the produc-
tivity of local firms. There is little agreement on the specific source of agglomeration economies,
with various papers arguing that they result from changes in population, overall employment,
or employment in particular industries (Melo, Graham and Noland 2009).43 For agglomeration
economies to be relevant in our context, we must observe effects of regional tariff reductions on
at least one of these agglomeration sources. In Table 2 and Appendix B.11, we show that neither
working-age population nor overall employment (sum of formal and informal) respond substan-
tially to RTRr, while Figure 4 shows that liberalization substantially affected formal employment.
For agglomeration economies to be relevant in our context, agglomeration must apply to regional
formal employment, since other potential sources of agglomeration do not significantly respond to
liberalization. This is plausible, as labor market pooling and knowledge spillovers are more likely to
apply in formal employment than in informal employment, which disproportionately includes agri-
cultural production. In this case, a negative labor demand shock decreases wages on impact, which
43Many papers argue that population or employment density is the relevant quantity, but since we utilize regionswith fixed boundaries, the change in log population or employment density is identical to the change in log populationor employment level.
21
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
endogenously decreases formal employment and therefore decreases regional productivity through
agglomeration economies. As shown in (7), this productivity decline amplifies the wage decline
from the initial shock, leading to further reductions in local formal employment and productivity,
etc. If this amplification occurs slowly over time, perhaps due to slow labor supply responses or
slow responses of productivity to formal employment (Kline and Moretti 2014), then the observed
effects of liberalization may also grow over time.
Therefore, given imperfect labor mobility across regions, both capital adjustment and agglom-
eration economies could qualitatively explain the earnings and employment patterns in Figures 3
and 4. To provide evidence for the relevance of dynamic labor demand, we rearrange (7) to infer the
labor demand shifts needed to rationalize the changes in earnings with the observed regional tariff
reductions and changes in formal employment. For consistency with the agglomeration literature,
we assume identical factor cost shares across industries (ϕi = ϕ ∀i and ζi = ζ ∀i, which implies
δr = ϕ).44 The economy-wide value of ϕ is 0.544 (see Appendix A.4), and we discuss the value of
ζ in Section 5.4.3. ∑i
βriAri + ϕ(1− ζ)∑i
λriKri = wr −∑i
βriPi + ϕLr︸ ︷︷ ︸observed
(8)
The left hand side of (8) captures the overall shifts in labor demand resulting from agglomeration
economies and capital adjustment, which we can measure as a residual using the observable quanti-
ties on the right hand side. We measure wr as the change in regional earnings premium, −∑
i βriPi
as RTRr, and Lr as the change in regional formal employment. Figure 6 (solid blue circles) shows
the relationship between this inferred labor demand measure and regional tariff reductions in each
year following the start of liberalization. We can infer that labor demand steadily declined in
regions facing larger tariff reductions and that these dynamics were complete by the late 2000s.
Given this evidence for dynamic labor demand in general, we examine evidence for the two specific
sources of dynamics: agglomeration economies and slow capital adjustment.
5.4.2 Evidence for Agglomeration Economies and Capital Adjustment
To examine these mechanisms in more detail, we follow the literature by imposing additional long-
run assumptions that allow us to compare our results to prior work and to quantify the roles of
agglomeration and slow capital adjustment. We assume a constant elasticity long-run agglomeration
function.45
Ari = κLr, κ ≥ 0 (9)
44When assuming identical factor cost shares across industries, our production function is identical to those in Klineand Moretti (2014) and Helm (2016). Hanlon and Miscio (2016) use a slightly different Cobb-Douglas productionfunction, but also assume constant cost shares across industries.
45Kline and Moretti (2014) provide empirical support for a constant agglomeration elasticity.
22
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table 2 shows that working-age population does not substantially respond to liberalization, indicat-
ing that the main margin of labor supply adjustment is workers’ choice of whether to pursue formal
employment within a given region. In earlier work, we showed that informal sector earnings do not
respond to liberalization (Dix-Carneiro and Kovak 2015b). Therefore, we assume that changes in
formal labor (Lr) depend upon changes in the regional formal wage (wr), and assume a constant
elasticity long-run local formal labor supply function.
Lr =1
ηwr, η ≥ 0 (10)
Finally, we assume perfectly mobile capital in the long run (Rr = R ∀r, where R is the price of
capital).46 We take 2010 to be the long run (20 years following the start of liberalization), consistent
with the flat earnings and employment responses by the late 2000s. Imposing these assumptions
on the model yields the following expressions for the long-run regional wage change and the change
in employment in a given region × industry combination (derived in Appendix C).
wr =η
η[1− ϕ(1− ζ)]− κ+ ϕζ
∑i
βriPi −ϕ(1− ζ)η
η[1− ϕ(1− ζ)]− κ+ ϕζR (11)
Lri =1
ϕζPi −
1
ϕζ· η[1− ϕ(1− ζ)]− κη[1− ϕ(1− ζ)]− κ+ ϕζ
∑i
βriPi −ϕ(1− ζ)
η[1− ϕ(1− ζ)]− κ+ ϕζR (12)
We test for the presence of agglomeration economies using the change in employment in each
region × industry combination, following an approach similar to that of Helm (2016). As shown in
(12), in the absence of agglomeration (κ = 0), holding fixed an industry’s own price decline, larger
regional tariff reductions increase local industry employment. Intuitively, if other industries in the
same region face larger tariff cuts, more laborers will locally transition into the reference industry
in equilibrium. However, in a setting with agglomeration economies (κ > 0), price reductions
in other industries in the same region reduce the local productivity of the reference industry. If
agglomeration forces are strong enough, larger regional tariff reductions can reduce local industry
employment conditional on the industry’s own price change. We therefore estimate the following
specification.
Lri = γ0 + γ1Pi + γ2RTRr + εri (13)
This expression is the reduced form of (12). γ0 captures the term for R, which does not vary
across industries or regions, and γ2 < 0 implies the presence of agglomeration economies.47 We
measure Lri using changes from 1991 to 2010 to capture long-run adjustment, and present the
46Perfect long-run capital mobility is a standard assumption in this literature (Hanlon and Miscio 2016, Helm 2016,Kline and Moretti 2014).
47Recall that RTRr ≡ −∑i βriPi, so γ2 < 0 implies η[1−ϕ(1−ζ)]−κ
η[1−ϕ(1−ζ)]−κ+ϕζ < 0 in (12), which in turn implies κ > 0.
23
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
results in Table 5. We control for industry price changes either directly using tariff changes, or
with industry fixed effects.48 In all cases, the coefficient on RTRr is negative and highly significant;
an industry’s local employment actually falls when other industries in the same region face larger
tariff reductions, implying the presence of agglomeration economies.
We also find evidence for slow capital adjustment. Although regional capital stock measures are
unavailable, we can observe changes in the number of formal establishments in a given region, which
are likely to approximate changes in regional capital stocks.49,50 Figure 7 shows that regions facing
larger tariff reductions experienced steady relative declines in the number of formal establishments,
with the effect increasing most quickly in the early 2000s and leveling out later in the sample period.
It is possible that capital simply reallocated from smaller exiting establishments to larger continuing
establishments in harder-hit locations. If this were the case, the change in number of establishments
would not be particularly informative about the change in regional capital stock. However, the
decline in the number of establishments was not offset by increases in the average size of remaining
establishments; if anything these establishments shrank on average. Moreover, Appendix B.13
shows that larger tariff declines drove increases in exit rates throughout the establishment size
distribution. These results strongly support the interpretation that trade shocks induced a gradual
reallocation of capital away from harder hit locations.
To reinforce this conclusion, we present evidence on the margins of capital adjustment. We
expect investment to respond immediately following liberalization, with new investment directed
toward more favorable markets and away from markets facing larger tariff reductions. In contrast,
depreciation takes time to erode the capital stock in a negatively affected region. We confirm these
patterns using measures of regional establishment entry and exit and job creation and destruc-
tion. We measure cumulative entry, exit, job creation, and job destruction by observing changes
from 1991 to each subsequent year, and calculate each measure following Davis and Haltiwanger
(1990).51 We then examine the relationship between the log of each measure and RTRr. Figure
8 reports the results for entry and exit, and Figure 9 shows the results for job creation and de-
struction. New investment, as observed in establishment entry and job creation, falls immediately
in negatively affected regions and stays low throughout the sample period. In contrast, the exit
48In columns (1) and (2), tradable industry price changes are measured using the change in log one plus the tariffrate, and in column (1) the nontradable industry price change is measured using RTRr, following Kovak (2013).
49It is not possible to construct regional capital stocks in Brazil during our sample period. Capital investmentin manufacturing firms could in principle be constructed from the Annual Manufacturing Survey (PIA) beginningin 1996, but the Brazilian Statistical Agency (IBGE) has a strict policy against constructing PIA variables at theregional level. Moreover, with investment data beginning in 1996, we would not have credible capital stock measuresuntil well after liberalization. Data sources covering non-manufacturing sectors also begin well after liberalization.
50Regional capital could slowly reallocate from firms in the formal sector to firms in the informal sector, but thisis unlikely, as firms in the informal sector are much less capital intensive than those in the formal sector (LaPortaand Schleifer 2014, Fajnzylber, Maloney and Montes-Rojas 2011)
51For establishment entry and exit, the Davis and Haltiwanger (1990) measure reduces to the number of establish-ments that entered or exited between 1991 and year t as a share of active establishments in year t. See AppendixA.7 for details.
24
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
and job destruction effects grow slowly over time as existing establishments in regions facing larger
tariff cuts allow their installed capital stocks to erode through depreciation, directing investment
elsewhere. Together, these results support the conclusion that capital slowly reallocated away from
regions facing larger tariff declines, steadily amplifying the earnings effects of liberalization.
5.4.3 Quantification
The preceding results provide evidence that both slow capital adjustment and agglomeration
economies play qualitatively important roles in driving the evolution of liberalization’s effects on
earnings and employment. We now investigate the extent to which these mechanisms can quanti-
tatively explain the long-run labor market effects we observe.
We begin by examining the extent to which regional capital adjustment can account for the
long-run evolution of labor demand inferred from (8). Capital’s contribution to overall adjustment
is given by ϕ(1− ζ)∑
i λriKri. We proxy for∑
i λriKri using the change in log number of regional
formal establishments (as discussed above) and measure ζ (fixed-factors’ share of non-labor input
costs), using estimates of equipment, structures, and land cost shares from Valentinyi and Her-
rendorf (2008).52 We consider three alternative values for ζ, defining fixed factors as i) land only
(ζ = 0.152), ii) land and structures (ζ = 0.545) and iii) land and half of structures (ζ = 0.349).53
Figure 6 shows the evolution of liberalization’s effect on these capital adjustment measures com-
pared to the overall labor demand adjustment inferred from (8). Although the shapes of the capital
adjustment and overall adjustment profiles are not identical, they both grow over time and have
similar scales. Depending on the value of ζ, capital adjustment can account for between 47 and 88
percent of the inferred labor demand adjustment in 2010. While this is a somewhat wide range,
it is clear that capital adjustment accounts for an important share of overall long-run labor de-
mand adjustment, but that it is unlikely to account for all of the adjustment in the absence of
agglomeration.
To quantify the strength of agglomeration economies needed to rationalize the data, we first
need to estimate the inverse labor supply elasticity, η. We do so following (10) by regressing the
1991-2010 change in log formal employment on the change in log regional earnings premium with
RTRr serving as an instrument for wr. The resulting estimate of 0.363 is shown in Panel A of
Table 6. Given this value for η, we estimate κ using non-linear least squares based on long-run
changes in regional earnings in (11) or long-run changes in employment in (12). In both cases, the
R term is captured by the intercept, and the regional weighted average price shocks are measured
by RTRr. When estimating equation (12), we include all industries and control for industry price
52Agglomeration estimation exercises regularly require cost share calibrations along these lines, e.g. Kline andMoretti (2014).
53While i) is likely an underestimate because there are fixed inputs other than land (e.g. heavy infrastructure),ii) is likely an overestimate, because some structures depreciate substantially at a 15 year time horizon. Thus, theintermediate value, iii), is our preferred estimate.
25
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
changes using tariff changes, as in column (1) of Table 5, though the results are nearly identical
when using the alternative approaches in columns (2) or (3) of Table 5. We show estimates for
each value of ζ and bootstrap the entire estimation procedure when calculating standard errors to
account for potential correlation between the η and κ estimates.
The resulting estimates of κ appear in Panel B of Table 6. All of the estimates are positive and
fall within the range of the prior literature (Melo et al. 2009). For example, Kline and Moretti (2014)
find an estimate of 0.2, which is quite close to our wage-based estimate of 0.188 for the intermediate
value of ζ. The value of ζ is important in determining the magnitude of the agglomeration elasticity,
which is unsurprising since Figure 6 showed that capital adjustment explains a smaller share of
overall adjustment for higher values of ζ, leaving a larger role for agglomeration economies.
The estimates in Table 6 and the patterns in Figure 6 show that capital adjustment and standard
agglomeration economies can quantitatively account for the long-run behavior of regional earnings in
response to liberalization. Along with this long-run evidence, Figures 4, 7, and 8 show that regional
labor and capital evolved slowly over time following liberalization and did so in a way consistent with
growing earnings effects of liberalization. In contrast to the other mechanisms that we considered,
dynamic labor demand, driven by slow capital adjustment and agglomeration economies, is both
qualitatively and quantitatively consistent with the earnings responses in Figure 3.
6 Conclusion
This paper documents regional labor market dynamics following the Brazilian trade liberalization
of the early 1990s. Using 25 years of administrative employment data, we find large and growing
effects of trade liberalization on regional formal earnings and employment. Contrary to conventional
wisdom, which assumes wage-equalizing labor adjustment, the regional effects of liberalization grow
for more than a decade before leveling off. This pattern is not driven by post-liberalization economic
shocks and is robust to a wide variety of alternative specifications. After ruling out a number of
potential mechanisms that could generate these growing effects over time, we find strong evidence
in support of a combination of imperfect interregional labor mobility and dynamic labor demand,
driven by slow capital adjustment and agglomeration economies.
Our results have important implications for our thinking about the labor market effects of trade
liberalization. A growing literature has shown in a variety of contexts that trade and trade policy
have heterogeneous effects across regions in the short-run. However, most researchers, ourselves
included, generally assumed that these effects would be upper bounds on the long-run effects, as
labor reallocation would arbitrage away regional differences. This paper finds precisely the opposite.
Short-run effects vastly underestimate the long-run effects, indicating that the costs and benefits
of liberalization remain sharply unevenly distributed across geography, even twenty years after the
policy began.
26
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Our empirical results also inform a large and growing literature using structural models of the
labor market to study trade-induced transitional dynamics. We document the importance of re-
gional adjustment to trade liberalization, even in the long run, and highlight margins of adjustment
that have received little attention by this line of work.54 We find evidence for slow capital adjust-
ment in response to trade liberalization, reinforcing the message of Dix-Carneiro (2014) that jointly
quantifying mobility frictions for labor and other factors such as capital is key to understanding
trade adjustment.55 We also find that agglomeration economies are quantitatively important in
accounting for the magnitudes of trade’s effects on regional earnings, suggesting another feature
for inclusion in models examining the effects of trade shocks on labor markets.
54With the exception of Caliendo et al. (2015), this literature has abstracted from geography.55Artuc, Bet, Brambilla and Porto (2014) take an initial step in this direction.
27
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
References
Abreu, Marcelo de Paiva, “The Political Economy of High Protection in Brazil before 1987,” Inter-AmericanDevelopment Bank Special Initiative on Trade and Integration Working Paper, 2004, (SITI-08A).
Adao, Rodrigo, “Worker Heterogeneity, Wage Inequality, and International Trade: Theory and Evidence fromBrazil,” Unpublished, 2015.
Arbache, Jorge Saba, Andy Dickerson, and Francis Green, “Trade Liberalisation and Wages in DevelopingCountries,” Economic Journal, 2004, 114 (493), F73–F96.
Arcidiacono, Peter, Gigi Foster, Natalie Goodpaster, and Josh Kinsler, “Estimating Spillovers Using PanelData, with an Application to the Classroom,” Quantative Economics, 2012, 3 (3), 421–470.
Artuc, Erhan, German Bet, Irene Brambilla, and Guido Porto, “Trade Shocks and Factor AdjustmentFrictions: Implications for Investment and Labor,” unpublished, 2014.
, Shubham Chaudhuri, and John McLaren, “Trade Shocks and Labor Adjustment: A Structural EmpiricalApproach,” American Economic Review, 2010, 100 (3), 1008–45.
Autor, David, David Dorn, and Gordon Hanson, “The China Syndrome: Local Labor Market Effects of ImportCompetition in the United States,” American Economic Review, 2013, 103 (6).
, , , and Jae Song, “Trade Adjustment: Worker Level Evidence,” Quarterly Journal of Economics,2014, 129 (4), 1799–1860.
Blanchard, Olivier Jean and Lawrence F. Katz, “Regional Evolutions,” Brookings Papers on Economic Activity,1992, (1), 1–75.
Bound, John and Harry J. Holzer, “Demand Shifts, Population Adjustments, and Labor Market Outcomesduring the 1980s,” Journal of Labor Economics, 2000, 18 (1), 20–54.
Bustos, Paula, Bruno Caprettini, and Jacopo Ponticelli, “Agricultural Productivity and Structural Transfor-mation. Evidence from Brazil,” American Economic Review, 2016, 106 (6), 1320–1365.
Caliendo, Lorenzo, Maximiliano Dvorkin, and Fernando Parro, “Trade and Labor Market Dynamics,”Unpublished, 2015.
Cosar, Kerem, “Adjusting to Trade Liberalization: Reallocation and Labor Market Policies,” unpublished, 2013.
Costa, Francisco J.M., Jason Garred, and Joao Paulo Pessoa, “Winners and Losers from a Commodities-for-Manufactures Trade Boom,” Journal of International Economics, 2016, 102, 50–69.
Davidson, Russell and James G. MacKinnon, “Bootstrap Methods in Econometrics,” in Terence C. Mills andKerry Patterson, eds., Palgrave Handbook of Econometrics, Vol. 1, Basingstoke: Palgrave MacMillan, 2006,pp. 812–838.
Davis, Steven J. and John Haltiwanger, “Gross Job Creation and Destruction: Microeconomic Evidence andMacroeconomic Implications,” NBER Macroeconomics Annual, 1990, 5.
de Carvalho, Jr., Mario C., “Alguns Aspectos da Reforma Aduaneira Recente,” FUNCEX Texto Para Discussao,1992.
De Negri, Joao Alberto, Paulo Furtado de Castro, Natalia Ribeiro de Souza, and Jorge Saba Arbache,“Mercado Formal de Trabalho: Comparacao entre os Microdados da RAIS e da PNAD,” IPEA Texto ParaDiscussao, 2001, (840).
Dix-Carneiro, Rafael, “Trade Liberalization and Labor Market Dynamics,” Econometrica, 2014, 82 (3).
and Brian K. Kovak, “Trade Liberalization and the Skill Premium: A Local Labor Markets Approach,”American Economic Review, 2015, 105 (5), 551–557.
and , “Trade Reform and Regional Dynamics: Evidence From 25 Years of Brazilian Matched Employer-Employee Data,” NBER Working Paper, 2015, (w20908).
Edmonds, Eric, Nina Pavcnik, and Petia Topalova, “Trade Adjustment and Human Capital Investment:Evidence from Indian Tariff Reform,” American Economic Journal: Applied Economics, 2010, 2 (4), 42–75.
Fajnzylber, Pablo, William F. Maloney, and Gabriel V. Montes-Rojas, “Does formality improve micro-firmperformance? Evidence from the Brazilian SIMPLES program,” Journal of Development Economics, 2011, 94(2), 262 – 276.
28
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Glaeser, Edward L. and Joseph Gyourko, “Urban Decline and Durable Housing,” Journal of Political Economy,2005, 113 (2), 345–375.
Goldberg, Pinelopi and Nina Pavcnik, “Trade, Wages, and the Political Economy of Trade Protection: Evidencefrom the Colombian Trade Reforms,” Journal of International Economics, 2005, 66 (1), 75–105.
and , “Distributional Effects of Globalization in Developing Countries,” Journal of Economic Literature,2007, XLV, 39–82.
Goldberg, Pinelopi Koujianou and Nina Pavcnik, “The response of the informal sector to trade liberalization,”Journal of Development Economics, 2003, 72 (3), 463–496.
Gonzaga, Gustavo, Naercio Menezes Filho, and Cristina Terra, “Trade liberalization and the evolution ofskill earnings differentials in Brazil,” Journal of International Economics, 2006, 68 (2), 345–367.
Hakobyan, Shushanik and John McLaren, “Looking for Local Labor Market Effects of NAFTA,” Review ofEconomics and Statistics, forthcoming.
Hanlon, W. Walker and Antonio Miscio, “Agglomeration: A Long-Run Panel Data Approach,” Unpublished,2016.
Hasan, Rana, Devasish Mitra, and Beyza P. Ural, “Trade Liberalization, Labor-Market Institutions, andPoverty Reduction: Evidence from Indian States,” India Policy Forum, 2006, 3.
, , Priya Ranjan, and Reshad N. Ahsan, “Trade Liberalization and Unemployment: Theory andEvidence from India,” Journal of Development Economics, 2012, 97 (2).
Helm, Ines, “National Industry Trade Shocks, Local Labor Markets, and Agglomeration Spillovers,” Unpublished,2016.
Helpman, Elhanan, Oleg Itskhoki, Marc-Andreas Muendler, and Steven J. Redding, “Trade and Inequal-ity: From Theory to Estimation,” Review of Economic Studies, forthcoming.
Hollweg, Claire H., Daniel Lederman, Diego Rojas, and Elizabeth Ruppert Bulmer, Sticky feet : howlabor market frictions shape the impact of international trade on jobs and wages., Washington D.C.: The WorldBank, 2014.
IBGE, Censo Demografico 2000: Documentacao das Microdados da Amostra, Instituto Brasileiro de Geografia eEstatıstica, 2002.
, Sistema Nacional de Indices de Precos ao Consumidor, Vol. 34 of Serie Relatorios Metodologicos, InstitutoBrasileiro de Geografia e Estatıstica, 2005.
Kambourov, Gueorgui, “Labor Market Regulations and the Sectoral Reallocation of Workers: The Case of TradeReforms,” Review of Economic Studies, 2009, 76 (4), 1321–1358.
Kline, Patrick and Enrico Moretti, “Local Economic Development, Agglomeration Economies and the Big Push:100 Years of Evidence from the Tennessee Valley Authority,” Quarterly Journal of Economics, 2014, 129 (1),275–331.
Kondo, Illenin O., “Trade Reforms, Foreign Competition, and Labor Market Adjustments in the U.S.,” unpublished,2014.
Kovak, Brian, “Regional Efects of Trade Reform: What is the Correct Measure of Liberalization?,” AmericanEconomic Review, 2013, 103 (5), 1960–1976.
Krishna, Pravin, Jennifer P. Poole, and Mine Zeynep Senses, “Wage Effects of Trade Reform with Endoge-nous Worker Mobility,” NBER Working Paper, 2011, (17256).
, , and , “Wage Effects of Trade Reform with Endogenous Worker Mobility,” Journal of InternationalEconomics, 2014, 93 (2), 239–252.
Kume, Honorio, “A Polıtica Tarifaria Brasileira no Perıodo 1980-88: Avaliacao e Reforma,” Serie Epico, March1990, (17).
, Guida Piani, and Carlos Frederico Braz de Souza, “A Polıtica Brasileira de Importacao no Perıodo1987-1998: Descricao e Avaliacao,” in Carlos Henrique Corseuil and Honorio Kume, eds., A Abertura ComercialBrasileira nos Anos 1990: Impactos Sobre Emprego e Salario, Rio de Janiero: MTE/IPEA, 2003, chapter 1,pp. 1–37.
29
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
LaPorta, Rafael and Andrei Schleifer, “Informality and Development,” Journal of Economic Perspectives,Summer 2014, 28 (3), 109–126.
Lopes de Melo, Rafael, “Firm Wage Differentials and Labor Market Sorting: Reconciling Theory and Evidence,”unpublished, 2013.
MacKinnon, James G., “Thirty Years of Heteroskedasticity-Robust Inference,” Queen’s Economics DepartmentWorking Paper, 2011, (1268).
McCaig, Brian, “Exporting Out of Poverty: Provincial Poverty in Vietnam and US Market Access,” Journal ofInternational Economics, 2011, 85 (1).
Melo, Patricia C., Daniel J. Graham, and Robert B. Noland, “A Meta-Analysis of Estimates of UrbanAgglomeration Economies,” Regional Science and Urban Economics, 2009, 39, 332–342.
Menezes-Filho, Naercio and Marc-Andreas Muendler, “Labor Reallocation in Response to Trade Reform,”NBER Working Paper, 2011, (17372).
Monte, Ferdinando, “The Local Incidence of Trade Shocks,” Unpublished, 2016.
Moretti, Enrico, “Real Wage Inequality,” American Economic Journal: Applied Economics, 2013, 5 (1), 65–103.
Neri, Marcelo and Rodrigo Moura, “Brasil: La institucionalidad del salario mınimo,” in Andres Marinakis andJuan Jacobo Velasco, eds., Para que sirve el salario mınimo?, Organizacion Internacional del Trabajo, 2006,pp. 105–158.
Notowidigdo, Matthew J., “The Incidence of Local Labor Demand Shocks,” Unpublished, 2013.
Pavcnik, Nina, Andreas Blom, Pinelopi Goldberg, and Norbert Schady, “Trade Liberalization and IndustryWage Structure: Evidence from Brazil,” World Bank Economic Review, 2004, 18 (3), 319–334.
Paz, Lourenco, “The impacts of trade liberalization on informal labor markets: an evaluation of the Brazilian case,”Journal of International Economics, 2014, 92 (2), 330–348.
Saboia, Joao L. M. and Ricardo M. L. Tolipan, “A relacao anual de informacoes sociais (RAIS) e o mercadoformal de trabalho no Brasil: uma nota,” Pesquisa e Planejamento Economico, 1985, 15 (2), 447–456.
Salem, Samira and John Benedetto, “The USITC’s Roundtable on the Labor Market Effects of Trade: DiscussionSummary,” Journal of International Commerce and Economics, August 2013.
Schor, Adriana, “Heterogeneous productivity response to tariff reduction. Evidence from Brazilian manufacturingfirms,” Journal of Development Economics, 2004, 75 (2), 373–396.
Soares, Rodrigo R. and Guilherme Hirata, “Competition and the Racial Wage Gap: Testing Becker’s Modelof Employer Discrimination,” IZA Discussion Paper, Februray 2016, (9764).
Stock, James H and Motohiro Yogo, “Testing for Weak Instruments in Linear IV Regression,” in Donald W. K.Andrews and James H. Stock, eds., Identification and Inference for Econometric Models: Essays in Honor ofThomas Rothenberg, Cambridge University Press, 2005, pp. 80–108.
Topalova, Petia, “Trade Liberalization, Poverty, and Inequality: Evidence from Indian Districts,” in Ann Harrison,ed., Globalization and Poverty, University of Chicago Press, 2007, pp. 291–336.
, “Factor Immobility and Regional Impacts of Trade Liberalization: Evidence on Poverty from India,” AmericanEconomic Journal: Applied Economics, 2010, 2 (4).
Traiberman, Sharon, “Occupations and Import Competition,” Unpublished, 2016.
Valentinyi, Akos and Berthold Herrendorf, “Measuring Factor Income Shares at the Sectoral Level,” Reviewof Economic Dynamics, 2008, 11, 820–835.
30
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 1: Tariff Changes
-0.25
-0.20
-0.15
-0.10
-0.05
0.00
Chan
ge in
ln(1
+tar
iff),
1990
-95
Agric
ultu
re
Met
als
Appa
rel
Food
Pro
cess
ing
Woo
d, F
urni
ture
, Pea
t
Text
iles
Nonm
etal
lic M
iner
al M
anuf
Pape
r, Pu
blish
ing,
Prin
ting
Min
eral
Min
ing
Foot
wear
, Lea
ther
Chem
icals
Auto
, Tra
nspo
rt, V
ehicl
es
Elec
tric,
Ele
ctro
nic
Equi
p.
Mac
hine
ry, E
quip
men
t
Plas
tics
Oth
er M
anuf
.
Phar
ma.
, Per
fum
es, D
eter
gent
s
Petro
leum
Refi
ning
Rubb
er
Petro
leum
, Gas
, Coa
l
Tariff data from Kume et al. (2003), aggregated to allow consistent industry definitions across data sources. SeeAppendix Table A1 for details of the industry classification. Industries sorted based on 1991 national employment(largest on the left, and smallest on the right)
31
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 2: Regional Tariff Reductions
BelémBelém
RecifeRecife
ManausManaus
CuritibaCuritiba
BrasíliaBrasília
SalvadorSalvador
FortalezaFortaleza
São PauloSão Paulo
Porto AlegrePorto Alegre
Belo HorizonteBelo Horizonte
8% to 15%4% to 8%3% to 4%1% to 3%-1% to 1%
Local labor markets reflect microregions defined by IBGE, aggregated slightly to account for border changes between1986 and 2010. Regions are colored based on the regional tariff reduction measure, RTRr, defined in (2). Regionsfacing larger tariff reductions are presented as lighter and yellower, while regions facing smaller cuts are shown asdarker and bluer. Dark lines represent state borders, gray lines represent consistent microregion borders, and cross-hatched migroregions are omitted from the analysis. These microregions were either i) part of a Free Trade Area ii)part of the state of Tocantins and not consistently identifiable over time, or iii) not included in the RAIS samplebefore 1990.
32
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 3: Regional log Formal Earnings Premia - 1992-2010
-‐2.0
-‐1.5
-‐1.0
-‐0.5
0.0
0.5
1.0
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Pre-‐liberaliza6on (chg. from 1986)
Liberaliza6on Post-‐liberaliza6on (chg. from 1991)
Each point reflects an individual regression coefficient, θt, following (3), where the dependent variable is the changein regional log formal earnings premium and the independent variable is the regional tariff reduction (RTR), definedin (2). Note that the RTR always reflects tariff reductions from 1990-1995. For blue circles, the changes are from1991 to the year listed on the x-axis. For purple diamonds, the changes are from 1986 to the year listed. Allregressions include state fixed effects, and post-liberalization regressions control for the 1986-1990 outcome pre-trend.Negative estimates imply larger earnings declines in regions facing larger tariff reductions. Vertical bars indicate thatliberalization began in 1991 and was complete by 1995. Dashed lines show 95 percent confidence intervals. Standarderrors adjusted for 112 mesoregion clusters.
33
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 4: Regional log Formal Employment - 1992-2010
-‐6.5
-‐5.5
-‐4.5
-‐3.5
-‐2.5
-‐1.5
-‐0.5
0.5
1.5
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Pre-‐liberaliza6on (chg. from 1986)
Liberaliza6on Post-‐liberaliza6on (chg. from 1991)
Liberaliza6on Post-‐liberaliza6on (chg. from 1991)
Liberaliza6on Post-‐liberaliza6on (chg. from 1991)
Each point reflects an individual regression coefficient, θt, following (3), where the dependent variable is the changein regional log formal employment and the independent variable is the regional tariff reduction (RTR), defined in(2). Note that the RTR always reflects tariff reductions from 1990-1995. For blue circles, the changes are from 1991to the year listed on the x-axis. For purple diamonds, the changes are from 1986 to the year listed. All regressionsinclude state fixed effects, and post-liberalization regressions control for the 1986-1990 outcome pre-trend. Negativeestimates imply larger employment declines in regions facing larger tariff reductions. Vertical bars indicate thatliberalization began in 1991 and was complete by 1995. Dashed lines show 95 percent confidence intervals. Standarderrors adjusted for 112 mesoregion clusters.
34
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 5: Regional Imports, Exports, and Net Exports Per Worker - 1992-2010
-‐0.4
-‐0.2
0
0.2
0.4
0.6
0.8
1
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Imports
Exports
Net Exports
Liberaliza6on Post-‐liberaliza6on (chg. from 1991)
Each point reflects an individual regression coefficient, θt, following (3), where the dependent variable is the changein regional imports per worker (blue circles), exports per worker (red triangles), or net exports per worker (greendiamonds), measured in $100,000 units. The independent variable is the regional tariff reduction (RTR), defined in(2). Note that the RTR always reflects tariff reductions from 1990-1995. All regressions include state fixed effects,but do not include pre-liberalization trends due to a lack of Comtrade trade data before 1989. Positive estimatesimply larger increases in trade flow per worker in regions facing larger tariff reductions. Vertical bar indicates thatliberalization was complete by 1995. Dashed lines show 95 percent confidence intervals. Standard errors adjusted for112 mesoregion clusters.
35
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 6: Inferred Adjustment and Capital Adjustment Quantification - 1992-2010
-‐3.0
-‐2.5
-‐2.0
-‐1.5
-‐1.0
-‐0.5
0.0
0.5
1.0
1.5
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Pre-‐liberaliza6on (chg. from 1986) Liberaliza6on Post-‐liberaliza6on
(chg. from 1991)
Inferred Adjustment
Capital (establishments) adjustment
ζ = 0.152
ζ = 0.349
ζ = 0.545
Each point reflects an individual regression coefficient, θt, following (3). For the blue profile with solid circles, thedependent variable is the inferred labor demand shifts from agglomeration and capital adjustment, defined in (8). Forthe gray profiles with hollow markers, the dependent variable is capital’s contribution to overall adjustment, using thechange in the number of regional formal establishments as a proxy for the change in regional capital,
∑i λriKri. We
present profiles for three values of ζ, specific factors’ share of non-labor inputs, based on Valentinyi and Herrendorf(2008). The independent variable is the regional tariff reduction (RTR), defined in (2). Note that the RTR alwaysreflects tariff reductions from 1990-1995. All regressions include state fixed effects, and post-liberalization regressionscontrol for the 1986-1990 outcome pre-trend. Negative estimates imply larger declines in residual labor demand orthe number of establishments in regions facing larger tariff reductions. Vertical bar indicates that liberalization wascomplete by 1995. Dashed lines show 95 percent confidence intervals. Confidence intervals for capital adjustmentprofiles shown in Appendix B.12. Standard errors adjusted for 112 mesoregion clusters.
36
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 7: Regional log Number of Formal Establishments and log Average Formal EstablishmentSize (Number of Workers)- 1992-2010
-‐5
-‐4
-‐3
-‐2
-‐1
0
1
2
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Pre-‐liberaliza5on (chg. from 1986)
Liberaliza5on Post-‐liberaliza5on (chg. from 1991)
Establishment Size
Establishments
Liberaliza5on Post-‐liberaliza5on (chg. from 1991)
Estab. Size Pretrend
Establishments Pretrend
Each point reflects an individual regression coefficient, θt, following (3), where the dependent variable is the changein regional log number of formal establishments or the change in regional log average formal establishment size.The independent variable is the regional tariff reduction (RTR), defined in (2). Note that the RTR always reflectstariff reductions from 1990-1995. For blue circles and red triangles, the changes are from 1991 to the year listed onthe x-axis. For purple diamonds and orange squares, the changes are from 1986 to the year listed. All regressionsinclude state fixed effects, and post-liberalization regressions control for the 1986-1990 outcome pre-trend. Negativeestimates imply larger declines in the number of establishments or average establishment size in regions facing largertariff reductions. Vertical bars indicate that liberalization began in 1991 and was complete by 1995. Dashed linesshow 95 percent confidence intervals. Standard errors adjusted for 112 mesoregion clusters.
37
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 8: Regional log Cumulative Formal Establishment Entry and Exit - 1992-2010
-‐3
-‐2
-‐1
0
1
2
3
4
5
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Exit
Entry
Pre-‐liberaliza5on (chg. from 1986)
Liberaliza5on Post-‐liberaliza5on (chg. from 1991)
Exit Pretrend
Entry Pretrend
Exit
Entry
Liberaliza5on Post-‐liberaliza5on (chg. from 1991)
Exit Pretrend
Entry Pretrend
Each point reflects an individual regression coefficient. The dependent variable is the log cumulative formal estab-lishment entry or exit from 1991 to the year listed on the x-axis (blue circles and red triangles) or from 1986 to theyear listed (purple diamonds and orange squares), calculated as in Davis and Haltiwanger (1990). The independentvariable is the regional tariff reduction (RTR), defined in (2). Note that the RTR always reflects tariff reductions from1990-1995. All regressions include state fixed effects, and post-liberalization regressions control for log cumulativeestablishment entry or exit during 1986-1990. Positive exit estimates and negative entry estimates imply larger ratesof exit and smaller rates of entry in regions facing larger tariff reductions. Vertical bars indicate that liberalizationbegan in 1991 and was complete by 1995. Dashed lines show 95 percent confidence intervals. Standard errors adjustedfor 112 mesoregion clusters.
38
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure 9: Regional log Cumulative Job Creation and Destruction - 1992-2010
-‐4
-‐3
-‐2
-‐1
0
1
2
3
4
5
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Job Crea)on
Job Destruc)on
Pre-‐liberaliza5on (chg. from 1986)
Liberaliza5on Post-‐liberaliza5on (chg. from 1991)
Job Crea)on Pretrend
Job Destruc)on Pretrend
Each point reflects an individual regression coefficient. The dependent variable is the log cumulative job creationor destruction rate from 1991 to the year listed on the x-axis (blue circles and red triangles) or from 1986 to theyear listed (purple diamonds and orange squares), calculated as in Davis and Haltiwanger (1990). The independentvariable is the regional tariff reduction (RTR), defined in (2). Note that the RTR always reflects tariff reductions from1990-1995. All regressions include state fixed effects, and post-liberalization regressions control for log cumulativejob creation or destruction during 1986-1990. Positive job destruction estimates and negative job creation estimatesimply larger rates of job destruction and smaller rates of job creation in regions facing larger tariff reductions. Verticalbars indicate that liberalization began in 1991 and was complete by 1995. Dashed lines show 95 percent confidenceintervals. Standard errors adjusted for 112 mesoregion clusters.
39
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table 1: Regional log Formal Earnings Premia and Employment - 2000, 2010
Change in outcome: (1) (2) (3) (4) (5) (6)
Panel A: log Formal Earnings PremiaRegional tariff reduction (RTR) -0.451*** -0.638*** -0.529*** -1.885*** -1.736*** -1.594***
(0.152) (0.154) (0.141) (0.316) (0.184) (0.169)Formal earnings pre-trend (86-90) -0.312** -0.418***
(0.149) (0.144)State fixed effects (26) ✓ ✓ ✓ ✓
R-squared 0.040 0.225 0.268 0.320 0.501 0.537
Panel B: log Formal Real Earnings Premia (regional deflators following Moretti (2013))Regional tariff reduction (RTR) -1.594*** -1.382*** -1.260***
(0.306) (0.180) (0.168)Formal earnings pre-trend (86-90) -0.359***
(0.133)State fixed effects (26) ✓ ✓
R-squared 0.238 0.449 0.477
Panel C: log Formal EmploymentRegional tariff reduction (RTR) -3.748*** -3.545*** -3.533*** -6.059*** -4.675*** -4.663***
(0.516) (0.563) (0.582) (0.560) (0.660) (0.679)Formal employment pre-trend (86-90) -0.0331 -0.0319
(0.147) (0.156)State fixed effects (26) ✓ ✓ ✓ ✓
R-squared 0.072 0.291 0.291 0.149 0.409 0.410
1991-2000 1991-2010
Negative coefficient estimates for the regional tariff reduction imply larger declines in formal earnings or employmentin regions facing larger tariff reductions. Microregion observations: Panels A and C, 475; Panel B, 456 (omits a fewsparsely populated locations with insufficient data to calculate regional price deflators. Regional earnings premiacalculated controlling for age, sex, education, and industry of employment. Panels A and B: efficiency weightedby the inverse of the squared standard error of the estimated change in log formal earnings premium. Pre-trendscomputed for 1986-1990. Standard errors (in parentheses) adjusted for 112 mesoregion clusters. *** Significant atthe 1 percent, ** 5 percent, * 10 percent level.
40
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table 2: Regional log Working-Age Population - 2000, 2010
Change in log Working-Age Population: OLS IV IV OLS IV IV
1980 level 1970-80 chg. 1980 level 1970-80 chg.(1) (2) (3) (4) (5) (6)
Regional tariff reduction (RTR) 0.333 0.111 0.085 0.392 0.038 0.027(0.243) (0.159) (0.166) (0.319) (0.241) (0.232)
Population pre-trend (80-91) 0.406** 0.585*** 0.606*** 0.632*** 0.918*** 0.927***(0.164) (0.100) (0.105) (0.225) (0.182) (0.148)
State fixed effects (26) ✓ ✓ ✓ ✓ ✓ ✓
R-squared 0.654 0.603 0.590 0.666 0.615 0.6121st stage F (Kleibergen-Paap) 18.23 32.71 17.31 31.90
1991-2000 1991-2010
Positive (negative) coefficient estimates for the regional tariff reduction imply larger increases (decreases) in populationin regions facing larger tariff reductions. Outcomes calculated using Census data. 405 microregion observations.Efficiency weighted by the inverse of the squared standard error of the dependent variable estimate. Pre-trendscomputed for 1980-1991. To address potential mechanical endogeneity due to pre-trend and dependent variableoverlap in 1991, columns (2) and (5) use the 1980 dependent variable level as an instrument for the pre-trend andcolumns (3) and (6) use the 1970-1980 change as an instrument. Standard errors (in parentheses) adjusted for 90mesoregion clusters. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
41
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table 3: Mechanisms: Changing Worker Composition - 1995, 2000, 2005, 2010
Change in log Formal Earnings Premia: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Main specificationRegional tariff reduction (RTR) -0.096 -0.529*** -1.294*** -1.594***
(0.120) (0.141) (0.139) (0.169)Panel B: Earnings premia controlling for individual fixed effects (fixed returns)
Regional tariff reduction (RTR) -0.193* -0.514*** -1.119*** -1.271***(0.115) (0.144) (0.147) (0.172)
Panel C: Earnigns premia controlling for individual fixed effects (time-varying returns)Regional tariff reduction (RTR) -0.230** -0.551*** -1.322*** -1.454***
(0.093) (0.098) (0.094) (0.119)Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal earnings inregions facing larger tariff reductions. Microregion observations: Panel A, 475; Panels B and C, 450 (omits regionswith insufficient observations to identify region-year fixed effects in any particular year). Regional earnings premia:Panel A: calculated controlling for age, sex, education, and industry of employment; Panels B and C: controlling forindividual fixed effects. Standard errors (in parentheses) adjusted for 112 mesoregion clusters. Efficiency weightedby the inverse of the squared standard error of the estimated change in log formal earnings premium. See text fordetailed description of each panel. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
42
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table 4: Mechanisms: Slow Response of Imports or Exports - 1995, 2000, 2005, 2010
Change in log Formal Earnings Premia: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Main specificationRegional tariff reduction (RTR) -0.096 -0.529*** -1.294*** -1.594***
(0.120) (0.141) (0.139) (0.169)
Panel B: Controls for trade quantity shocks (OLS)Regional tariff reduction (RTR) -0.089 -0.521*** -1.287*** -1.562***
(0.112) (0.138) (0.181) (0.221)Import quantity shock
Export quantity shock
Panel C: Latin America IVRegional tariff reduction (RTR) -0.129 -0.569*** -1.342*** -1.757***
(0.106) (0.129) (0.173) (0.212)Import quantity shock
Export quantity shock
First-‐stage F (Kleibergen-‐Paap)
Panel D: Colombia IV Regional tariff reduction (RTR) -0.049 -0.488*** -1.372*** -1.502***
(0.108) (0.132) (0.161) (0.213)Import quantity shock
Export quantity shock
First-‐stage F (Kleibergen-‐Paap)
Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
-0.382(2.242)0.142
(3.355)
(2.631)
(3.861)93.04
-3.489(2.427)5.379*(3.268)876.2
1.668
-0.149
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal earnings in regionsfacing larger tariff reductions. Panel A replicates the earnings results in columns (3) and (6) of Table 1. Panels B-Dcontrol for regional import and export quantity shocks as in (5). We instrument for the potentially endogenous importand export shocks using regional measures of commodity price growth from Adao (2015) and with regional tradeflows for other countries. “Latin America” consists of Argentina, Chile, Colombia, Paraguay, Peru, and Uruguay. Wemeasure imports and exports between Latin America or Colombia and the rest of the world excluding Brazil. Due toComtrade data availability, changes in Colombian trade flows are measured from 1991 to each subsequent year andLatin American trade flows from 1994. We allow for time-varying first-stage coefficients, so we have 2 endogenousvariables (RegImprt and RegExprt) and 57 instruments for Colombia (3 instruments × 19 years) and 48 instrumentsfor Latin America (3 instruments × 16 years). First-stage Kleinbergen-Paap F statistics are compared to the Stockand Yogo (2005) critical value of 21 to reject 5 percent bias relative to OLS. Standard errors (in parentheses) adjustedfor 112 mesoregion clusters. Efficiency weighted by the inverse of the squared standard error of the estimated changein log formal earnings premium. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
43
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table 5: Test for Agglomeration Economies
Change in log Region x Industry Employment: All industries Tradable industries Tradable industries(1) (2) (3)
Regional tariff reduction (RTR) -‐6.183*** -‐6.708*** -‐6.704***(0.631) (0.675) (0.694)
Industry price change controls ✓ ✓Industry fixed effects (20) ✓
Observations 4,648 4,174 4,174R-‐squared 0.119 0.120 0.222
Negative coefficient estimates for the regional tariff reduction imply the presence of agglomeration economies, following(13). Observations represent region × industry pairs. The dependent variable is the change in log formal employmentin a given region × industry pair from 1991 to 2010. Regressions include 1986-1990 pre-liberalization employmenttrend controls and state fixed effects. Column (1) covers all industries, including the nontradable sector. In thatcolumn, tradable industry price changes are measured using the change in log one plus the tariff rate, and thenontradable industry price change is measured using RTRr. Columns (2) and (3) include only tradable industries,omitting the nontradable sector. In column (2), tradable industry prices are measured using the change in log oneplus the tariff rate, while in column (3) they are controlled for using industry fixed effects. Standard errors (inparentheses) adjusted for 112 mesoregion clusters. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
Table 6: Agglomeration Elasticity Estimates
Panel A: Inverse labor supply elasticity (η) 0.363***(0.060)
Panel B: Agglomeration elasticity (κ)(1) (2) (3)
Specific factors' share of non-‐labor inputs (ζ): low (0.152) mid (0.349) high (0.545)
Wage-‐based agglomeration elasticity (κ) 0.042* 0.188*** 0.333***(0.023) (0.023) (0.025)
Employment-‐based agglomeration elasticity (κ) 0.215*** 0.330*** 0.461***(0.032) (0.038) (0.043)
Labor supply elasticity, η, estimated from (10) using RTRr as an instrument for the change in regional log earningspremium. The first-stage partial F-statistic (Kleibergen-Paap) for this regression is 59.14. Given the estimate of η,the agglomeration elasticity, κ, is estimated using two alternative methods. The earnings-based approach estimates(11), and the employment-based approach estimates (12), both using nonlinear least squares, and both including1986-1990 pre-liberalization outcome trends and state fixed effects. The employment-based estimates control forindustry price changes as in column (1) of Table 5, and results using the other two approaches are very similar. Wepresent estimates for three different values of ζ, specific factors’ share of non-labor inputs, based on Valentinyi andHerrendorf (2008). See text for details. Standard errors (in parentheses) bootstrapped by regional resampling. ***Significant at the 1 percent, ** 5 percent, * 10 percent level.
44
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Online Appendices
(Not for publication)
A Data and Definitions 46A.1 Tariffs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46A.2 RAIS Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49A.3 Demographic Census . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49A.4 Regional Tariff Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50A.5 Local Price Indexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53A.6 Regional Change in Imports and Exports . . . . . . . . . . . . . . . . . . . . . . . . 54A.7 Entry, Exit, Job Creation, and Job Destruction . . . . . . . . . . . . . . . . . . . . . 54
B Supplemental Empirical Results 56B.1 Industry-Level Outcome Pre-Trends vs. Tariff Reductions . . . . . . . . . . . . . . . 56B.2 Informal Employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58B.3 Regional Earnings Premium Regressions . . . . . . . . . . . . . . . . . . . . . . . . . 61B.4 Formal Earnings Regression Scatterplots . . . . . . . . . . . . . . . . . . . . . . . . . 63B.5 Census Earnings, Wage, and Employment Results . . . . . . . . . . . . . . . . . . . 65B.6 Robustness Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66B.7 Potential Confounders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69B.8 Earnings and Employment Sample Splits . . . . . . . . . . . . . . . . . . . . . . . . . 76B.9 Earnings Premia with Individual Worker Fixed Effects . . . . . . . . . . . . . . . . . 78B.10 Regional Change in log Imports and Exports . . . . . . . . . . . . . . . . . . . . . . 79B.11 Overall Employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85B.12 Capital Adjustment Confidence Intervals . . . . . . . . . . . . . . . . . . . . . . . . . 86B.13 Exit by Establishment Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
C Model 92C.1 Baseline Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92C.2 Agglomeration Economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
45
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
A Data and Definitions
A.1 Tariffs
Tariff data come from Kume et al. (2003), who report nominal tariffs and effective rates of pro-tection from 1987 to 1998 using the Brazilian industry classification Nıvel 50. We aggregate thesetariffs slightly to an industry classification that is consistent with the Demographic Census dataused to construct local tariff shock measures. The classification is presented in Table A1. In ag-gregating, we weight each Nıvel 50 industry by its 1990 industry value added, as reported in IBGENational Accounts data. Figure A1 shows the evolution of nominal tariffs from 1987 to 1998 forthe ten largest industries. The phases of Brazilian liberalization are visible (see Section 2 for adiscussion and citations). Large nominal tariff cuts from 1987-1989 had little effect on protection,due to the presence of substantial nontariff barriers and tariff exemptions. In 1990, the majority ofnontariff barriers and tariff exemptions were abolished, being replaced by tariffs providing equiva-lent protection; note the increase in tariffs in some industries in 1990. During liberalization, from1990 to 1994, tariffs fell in all industries, then were relatively stable from 1995 onward.
In Section 4.2 we calculate post-liberalization tariff changes using UNCTAD TRAINS. SeeAppendix B.7 for details.
46
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Tab
leA
1:C
onsi
sten
tIn
du
stry
Cla
ssifi
cati
onA
cros
sC
ensu
ses
and
Tar
iffD
ata
Indu
stry
Indu
stry
Nam
eN
ível
50
1970
, 198
0, 1
991
Cen
sus (atividade)
2000
, 201
0 C
ensu
s (CNAE-Dom
)1
Agr
icul
ture
101
1-03
7, 0
41, 0
42, 5
8111
01-1
118,
120
1-12
09, 1
300,
140
1, 1
402,
200
1, 2
002,
50
01, 5
002
2M
iner
al M
inin
g (e
xcep
t com
bust
ible
s)2
050,
053
-059
1200
0, 1
3001
, 130
02, 1
4001
-140
043
Petro
leum
and
Gas
Ext
ract
ion
and
Coa
l Min
ing
305
1-05
210
000,
110
004
Non
met
allic
Min
eral
Goo
ds M
anuf
actu
ring
410
026
010,
260
91, 2
6092
5Ir
on a
nd S
teel
, Non
ferr
ous,
and
Oth
er M
etal
Pro
duct
ion
and
Proc
essi
ng5-
711
027
001-
2700
3, 2
8001
, 280
028
Mac
hine
ry, E
quip
men
t, C
omm
erci
al In
stal
latio
n M
anuf
actu
ring,
and
Tra
ctor
Man
ufac
turin
g8
120
2900
110
Elec
trica
l, El
ectro
nic,
and
Com
mun
icat
ion
Equi
pmen
t and
Com
pone
nts M
anuf
actu
ring
10-1
113
029
002,
300
00, 3
1001
, 310
02, 3
2000
, 330
0312
Aut
omob
ile, T
rans
porta
tion,
and
Veh
icle
Par
ts M
anuf
actu
ring
12-1
314
034
001-
3400
3, 3
5010
, 350
20, 3
5030
, 350
9014
Woo
d Pr
oduc
ts, F
urni
ture
Man
ufac
turin
g, a
nd P
eat P
rodu
ctio
n14
150,
151
, 160
2000
0, 3
6010
15Pa
per M
anuf
actu
ring,
Pub
lishi
ng, a
nd P
rintin
g15
170,
290
2100
1, 2
1002
, 220
0016
Rub
ber P
rodu
ct M
anuf
actu
ring
1618
025
010
17C
hem
ical
Pro
duct
Man
ufac
turin
g17
,19
200
2301
0, 2
3030
, 234
00, 2
4010
, 240
9018
Petro
leum
Ref
inin
g an
d Pe
troch
emic
al M
anuf
actu
ring
1820
1, 2
02, 3
52, 4
7723
020
20Ph
arm
aceu
tical
Pro
duct
s, Pe
rfum
es a
nd D
eter
gent
s Man
ufac
turin
g20
210,
220
2402
0, 2
4030
21Pl
astic
s Pro
duct
s Man
ufac
turin
g21
230
2502
022
Text
iles M
anuf
actu
ring
2224
0, 2
4117
001,
170
0223
App
arel
and
App
arel
Acc
esso
ries M
anuf
actu
ring
2325
0,53
218
001,
180
0224
Foot
wea
r and
Lea
ther
and
Hid
e Pr
oduc
ts M
anuf
actu
ring
2419
0, 2
5119
011,
190
12, 1
9020
25Fo
od P
roce
ssin
g (C
offe
e, P
lant
Pro
duct
s, M
eat,
Dai
ry, S
ugar
, Oils
, Bev
erag
es, a
nd O
ther
)25
-31
260,
261
, 270
, 280
1501
0, 1
5021
, 150
22, 1
5030
, 150
41-1
5043
, 150
50, 1
6000
32M
isce
llane
ous O
ther
Pro
duct
s Man
ufac
turin
g32
300
3300
1, 3
3002
, 330
04, 3
3005
, 360
90, 3
7000
91U
tiliti
es33
351,
353
4001
0, 4
0020
, 410
0092
Con
stru
ctio
n34
340,
524
4500
1-45
005
93W
hole
sale
and
Ret
ail T
rade
3541
0-42
4, 5
82, 5
8350
010,
500
30, 5
0040
, 500
50, 5
3010
,530
20, 5
3030
, 530
41,
5304
2, 5
3050
, 530
61-5
3068
, 530
70, 5
3080
, 530
90, 5
3101
, 53
102,
550
2094
Fina
ncia
l Ins
titut
ions
3845
1-45
3, 5
85, 6
1265
000,
660
00, 6
7010
, 670
2095
Rea
l Est
ate
and
Cor
pora
te S
ervi
ces
40, 4
146
1-46
4, 5
43, 5
52, 5
71-5
78, 5
84, 5
8963
022,
700
01, 7
1020
, 720
10, 7
4011
, 740
12, 7
4021
, 740
22,
7403
0, 7
4040
, 740
50, 7
4090
, 920
13, 9
2014
, 920
15, 9
2020
96Tr
ansp
orta
tion
and
Com
mun
icat
ions
36, 3
747
1-47
6, 4
81, 4
82, 5
8860
010,
600
20, 6
0031
, 600
32, 6
0040
, 600
91, 6
0092
, 610
00,
6200
0, 6
3010
, 630
21 ,6
4010
,640
20, 9
1010
97Pr
ivat
e Se
rvic
es39
, 43
511,
512
, 521
-523
, 525
, 531
, 533
, 541
, 542
. 544
, 54
5, 5
51, 5
77, 5
86, 5
87, 6
13-6
19, 6
22-6
24, 6
32, 9
01,
902
1500
, 500
20, 5
3111
, 531
12, 5
3113
, 550
10, 5
5030
, 630
30,
7000
2, 7
1010
, 710
30, 7
2020
, 730
00, 7
4060
, 800
11, 8
0012
, 80
090,
850
11, 8
5012
, 850
13, 8
5020
, 850
30, 9
0000
, 910
20,
9109
1, 9
1092
, 920
11, 9
2012
, 920
30, 9
2040
, 930
10, 9
3020
, 93
030,
930
91, 9
3092
, 950
0098
Publ
ic A
dmin
istra
tion
4235
4, 6
10, 6
11, 6
21, 6
31, 7
11-7
17, 7
21-7
2775
011-
7501
7, 7
5020
NontradableTradable
Consi
sten
tin
dust
rycl
ass
ifica
tion
use
din
gen
erati
ng
loca
lta
riff
shock
sfr
om
Nıv
el50
tari
ffdata
inK
um
eet
al.
(2003)
and
Dec
ennia
lC
ensu
sdata
.
47
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure A1: Tariffs - 1987-1998
Tex$les
Auto, Transport, Vehicles
Food Processing Nonmetallic Mineral Manuf
Electric, Electronic Equip.
Machinery, Equipment Metals
Agriculture
Chemicals
Petroleum Refining
0
10
20
30
40
50
60
70
80
90
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
Nominal tariffs from Kume et al. (2003), aggregated to the industry classification presented in Table A1. The tenlargest industries by 1990 value added are shown.
48
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
A.2 RAIS Data
The Relacao Anual de Informacoes Sociais (RAIS) is a high quality census of the Brazilian formallabor market. Originally, RAIS was created as an operational tool for the Brazilian governmentto i) monitor the entry of foreign workers into the labor market; ii) oversee the records of theFGTS (Fundo de Garantia do Tempo de Servico) program, a national benefits program consistingof employers’ contributions to each of its employees; iii) provide information for administeringseveral government benefits programs such as unemployment insurance; and iv) generate statisticsregarding the formal labor market. Today it is the main tool used by the government to enablethe payment of the ”abono salarial” to eligible workers. This is a government program that paysone additional minimum wage at the end of the year to workers whose average monthly wage wasnot greater than two times the minimum wage, and whose job information was correctly declaredin RAIS, among other minor requirements. Thus, workers have an incentive to ensure that theiremployer is filing the required information. Moreover, firms are required to file, and face fines untilthey do so. Together, these requirements ensure that the data in RAIS are accurate and complete.
Observations in the data are indexed by a worker ID number, the Programa de Integracao So-cial (PIS), and an establishment registration number, the Cadastro Nacional da Pessoa Jurıdica(CNPJ). Both of these identifiers are consistent over time, allowing one to track workers and estab-lishments across years. Establishment industry is reported using the Subsetor IBGE classification,which includes 12 manufacturing industries, 2 primary industries, 11 nontradable industries, and1 other/ignored.56 Worker education is reported using the following 9 education categories (list-ing corresponding years of education in parentheses): illiterate (0), primary school dropout (1-3),primary school graduate (4), middle school dropout (5-7), middle school graduate (8), high schooldropout (9-11), high school graduate (12), college dropout (13-15), and college graduate (≥ 16).
In each year, and for each job, RAIS reports average earnings throughout the year, and earningsin December.57 We focus on labor market outcomes reported in December of each year. This choiceensures that earnings and formal employment status are measured at the same time for all workersand all jobs. It avoids the potential confounding effects on average yearly earnings that might arisein situations where some workers begin working in early in the year and others begin late in theyear.
A.3 Demographic Census
We utilize information from the long form of the Demographic Censuses (Censo Demografico) for1970, 1980, 1991, 2000, and 2010. The long form micro data reflect a 5 percent sample of thepopulation in 1970, 1980, and 2010, a 5.8 percent sample in 1991, and a 6 percent sample in 2000.The primary benefit of the Census for our purposes is the ability to observe those outside formalemployment, who are not present in the RAIS database.
Although our main analysis focuses on monthly earnings, following the information availablein RAIS, the Census provides weekly hours information from 1991-2010, allowing us to calculatehourly wages as monthly earnings divided by 4.33 times weekly hours. Census results for monthlyearnings and hourly wages are very similar. In 1970 and 1980, hours information is presented in
56A less aggregate industry classification (CNAE) is available from 1994 onward, but we need a consistent classifi-cation from 1986-2010, so we use Subsetor IBGE.
57From 1994 onward, RAIS reports hours, making it possible to calculate hourly wages. However, since we need aconsistent measure from 1986-2010, we focus on monthly earnings.
49
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
5 rough bins. Thus, when calculating pre-liberalization trends using data from 1970 and 1980, weuse monthly earnings even when examining hourly wage outcomes.
In 1991-2010, the Census asks whether each worker has a signed work card. This is the standarddefinition of formal employment, and is necessary for a worker to appear in the RAIS sample.Thus, we use this as our primary definition of formal employment. In 1980 and 1991, there is analternative proxy for formal employment, reporting whether the worker’s job includes contributionsto the national social security system. When calculating pre-liberalization outcome trends for 1980-1991, we use this alternative measure to identify formally employed workers. The social securitycontributions proxy appears to be a good one; in 1991, when both measures are available, 95.9percent of workers would be classified identically when using either measure. In 1970, there is noinformation on formality, so pre-liberalization outcome trends for 1970-1980 are calculated for allworkers.
The definition of employment changes across Census years. In 1970 it includes those reportingworking or looking for work during August 1970 (the questionnaire does not separately identifyworking vs. looking for work). In 1980 it includes those who report working during the yearprior to September 1, 1980. In 1991 it includes those reporting working regularly or occasionallyduring the year prior to September 1, 1991. In 2000 and 2010 it includes those who report paidwork, temporary leave, unpaid work, or cultivation for own consumption during the week of July23-29 in 2000 and July 25-31 in 2010. Note that the employment concept changes substantiallyacross years. This highlights yet another benefit of using RAIS as our primary data source, sincethe employment concept in RAIS is consistent throughout the sample. Yet, while the changescomplicate the interpretation of Census-based employment rates over time, there is no reasonto expect systematic differences across regions to result from the changing employment concept.Thus, our cross-region identification strategy should be valid when using the Census to measureemployment in spite of these measurement issues.
A.4 Regional Tariff Changes
Regional tariff reductions, defined in (2), are constructed using information from various sources.Tariff changes come from Kume et al. (2003), and are aggregated from the Nıvel 50 level to the in-dustry classification presented in Table A1 using 1990 value-added weights from the IBGE NationalAccounts. Figure 1 shows the resulting industry-level variation in tariff changes.
The weights, βri in (2) depend upon the initial regional industry distribution (λri) and thespecific-factor share in production (ϕi). We calculate the λri using the 1991 Census. We use theCensus because it provides a less aggregate industry definition than what is available in RAIS, andbecause the Census allows us to calculate weights that are representative of overall employment,rather than just formal employment. However, note that shocks using formal employment weightsyield very similar results (see Panel D of Table B5). We calculate the ϕi using data from the UseTable of the 1990 National Accounts from IBGE. The table “Componentes do Valor Adicionado”provides the wagebill (Remuneracoes) and gross operating surplus (Excedente Operacional BrutoInclusive Rendimento de Autonomos), which reflects the share of income earned by capital. Wedefine ϕi as capital’s share of the sum of these two components. When imposing equal cost sharesacross industries (see Section 5.4.2), we calculate ϕ using the economy-wide wagebill and grossoperating surplus, yielding a value of ϕ = 0.544.
Because Brazilian local labor markets differ substantially in the industry distribution of their
50
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
employment, the weights βri vary across regions. Figure A2 demonstrates how variation in industrymix leads to variation in RTRr. The figure shows the initial industry distribution of employment forthe regions facing the largest tariff reduction (Rio de Janeiro) the median tariff reduction (Alfenasin southwestern Minas Gerais state), and the smallest tariff reduction (actually a small increase,Mata Grande in northwest Alagoas state). The industries on the x-axis are sorted from the mostnegative to the most positive tariff change. Rio de Janeiro has more weight on the left side of thediagram, by virtue of specializing in manufacturing, particularly in apparel and food processingindustries, which faced quite large tariff reductions. Thus, its regional tariff reduction is quitelarge. Alfenas is a coffee growing and processing region, which also has some apparel employment,balancing the large tariff declines in apparel and food processing against the small tariff increasein agriculture. Mata Grande is located in a sparsely populated mountainous region, and is almostexclusively agricultural, leading it to experience a small tariff increase overall. Thus, although allregions faced the same set of tariff reductions across industries, variation in the industry distributionof employment in each region generates substantial variation in RTRr.
51
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure A2: Variation Underlying Regional Tariff Reduction
0.75 0.97
0.00
0.10
0.20
0.30
0.40
0.50
Indu
stry
Wei
ght
Rubb
er
Appa
rel
Oth
er M
anuf
.
Phar
ma.
, Per
fum
es, D
eter
gent
s
Plas
tics
Auto
, Tra
nspo
rt, V
ehicl
es
Nonm
etal
lic M
iner
al M
anuf
Elec
tric,
Ele
ctro
nic
Equi
p.
Food
Pro
cess
ing
Mac
hine
ry, E
quip
men
t
Petro
leum
Refi
ning
Text
iles
Chem
icals
Woo
d, F
urni
ture
, Pea
t
Pape
r, Pu
blish
ing,
Prin
ting
Met
als
Foot
wear
, Lea
ther
Min
eral
Min
ing
Petro
leum
, Gas
, Coa
l
Agric
ultu
re
Industries sorted from most negative to most positive tariff change
Rio de Janeiro, RJ (.15) Alfenas MG (.03) Mata Grande, AL (-.01)
Industry distribution of 1991 employment in the regions facing the largest (Rio de Janeiro, RJ), median (Alfenas,MG) and smallest (Mata Grande, AL) regional tariff reduction. Industries sorted from the most negative to themost positive tariff change (see Figure 1). More weight on the left side of the figure leads to a larger regional tariffreduction, and more weight on the right side leads to a smaller regional tariff reduction.
52
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
A.5 Local Price Indexes
Moretti (2013) calculates local price indexes for the U.S. using the change in monthly rents for 2 or3 bedroom apartments. We adjust this approach to the Brazilian context in a few ways. First, wefocus on 1 or 2 bedroom apartments, which are far more common in the Brazilian setting, accountingfor more than 85 percent of the stock of rental units in 1991 and 2010. Many Brazilian cities includefavelas with somewhat improvised structures, and rural areas often feature less formal dwellings.We restrict the sample to include only units with modern construction materials (masonry or woodframing), with at least one bathroom, and with modern sanitation (sewer or septic tank). Theserestrictions allow us to avoid comparing modern apartments to informal dwellings. Using thissample of apartments, we calculate the change in log average monthly rent in each region. 19very sparsely populated microregions do not have observations for any rental units satisfying thesecharacteristics in either 1991 or 2010, so we have rent indexes for 456 microregions in our sample.
We then need to transform the change in rental prices into a regional price index. Given thecross-sectional nature of our analysis, we only need to be concerned with prices that vary at thelocal level, i.e. nontradables, since tradable goods prices move together across regions, and thusdo not affect this exercise. Using local Consumer Price Indexes produced by the Bureau of LaborStatistics for 23 U.S. metropolitan areas, Moretti (2013) shows that, as expected, local non-housingnontradables’ prices move with local rental prices. He estimates a slope of 0.35 for the effect ofhousing prices on non-housing nontradables’ prices. The Brazilian Consumer Price Index (Indicesde Precos ao Consumidor - IPC) system reports that in 2002-03, housing’s share of consumptionwas 16.24 percent and that the share for other nontradable goods was 39.94 percent (IBGE 2005).Together, these figures imply that the effective weight on housing prices in the consumer price indexis 0.1624 + 0.3994 · 0.35 = 0.3022. Our local price deflator is therefore 0.3022 times the change inlog rental prices in the region.
53
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
A.6 Regional Change in Imports and Exports
Import and export data between Brazil and the rest of the world come from Comtrade, at the6-digit HS level. We map from HS codes to the industries presented in Table A1 and calculate totalBrazilian trade flows by industry and year.
In the main text, we follow Autor et al. (2013) (ADH) by generating regional weighted averagesof changes in imports and exports per worker. For each industry, we calculate the change in imports(Mit) and exports (Xit) from 1990 to each subsequent year t. These trade flows are measured in$100,000 units. We then generate the regional change in imports and exports per worker as follows.
RegImprt =∑i
Lrit0Lrt0
∆1990−tMit
Lit0(14)
RegExprt =∑i
Lrit0Lrt0
∆1990−tXit
Lit0(15)
The rightmost ratios in these expressions are industry-level shocks, measuring the change in importsor exports per worker initially employed in the industry, in year t0 = 1991. The preceding ratiosrepresent industry weights for each region, reflecting industry i’s share of tradable employment inregion r in 1991. These weights are equivalent to λri in (1). We then generate weighted averagesby summing these terms over tradable industries. Finally, we construct regional net exports as thedifference in regional exports and imports.
RegNetExprt = RegExprt −RegImprt
In Appendix B.10, we present an alternative set of results based on the change in log tradeflows rather than the change in trade flows per worker.
RegLnImprt =∑i
Lrit0Lrt0
∆1990−t ln(Mit) (16)
RegLnEmprt =∑i
Lrit0Lrt0
∆1990−t ln(Xit) (17)
We emphasize that this measure is presented only for descriptive purposes and as a statisticalrobustness test since it does not have the same theoretical underpinnings as the measures followingAutor et al. (2013).
A.7 Entry, Exit, Job Creation, and Job Destruction
We calculate cumulative job creation and job destruction following Davis and Haltiwanger (1990).
job creationrt ≡∑
e∈Ert, get>0
xetXrt
get, (18)
job destructionrt ≡∑
e∈Ert, get<0
xetXrt|get|, (19)
54
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
where get ≡Let − Le,1991
xet, xet ≡
1
2(Let + Le,1991), Xrt ≡
∑e∈Ert
Let,
Let is employment at establishment e in year t and Ert is the set of active establishments in regionr in year t. Note that employment growth, get, is calculated from 1991 to year t. The dependentvariables for the regressions underlying Figure 9 are ln(job creationrt) and ln(job destructionrt).
Entry and exit are calculated analogously, replacing establishment employment with an indica-tor for the establishment being active in the relevant year, ιet.
entryrt ≡∑
e∈Ert, get>0
xet
Xrt
get, (20)
exitrt ≡∑
e∈Ert, get<0
xet
Xrt
|get|, (21)
where get ≡ιet − ιe,1991
xet, xet ≡
1
2(ιet + ιe,1991), Xrt ≡
∑e∈Ert
ιet,
These definitions yield intuitive results. entryrt is equivalent to the share of active firms in regionr in year t that were not active in 1991. exitrt is equivalent to the number of firms in region rthat were active in 1991 and not in year t, divided by the number of firms active in year t. Thedependent variables for the regressions underlying Figure 8 are ln(entryrt) and ln(exitrt)
55
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B Supplemental Empirical Results
B.1 Industry-Level Outcome Pre-Trends vs. Tariff Reductions
Along with regional variation in the industrial composition of employment, our analysis relies onvariation in tariff cuts across industries. Here we analyze the relationship between tariff cuts duringliberalization (1990-1995) and trends in industry wages and employment before liberalization, 1980-1991. We calculate these pre-liberalization outcome trends using the Demographic Census, toprovide a longer pre-liberalization period than what is available in RAIS, which starts in 1986.
We implemented a variety of specifications, with results reported in Table B1. In all spec-ifications, the independent variable is the proportional reduction in one plus the tariff rate (-∆1990−95 ln(1 + τi)). In panels A-C the dependent variable is the change in log industry earnings.Panel A uses average log earnings; Panel B uses average log earnings residuals controlling for in-dividual age, sex, education, and formal status; and Panel C uses average log earnings residualscontrolling for these individual characteristics and region fixed effects. In Panel D, the dependentvariable is the change in industry log employment. Column (1) weights industries equally, andpresents standard errors based on pairwise bootstrap of the t-statistic, to improve small sampleproperties with only 20 tradable industry observations. Column (2) uses the same estimator, butdrops agriculture. Column (3) uses heteroskedasticity weights and presents heteroskedasticity-robust standard errors, which are likely understated in this small sample (MacKinnon 2011). Col-umn (4) uses the same estimator, but drops agriculture. In all cases, the results should be seenprimarily as suggestive, because the analysis uses only 19 or 20 observations.
Nearly all of the earnings estimates are positive, indicating larger tariff reductions in indus-tries experiencing more positive wage growth prior to liberalization. The majority of the estimatesare insignificantly different from zero, with the exception of weighted results in Panels A and B.These specifications heavily weight agriculture, which exhibited negative wage growth prior to lib-eralization and experienced essentially no tariff decline during liberalization, driving the strongnegative relationship. By dropping agriculture, Column (4) confirms that the significant rela-tionship is driven by agriculture. The employment estimates are larger, and change sign acrosscolumns. Given the diversity of findings across earnings and employment specifications, this ex-ercise is somewhat inconclusive. Tariff cuts may or may not have been substantially correlatedwith pre-liberalization outcome trends. These findings motivate us to control for pre-liberalizationoutcome trends whenever possible throughout the paper. This ensures that our results are robustto potential spurious correlation between liberalization-induced labor demand shocks and ongoingtrends.
56
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B1: Pre-Liberalization Industry Trends - 1980-1991
unweighted, bootstrapped
unweighted, bootstrapped, omitting
agriculture
weighted weighted, omitting agriculture
1980-1991 change in log: (1) (2) (3) (4)
Panel A: average earningsIndustry tariff reduction 0.345 0.111 1.029*** 0.510
(0.322) (0.354) (0.139) (0.582)Panel B: earnings premia (with individual controls)
Industry tariff reduction 0.203 -0.017 0.610*** -0.235(0.273) (0.311) (0.157) (0.350)
Panel C: earnings premia (with individual and region controls)Industry tariff reduction 0.135 0.044 0.184 0.018
(0.177) (0.209) (0.158) (0.222)
Panel D: employmentIndustry tariff reduction -1.624 -2.696** 0.687 -1.651
(1.272) (1.361) (0.417) (1.894)
Observations 20 19 20 19
Decennial Census data. 20 industry observations (19 omitting agriculture). See text for details of dependent andindependent variable construction. Column (1) weights industries equally, and presents standard errors based onpairwise bootstrap of the t-statistic. Column (2) uses the same estimator as Column (1), but drops agriculture.Column (3) uses heteroskedasticity weights and presents heteroskedasticity-robust standard errors. Column (4) usesthe same estimator as Column (3), but drops agriculture. *** Significant at the 1 percent, ** 5 percent, * 10 percentlevel.
57
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.2 Informal Employment
The following results provide some descriptive evidence on the informal sector in Brazil. Informalityis defined as working without a signed work card (Carteira de Trabalho e Previdencia Social), whichentitles workers to benefits and labor protections afforded them by the legal employment system.Table B2 shows that the overall rate of informality increased from 1991 to 2000, before decreasingsubstantially from 2000 to 2010. Rates of informality are highest in agriculture and much lower inmanufacturing. Table B1 breaks out informality rates in the manufacturing sector into individualindustries. Finally, Table B2 focuses on the year 2000 and shows the industry distribution of formaland informal employment. There is very substantial overlap in the industry distributions of formaland informal employment. The biggest differences occur in agriculture, which comprises a muchlarger share of informal employment, and food processing and metals, which comprise larger sharesof formal employment. In contrast, the nontradable share is nearly identical for formal and informalemployment.
58
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B2: Informal Share of Employment - 1991-2010
1991 2000 2010
Overall 0.58 0.64 0.49
Agriculture 0.89 0.86 0.83Mining 0.61 0.45 0.21Manufacturing 0.28 0.39 0.29Nontradable 0.55 0.64 0.48
Author’s calculations using Brazilian Demographic Census data for workers age 18-64. Informality defined as nothaving a signed work card.
Figure B1: Informal Share of Employment by Industry - 1991-2010
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Rubb
er
Apparel
Other M
anuf.
Pharma., Perfumes, D
etergents
PlasBcs
Auto, Transpo
rt, V
ehicles
Non
metallic M
ineral M
anuf
Electric, Electronic Equip.
Food
Processing
Machine
ry, Equ
ipmen
t
Petroleu
m Refi
ning
TexBles
Chem
icals
Woo
d, Furniture, Peat
Pape
r, Pu
blish
ing, Prin
Bng
Metals
Footwear, Leathe
r
Mineral M
ining
Petroleu
m, G
as, Coal
Agriculture
Non
traded
Inform
al Sha
re of Ind
ustry Em
ploymen
t 1991
2000
2010
Authors’ calculations using Brazilian Demographic Census data for workers age 18-64. Informality defined as nothaving a signed work card. Industries sorted from most negative to most positive tariff change (with the exceptionof the nontraded sector).
59
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B2: Industry Distribution of Formal and Informal Employment - 2000
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
Rubb
er
Apparel
Other M
anuf.
Pharma., Perfumes, D
etergents
Plas8cs
Auto, Transpo
rt, V
ehicles
Non
metallic M
ineral M
anuf
Electric, Electronic Equip.
Food
Processing
Machine
ry, Equ
ipmen
t
Petroleu
m Refi
ning
Tex8les
Chem
icals
Woo
d, Furniture, Peat
Pape
r, Pu
blish
ing, Prin
8ng
Metals
Footwear, Leathe
r
Mineral M
ining
Petroleu
m, G
as, Coal
Agriculture
Non
traded
Indu
stry Sha
re of S
ector E
mploymen
t
formal
informal
0.23 0.70 0.69
Authors’ calculations using year 2000 Brazilian Demographic Census data for workers age 18-64. Informality definedas not having a signed work card. Industries sorted from most negative to most positive tariff change (with theexception of the nontraded sector).
60
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.3 Regional Earnings Premium Regressions
As discussed in Section 4.1, we calculate regional earnings premia by regressing workers’ log De-cember earnings on flexible demographic and educational controls, industry fixed effects, and regionfixed effects, separately in each year. Table B3 shows the coefficient estimates from these earningspremium regressions for 1991, 2000, and 2010. The region fixed effect estimates provide average logearnings for formally employed workers in the region, controlling for the age, sex, education, andindustry composition of the region’s employment. These regional premia then form the outcomevariable in our earnings analyses.
Note that the coefficient estimates on the controls conform to expectations. Women are paidless than otherwise similar men, and this earnings gap declines over time. Workers exhibit aninverted U-shaped wage profile as they age, as is standard in Mincerian regressions. The returnsto education are monotonically positive.
61
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B3: Regional Earnings Premium Regressions - 1991, 2000, 2010
(1) (2) (3)1991 2000 2010
Female -0.349*** -0.295*** -0.261***(0.000) (0.000) (0.000)
Age25-29 0.210*** 0.209*** 0.148***
(0.000) (0.000) (0.000)30-39 0.407*** 0.374*** 0.273***
(0.000) (0.000) (0.000)40-49 0.527*** 0.525*** 0.382***
(0.001) (0.000) (0.000)50-64 0.435*** 0.507*** 0.474***
(0.001) (0.001) (0.000)
Education (years)Primary School Dropout (1-3) 0.015*** 0.007*** 0.130***
(0.001) (0.001) (0.001)Primary School Graduate (4) 0.111*** 0.075*** 0.182***
(0.001) (0.001) (0.001)Middle School Dropout (5-7) 0.188*** 0.129*** 0.206***
(0.001) (0.001) (0.001)Middle School Graduate (8) 0.297*** 0.181*** 0.236***
(0.001) (0.001) (0.001)High School Dropout (9-11) 0.454*** 0.305*** 0.289***
(0.001) (0.001) (0.001)High School Graduate (12) 0.711*** 0.523*** 0.430***
(0.001) (0.001) (0.001)College Dropout (13-15) 0.967*** 0.902*** 0.792***
(0.002) (0.001) (0.001)College Graduate (≥16) 1.374*** 1.384*** 1.368***
(0.001) (0.001) (0.001)
Fixed EffectsIndustry (24) X X XRegion (475) X X X
Observations 13,582,443 17,733,492 30,662,075R-squared 0.858 0.842 0.759
dependent variable: log monthly earnings
Individual worker observations from RAIS. Earnings premium regressions were run for each year from 1986-2010.Here we show three years as examples. The region fixed effect estimates provide average log earnings for formallyemployed workers in the region, controlling for the age, sex, education, and industry composition of the region’semployment. These regional premia then form the outcome variable in our regional earnings analyses. The omittedcategory is a male, age 18-24, with 0 years of education (illiterate). Robust standard errors in parentheses. ***Significant at the 1 percent, ** 5 percent, * 10 percent level.
62
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.4 Formal Earnings Regression Scatterplots
Figure B3 shows scatter plots underlying the formal earnings regression estimates in Figure 3 for1995, 2000, 2005, and 2010. Each marker represents a microregion, and microregions in each majorregion are shown with a separate type of marker. The size of each marker is proportional tothe weight the relevant microregion receives in the estimation. The mean value of the dependentvariable is normalized to zero in each year to focus attention on the slope.
These scatter plots make clear three important points about the earnings estimates. First,as shown in Figure 3, the magnitude of the slope increases substantially and steadily as timepasses following liberalization. Second, the relationship between changes in formal earnings premiaand regional tariff reductions is approximately linear in all time periods, justifying our choice offunctional form. Third, the increasing magnitude slope is driven by shifts in earnings across largenumbers of microregions in various parts of the country, rather than by a few outliers.
63
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Fig
ure
B3:
For
mal
Ear
nin
gsP
rem
iavs.
Reg
ion
alT
ariff
Red
uct
ion
s-
1995
,20
00,
2005,
2010
-1-.50.51
0.05
.1.15
1995
-1-.50.51
0.05
.1.15
2000
-1-.50.51
0.05
.1.15
2005
-1-.50.51
0.05
.1.15
2010
Sca
tter
plo
tsof
regre
ssio
ns
beh
ind
Fig
ure
3es
tim
ate
s.475
mic
rore
gio
nobse
rvati
ons.
Y-a
xis
isth
ech
ange
inlo
gfo
rmal
earn
ings
pre
miu
min
agiv
enre
gio
nfr
om
1991
toth
eyea
rlist
edin
the
titl
e.D
epen
den
tva
riable
dem
eaned
inea
chyea
r.X
-axis
isth
ere
gio
nal
tari
ffre
duct
ion,RTRr.
The
size
of
each
mark
erre
pre
sents
the
inver
seof
the
vari
ance
of
the
dep
enden
tva
riable
esti
mate
,use
dto
wei
ght
the
regre
ssio
ns.
Majo
rre
gio
ns
show
nw
ith
diff
eren
tsy
mb
ols
.South
east
:dark
blu
eci
rcle
s,N
ort
hea
st:
ora
nge
dia
monds,
Cen
ter-
Wes
t:re
dtr
iangle
s,South
:light
blu
eX
s,N
ort
h:
gre
ensq
uare
s.
64
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.5 Census Earnings, Wage, and Employment Results
Table B4 estimates versions of equation (3) using formal sector outcomes calculated using Censusdata. Because the Census includes hours information, we are able to examine both earnings andwage premia. In all cases, we find negative coefficients onRTRr, indicating that regions facing largertariff reductions experienced relative declines in monthly earnings, hourly wages, or employment.We also find substantial growth in the magnitude of these effects, corroborating the results in Table1, which uses RAIS outcomes.
Table B4: Census Regional log Formal Earnings, Wages, and Employment - 2000, 2010
Change in outcome: (1) (2) (3) (4) (5) (6)
Panel A: log Formal Earnings PremiaRegional tariff reduction (RTR) -‐0.397 -‐0.293** -‐0.261** -‐1.384** -‐0.890*** -‐0.855***
(0.335) (0.120) (0.116) (0.572) (0.198) (0.186)Formal earnings pre-‐trend (86-‐90) -‐0.0896* -‐0.0994
(0.0528) (0.0728)State fixed effects (26) ✓ ✓ ✓ ✓
R-‐squared 0.031 0.579 0.583 0.156 0.718 0.720
Panel B: log Formal Wage PremiaRegional tariff reduction (RTR) -‐0.630* -‐0.533*** -‐0.495*** -‐1.320** -‐0.765*** -‐0.721***
(0.355) (0.124) (0.118) (0.525) (0.173) (0.163)Formal earnings pre-‐trend (86-‐90) -‐0.108* -‐0.124*
(0.0555) (0.0642)State fixed effects (26) ✓ ✓ ✓ ✓
R-‐squared 0.071 0.605 0.609 0.136 0.718 0.721
Panel B: log Formal EmploymentRegional tariff reduction (RTR) -‐2.478*** -‐1.756*** -‐1.619*** -‐3.913*** -‐2.865*** -‐2.725***
(0.487) (0.281) (0.258) (0.758) (0.443) (0.417)Formal employment pre-‐trend (86-‐90) 0.211*** 0.227**
(0.0704) (0.0926)State fixed effects (26) ✓ ✓ ✓ ✓
R-‐squared 0.317 0.612 0.630 0.319 0.629 0.638
1991-‐2000 1991-‐2010
Outcomes calculated using Census data. Negative coefficient estimates for the regional tariff reduction imply largerdeclines in formal earnings, wages, or employment in regions facing larger tariff reductions. 475 microregion observa-tions. Regional earnings premia calculated controlling for age, sex, education, and industry of employment. Efficiencyweighted by the inverse of the squared standard error of the estimated outcome. RAIS pre-trends computed for 1986-1990. Standard errors (in parentheses) adjusted for 112 mesoregion clusters. *** Significant at the 1 percent, ** 5percent, * 10 percent level.
65
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.6 Robustness Tests
Table B5 presents the various earnings robustness tests summarized in Section 4.2. For comparison,Panel A shows the main specification, corresponding to the estimates in Table 1 and Figure 3.
All panels in the table control for pre-liberalization earnings growth from 1986-1990, calculatedusing RAIS. Panel B additionally controls for longer pre-liberalization earnings trends calculatedusing the Census. The 1980-1991 control reflects the growth in formal earnings premium, whereformality is defined based on whether the worker’s job included social security contributions. The1970-1980 control reflects growth in the earnings premium for all workers, since there is no formalityinformation in the 1970 Census. See Appendix A.3 for more detail on Census data. A potentialproblem with Panel B is mechanical endogeneity, because the 1980-1991 pre-trend and the 1991-tearnings growth dependent variable both include the 1991 earnings premium. Panel C resolves thisissue by using earnings growth from 1992 to year t as the dependent variable, while including theCensus pre-trend controls.
Panel D calculates the regional tariff reduction (RTRr) in (2) using weights based on the initialindustry distribution of regional formal employment, rather than overall employment. Panel Ecalculates RTRr using effective rates of protection rather than nominal tariffs. Effective rates ofprotection capture the overall effect of liberalization on producers in a given industry, accountingfor tariff changes on industry inputs and outputs. Kume et al. (2003) provide effective rates ofprotection along with the nominal tariffs used in our main analysis. The magnitude of the changesin effective rates of protection is larger than for nominal tariffs, so the coefficients in Panel 3smaller by the same proportion. Since versions of RTRr based on effective rates of protection andnominal tariffs are nearly perfectly correlated (correlation = 0.993), the variation in earnings growthexplained by both versions is nearly identical. Panel F calculates RTRr including the nontradablesector with a tariff change value of 0. This measure ignores the fact that nontradable prices movewith tradable prices (see Appendix A.5 and Panel B of Table 1), and in doing so underestimatesthe magnitude of the average liberalization-induced price change faced by each region. Because themagnitude of RTRr is reduced, the coefficient estimates are inflated by the same proportion.
Panel G omits industry fixed effects when calculating regional earnings premia. This maintainsthe national industry-level variation in earnings in the outcome measure, rather than restrictingattention to the regional equilibrium earnings used in our main specifications. Panel H omits allcontrols from the earnings premium regressions, using simple average log earnings for workers inthe relevant region. While the main analysis weights by the inverse of the squared standard errorof the estimated growth in regional wage premium, Panel I weights all regions equally, and PanelJ weights by 1991 formal employment.
In all cases, the effects grow substantially over time, as in our main specification. In fact, in allbut two of these robustness tests, the long run effect of liberalization on earnings is larger than itis in the main specification.
Table B6 shows that the formal employment results in Table 1 and Figure 4 are similarly robust.The panel labels correspond to Table B5, so see above for descriptions of each specification. Notethat Panels G and H do not apply to employment, since they relate to earnings premia. Panel Ialso does not apply, because the main employment specification is unweighted, since RAIS containsthe population of formally employed workers.
66
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B5: Robustness: Regional log Formal Earnings Premia - 1995, 2000, 2005, 2010
Change in log Formal Earnings Premia: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Main specificationRegional tariff reduction (RTR) -0.096 -0.529*** -1.294*** -1.594***
(0.120) (0.141) (0.139) (0.169)Panel B: Long pre-trends (Census: 1970-80, 1980-91, RAIS: 1986-90)
Regional tariff reduction (RTR) -0.243* -0.770*** -1.498*** -1.814***(0.130) (0.184) (0.186) (0.199)
Panel C: Long pre-trends, earnings growth from 1992 to tRegional tariff reduction (RTR) -0.478*** -1.009*** -1.737*** -2.039***
(0.115) (0.197) (0.195) (0.215)Panel D: RTR using formal employment industry weights
Regional tariff reduction (RTR) -0.089 -0.358 -1.270*** -1.665***(0.189) (0.217) (0.246) (0.298)
Panel E: RTR using effective rates of protectionRegional tariff reduction (RTR) -0.047 -0.328*** -0.823*** -1.017***
(0.076) (0.091) (0.090) (0.107)Panel F: RTR including zero nontradable price change
Regional tariff reduction (RTR) -0.798 -1.758*** -3.350*** -4.625***(0.489) (0.577) (0.643) (0.696)
Panel G: Earnings premium without industry fixed effectsRegional tariff reduction (RTR) 0.131 -0.422*** -1.420*** -1.895***
(0.148) (0.151) (0.163) (0.209)Panel H: Earnings premium with no controls (mean log earnings)
Regional tariff reduction (RTR) 0.317* 0.046 -1.192*** -1.905***(0.171) (0.214) (0.147) (0.182)
Panel I: Unweighted (equally weighted)Regional tariff reduction (RTR) -0.244 -0.490** -1.074*** -1.546***
(0.169) (0.197) (0.215) (0.224)Panel J: Weighted by 1991 formal employment
Regional tariff reduction (RTR) -0.020 -0.345** -1.217*** -1.631***(0.128) (0.147) (0.143) (0.175)
Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal earnings in regionsfacing larger tariff reductions. 475 microregion observations, except Panels B and C, which use a more aggregateregion definition with 405 observations for consistency with 1970 and 1980 Census data. Regional earnings premiacalculated controlling for age, sex, education, and industry of employment except in Panels G and H. Standard errors(in parentheses) adjusted for 112 mesoregion clusters, except Panels B and C with 90 mesoregion clusters. Efficiencyweighted by the inverse of the squared standard error of the estimated change in log formal earnings premium exceptin Panels I and J. See text for detailed description of each panel. *** Significant at the 1 percent, ** 5 percent, * 10percent level.
67
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B6: Robustness: Regional log Formal Employment - 1995, 2000, 2005, 2010
Change in log Formal Employment: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Main specificationRegional tariff reduction (RTR) -1.900*** -3.533*** -4.517*** -4.663***
(0.422) (0.582) (0.685) (0.679)Panel B: Long pre-trends (Census: 1970-80, 1980-91, RAIS: 1986-90)
Regional tariff reduction (RTR) -1.157 -3.393*** -4.687*** -4.537***(0.787) (0.930) (1.019) (1.007)
Panel C: Long pre-trends, earnings growth from 1992 to tRegional tariff reduction (RTR) -0.722 -2.957*** -4.252*** -4.102***
(0.804) (0.972) (1.084) (1.070)Panel D: RTR using formal employment industry weights
Regional tariff reduction (RTR) -1.728*** -2.690*** -4.491*** -4.362***(0.598) (0.793) (0.782) (0.789)
Panel E: RTR using effective rates of protectionRegional tariff reduction (RTR) -1.201*** -2.336*** -2.959*** -3.074***
(0.274) (0.369) (0.438) (0.430)Panel F: RTR including zero nontradable price change
Regional tariff reduction (RTR) -5.677*** -8.574*** -10.874*** -12.507***(1.571) (2.441) (2.752) (2.750)
Panel G: Not applicablePanel H: Not applicablePanel I: Not applicable
Panel J: Weighted by 1991 formal employmentRegional tariff reduction (RTR) -1.195*** -2.119*** -3.406*** -2.842***
(0.195) (0.358) (0.340) (0.397)Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal employmentin regions facing larger tariff reductions. 475 microregion observations, except Panels B and C, which use a moreaggregate region definition with 405 observations for consistency with 1970 and 1980 Census data. Panel labelscorrespond to Table B5, so Panels G and H, which relate to earnings premia, are not applicable here, nor is PanelI, since the main specification is unweighted. Standard errors (in parentheses) adjusted for 112 mesoregion clusters,except Panels B and C with 90 mesoregion clusters. See text for detailed description of each panel. *** Significantat the 1 percent, ** 5 percent, * 10 percent level.
68
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.7 Potential Confounders
B.7.1 Post-Liberalization Tariff Reductions
We calculate post-liberalization regional tariff reductions as in (2), but use tariff reductions between1995 and year t > 1995. Because the Kume et al. (2003) data end in 1998, we use UNCTADTRAINS to construct post-liberalization tariff reductions. The TRAINS data are reported by 6-digit HS codes. In order to maintain as much industry variation as possible, we created an industrymapping from HS codes to Census industry codes, which yields 44 consistently identifiable tradableindustries. This provides more industry detail than the main industry definition in Table A1.The concordance is available upon request. Panel B of Table B7 includes these post-liberalizationtariff reduction controls in the regional earnings growth regression. The post-liberalization controlhas the expected negative coefficient, but its inclusion has very little effect on the liberalizationcoefficient.
B.7.2 Real Exchange Rates
We construct regional real exchange rate shocks as follows. We begin with real exchange ratesbetween Brazil and its trading partners, calculated from Revision 7.1 of the Penn World Tables.We then calculate each country’s 1989 shares of Brazil’s imports and exports in each industryusing Comtrade. As in the prior section, we use the industry definition mapping from HS codes toCensus industries. Industry-specific real exchange rates are weighted averages of country-specificreal exchange rates, weighting either by the 1989 import share or export share. We define industry-level real exchange rate shocks as the change in log industry real exchange rate from 1990 toeach subsequent year. Finally we create regional real exchange rate shocks as weighted averagesof industry real exchange rate shocks, where the region’s industry weights are given by the 1991industry distribution of employment. Panel C of Table B7 includes both the import-weighted andexport-weighted real exchange rate controls. With these controls, the earnings effects grow evenmore than in the main specification.
B.7.3 Privatization
Substantial privatization in Brazil began in 1991 with the administration of President Collor, butsignificantly increased during President Cardoso’s administration (1995-2002). Beginning in 1995,the RAIS data allow us to identify as state-owned any firm at least partly owned by the government.In panels D and E of Table B7, we include different controls for the regional effects of privatization.Panel D includes quartile indicators for the 1995 share of regional employment in state-owned firms,controlling flexibly for the initial share of employment subject to potential privatization. Panel Econtrols for the change in state-owned firm employment share from 1995 to subsequent year t. Inboth cases, the privatization controls have no meaningful effect on the RTRr coefficients.
69
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B7: Potential Confounders: Regional log Formal Earnings Premia - 1995, 2000, 2005, 2010
Change in log Formal Earnings Premia: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Main specificationRegional tariff reduction (RTR) -0.096 -0.529*** -1.294*** -1.594***
(0.120) (0.141) (0.139) (0.169)Panel B: Post-liberalization tariff reductions
Regional tariff reduction (RTR) -0.096 -0.542*** -1.234*** -1.809***(0.120) (0.137) (0.171) (0.218)
Post-liberalization (1995 to t) regional n/a -3.705 -2.415 -2.124tariff reductions (3.273) (3.872) (1.425)
Panel C: Exchange ratesRegional tariff reduction (RTR) -0.113 -0.482*** -1.475*** -1.728***
(0.118) (0.161) (0.184) (0.207)Import-weighted real exchange rate 0.136 0.570* 0.243* 0.374*
(0.133) (0.327) (0.141) (0.201)Export-weighted real exchange rate 0.051 -0.164 0.084 -0.166
(0.160) (0.280) (0.384) (0.355)Panel D: Privatization: initial state-owned employment share
Regional tariff reduction (RTR) -0.090 -0.490*** -1.235*** -1.580***(0.134) (0.163) (0.172) (0.205)
Quartile indicators, 1995 state-owned ✓ ✓ ✓ ✓employment share distirbution
Panel E: Privatization: change in state-owned employment share, 1995 to tRegional tariff reduction (RTR) -0.096 -0.514*** -1.243*** -1.558***
(0.120) (0.149) (0.146) (0.182)Change in state-owned employment share 0.095 0.286 0.176
(0.178) (0.214) (0.227)Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal earnings inregions facing larger tariff reductions. 475 microregion observations. Regional earnings premia calculated controllingfor age, sex, education, and industry of employment. Standard errors (in parentheses) adjusted for 112 mesoregionclusters. Efficiency weighted by the inverse of the squared standard error of the estimated change in log formalearnings premium. See text for detailed description of each panel and for control construction. *** Significant at the1 percent, ** 5 percent, * 10 percent level.
70
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.7.4 Commodity Price Boom
Figure B4 shows price indexes for major commodity classifications from 1991 to 2010, taken fromthe IMF International Financial Statistics. We rescale the series to equal 100 in 1991. All of thecommodity prices were relatively stable through 2003; agricultural prices were within 1.7 percentof their 1991 level, and all other series were below their 1991 levels. In 2004 and later, all of thecommodities saw substantial price growth. Note the contrast between this time series pattern andthe effects of regional tariff reductions on formal earnings and employment growth in Figures 3 and4. The earnings effects grow steadily from 1996 to 2003, in spite of the fact that most commodityprices actually fell a bit during that time span. Commodity prices start growing very quickly in2004 and later, during which the earnings effects start to level off. A similar argument applies tothe employment effects, which grow from 1994 to 2004 and level off subsequently. Thus, the timingof the commodity price boom does not conform with the timing of the earnings and employmenteffects, making it very unlikely that commodity prices drive our results.
To reinforce this time-series evidence, in Table B8 we implement a wide variety of tests torule out the commodity price boom as a potential confounder. Panel A reproduces the mainspecification for comparison. Panels B and C respectively restrict the sample of regions to thosewith below median and bottom quartile employment shares in agriculture and mining, the sectorsaffected by the commodity price boom. Note that mining includes fuel extraction. When focusingon regions with minimal exposure to commodity sectors, we find even larger growth in the effects ofliberalization on earnings than in the entire sample. In Panel D, we maintain all regions, but onlycalculate earnings premia for workers employed in the manufacturing sector, omitting workers incommodity and nontradables sectors. Once again, the earnings effects continue grow substantiallyover time given this restriction. As an aside, note that earnings in the manufacturing sector, whichmost directly experienced the effects of trade liberalization, exhibit significant effects on impact,in 1995. This finding suggests that the very short-run effects of liberalization were concentrated inthe industries facing the largest tariff cuts, but that the earnings effects spread out to other sectorsover time through labor market equilibrium.
In Panel E, rather than restricting the sample of regions or workers, we control for the com-modity price boom directly. We flexibly control for the region’s potential exposure to commodityprice changes by including quartile indicators for the region’s 1991 share of regional employmentin agriculture and mining. We also control directly for regional commodity price shocks, followingthe approach of Adao (2015). Special thanks to Rodrigo Adao for sharing his data and code. SeeAppendix C in Adao (2015) for details on the data source and his equation (16) for the shockconstruction. To summarize, we calculate commodity-specific changes in log price from 1991 toeach subsequent year using data from the Commodity Research Bureau, and construct regionalweighted averages of these commodity price shocks. The weights reflect each commodity sector’sshare of total labor payments in all commodity sectors in the region in 1991. As seen in PanelE, controlling directly for these commodity price movements has little influence on the increasingprofile of earnings effects, which is not surprising given the timing argument above.
The rise of China appears to have played a substantial role in driving up commodity prices inthe late 2000s. As a final test of the commodity price boom hypothesis, we follow Costa et al.(2016), who study the regional effects of import competition from China and Chinese demand forexports. Rather than focusing on commodity prices, Costa et al. study the effects of import andexport quantity shocks, along the lines of Autor et al. (2013). They construct industry-level Chinese
71
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
import supply (IS) and export demand (XD) shocks as the growth in industry imports from orexports to China from 2000 to 2010, divided by Brazilian employment in the industry in 2000.They then generate regional weighted average shocks using the year 2000 industry distribution ofemployment in each region. Finally, they instrument for these shocks using similar measures basedon the growth in Chinese trade to countries other than Brazil. Special thanks to Francisco Costafor providing us with their shock and instrument measures.
Because Costa et al. examine shocks and outcomes between 2000 and 2010, in Panel A of TableB9 we provide our baseline earnings estimate for this time period, with a base year of 2000 ratherthan 1991. We use a slightly more aggregate region definition to match theirs, yielding 405 regionobservations. The coefficient estimate of -1.068 in column (1) is nearly identical to the differencebetween the estimates for 2000 and 2010 in columns (3) and (6), respectively, of Panel A in Table1 of -1.065. The slight difference results from the difference in region definitions between the twotables. In columns (2)-(4) of Table B9, we introduce the Chinese import supply and export demandshocks, instrumented following Costa et al.. The two shocks have the expected sign, with increasedimport competition lowering regional earnings and increased export demand increasing them (veryslightly), though only the import supply shock is statistically different from zero. This result mightseem surprising, given that Costa et al. find significant effects of export demand on wages. However,when they control for the regional composition of workers and for outcome pre-trends, as we dohere, the export result disappears (see their Table 2, Panel B, column (5) in their paper). Whenwe include these controls, they have only a very small effect on our coefficient of interest, furtherconfirming that the divergence in earnings growth between regions facing larger and smaller tariffreductions was not driven by China’s effects on commodity markets.
72
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B4: Commodity Prices - 1991-2010
50
100
150
200
250
300
350
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Price Inde
x (1991 = 100)
Food Agricultural raw materials Metals Non-‐Fuel Commodi@es Food and Beverages
Commodity price series from IMF International Financial Statistics in US dollars, rescaled to equal 100 in 1991.
73
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B8: Commodity Price Boom: Regional log Formal Earnings Premia - 1995, 2000, 2005, 2010
Change in log Formal Earnings Premia: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Main specificationRegional tariff reduction (RTR) -0.096 -0.529*** -1.294*** -1.594***
(0.120) (0.141) (0.139) (0.169)Panel B: Below-median agriculture/mining employment share (238 obs)
Regional tariff reduction (RTR) -0.017 -0.534*** -1.424*** -1.829***(0.148) (0.165) (0.163) (0.213)
Panel C: Bottom quartile agriculture/mining employment share 118 obs)Regional tariff reduction (RTR) 0.006 -0.347 -1.379*** -2.153***
(0.267) (0.262) (0.280) (0.373)Panel D: Manufacturing sector earnings
Regional tariff reduction (RTR) -0.501*** -0.965*** -1.878*** -2.252***(0.158) (0.192) (0.214) (0.262)
Panel E: Direct commodity price controls per Adao (2015)Regional tariff reduction (RTR) -0.052 -0.290 -1.269*** -1.926***
(0.259) (0.257) (0.276) (0.372)Regional commodity price shocks 0.033 -0.039 0.118 0.045
(Adao 2015) (0.207) (0.167) (0.092) (0.127)Quartile indicators, 1991 agriculture/mining ✓ ✓ ✓ ✓
employment share distribution
Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal earnings in regionsfacing larger tariff reductions. 475 microregion observations unless otherwise noted (Panels B and C). Regionalearnings premia calculated controlling for age, sex, education, and industry of employment. Standard errors (inparentheses) adjusted for 112 mesoregion clusters. Efficiency weighted by the inverse of the squared standard error ofthe estimated change in log formal earnings premium. See text for detailed description of each panel and for controlconstruction. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
74
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B9: Regional log Formal Earnings Premia with Costa et al (2015) controls
Change in log Formal Earnings Premia, 2000-2010: (1) (2) (3) (4)OLS IV IV IV
Regional tariff reduction (RTR) -1.068*** -0.929*** -1.069*** -0.931***(0.111) (0.123) (0.107) (0.122)
Formal earnings pre-trend (86-90) -0.077 -0.064 -0.076 -0.063(0.053) (0.055) (0.051) (0.055)
China import supply (Costa et al. 2015) -0.034*** -0.034***(0.010) (0.010)
China export demand (Costa et al. 2015) 0.001 0.001(0.002) (0.002)
State fixed effects (26) ✓ ✓ ✓ ✓
R-squared 0.733 0.738 0.733 0.7391st stage F (Kleibergen-Paap) 22.16 441.6 11.31
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal earnings inregions facing larger tariff reductions. See text for description of China import supply and export demand controlsand associated instruments from Costa et al. (2015). First stage partial F-statistics reported in brackets. 405microregion observations. Regional earnings premia calculated controlling for age, sex, education, and industry ofemployment. Standard errors (in parentheses) adjusted for 90 mesoregion clusters. Efficiency weighted by the inverseof the squared standard error of the estimated change in log formal earnings premium. *** Significant at the 1percent, ** 5 percent, * 10 percent level.
75
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.8 Earnings and Employment Sample Splits
Tables B10 and B11 present earnings results splitting the sample of workers into those employed inthe tradable sector (Panel B), those employed in the nontradable sector (Panel C), more educated(Panel D), and less educated (Panel E). Note that earnings and employment effects grow for allsubsamples. The employment effects are concentrated in the tradable sector and among less skilledworkers, though panels D and E in both tables should be interpreted with care, as the regionaltariff reduction shocks are derived from a model with a single type of labor. For a more generalmodel with two skill types, see Dix-Carneiro and Kovak (2015a).
Table B10: Sample Splits: Regional log Formal Earnings Premia - 1995, 2000, 2005, 2010
Change in log Formal Earnings Premia: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Full sampleRegional tariff reduction (RTR) -0.096 -0.529*** -1.294*** -1.594***
(0.120) (0.141) (0.139) (0.169)Panel B: Tradable sector workers
Regional tariff reduction (RTR) -0.287* -0.754*** -1.623*** -1.934***(0.149) (0.184) (0.203) (0.254)
Panel C: Nontradable sector workersRegional tariff reduction (RTR) -0.060 -0.389** -1.143*** -1.401***
(0.150) (0.179) (0.169) (0.183)Panel D: More educated workers (high school or more)
Regional tariff reduction (RTR) 0.310* -0.539*** -1.611*** -2.053***(0.160) (0.173) (0.192) (0.224)
Panel E: Less educated workers (less than high school)Regional tariff reduction (RTR) -0.218* -0.626*** -1.354*** -1.758***
(0.121) (0.143) (0.137) (0.180)Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal earnings inregions facing larger tariff reductions. 475 microregion observations. Regional earnings premia calculated controllingfor age, sex, education, and industry of employment. Standard errors (in parentheses) adjusted for 112 mesoregionclusters. Efficiency weighted by the inverse of the squared standard error of the estimated change in log formalearnings premium. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
76
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B11: Sample Splits: Regional log Formal Employment - 1995, 2000, 2005, 2010
Change in log Formal Employment: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Full sampleRegional tariff reduction (RTR) -1.900*** -3.533*** -4.517*** -4.663***
(0.422) (0.582) (0.685) (0.679)Panel B: Tradable sector workers
Regional tariff reduction (RTR) -5.790*** -8.416*** -10.097*** -10.156***(0.850) (0.993) (1.101) (1.140)
Panel C: Nontradable sector workersRegional tariff reduction (RTR) 0.726 -0.733 -1.500* -1.600**
(0.455) (0.664) (0.778) (0.749)Panel D: More educated workers (high school or more)
Regional tariff reduction (RTR) 0.141 -1.219* -1.637** -2.195***(0.450) (0.644) (0.772) (0.733)
Panel E: Less educated workers (less than high school)Regional tariff reduction (RTR) -2.507*** -4.556*** -6.328*** -6.910***
(0.465) (0.626) (0.770) (0.787)Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal employment inregions facing larger tariff reductions. 475 microregion observations. Standard errors (in parentheses) adjusted for112 mesoregion clusters. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
77
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.9 Earnings Premia with Individual Worker Fixed Effects
In order to calculate the earnings premia in Panels B and C of Table 3, we pool data acrossyears and include worker-specific fixed effects. To make the estimation procedure computationallyfeasible, we begin by drawing a 3 percent sample of all valid individual IDs that appear in RAISwith a positive earnings observation between 1986 and 2010. This sampling method yields 450microregions in which we observe formally employed workers earning labor income in December forall years in our sample.
The specification in Panel B of Table 3 is a straightforward fixed effects regression.
ln(earnjairt) = αj + ψa + φit + µrt + εjairt,
where αj are worker fixed effects; ψa are age effects (indicators for falling within each age binshown in Table B3); φit are time varying industry effects; and µrt are time varying region effects.These region × year fixed effect estimates (µrt) represent the regional log earnings premia used togenerate the dependent variable in Panel B.
While the preceding specification has the benefit of controlling for time-invariant worker charac-teristics, including unobservables, it restricts the returns to worker characteristics to be fixed overtime. Table B3 found substantial changes in the returns to observable characteristics, contradictingthis restriction. The earnings premia in Panel C of Table 3 relax this assumption by estimating thefollowing specification, which allows the returns to worker characteristics (δt) to vary over time.
ln(earnjairt) = δtαj + ψa + φit + µrt + εjairt.
This specification is difficult to estimate because the parameter space is high dimensional, andthe model is nonlinear in the parameters δt and αj . We estimate this model using the iterativealgorithm developed by Arcidiacono, Foster, Goodpaster and Kinsler (2012) and calculate standarderrors using the wild bootstrap method suggested by Davidson and MacKinnon (2006), with 500bootstrap iterations. Once again, we use the resulting region × year fixed effect estimates (µrt) togenerate the dependent variable in Panel C.
78
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.10 Regional Change in log Imports and Exports
This section presents versions of Figure 5 and Table 4 using alternative trade quantity measuresreflecting regional weighted averages of the change in log imports or exports rather than the changein imports or exports per worker. These alternative measures are presented only for descriptivepurposes and as a statistical robustness test.
We use the change in trade per worker in the main text both because it is theoretically justified,as shown by Autor et al. (2013), and because it intuitively captures the effects of changing tradequantities on local labor market outcomes. Figures B5 and B6, which show scatter plots relatingthe industry-level change in trade per worker and the change in log trade for imports and exportsin 2000 and 2010. Markers are proportional to 1991 employment, and industry labels correspondto Table A1. These plots show that the change in log trade often deviates substantially from thechange in trade per worker. This deviation occurs primarily in industries with relatively small tradeflows and relatively large employment. As an example, consider the Wood, Furniture, and Peatindustry (code 14) in Panel A of Figure B5. In this industry, Brazil imported R$71 million in 1990and R$249 million in 2000 (all values in year 2005 Reais). This very large proportional growthin imports corresponds to the large value for the change in log imports of 1.25. However, initialemployment in this industry was also quite large, 822,579, so the change in imports per worker wasonly R$216, much smaller than the values in the thousands or tens of thousands in other industries.Therefore, although the amount of imports increased very much in proportional terms, it was stillinsignificant compared to the number of workers in the industry. The change in trade per workercaptures the relative scale of trade and employment, while the change in log trade does not.
Nonetheless, figure B7 shows the relationships between RTRr and the regional change in logimports and exports (paralleling Figure 5). See (16) and (17) in Appendix A.6 for details onconstructing the change in log trade measures. Regions facing larger tariff reductions experiencelarger increases in log imports and larger declines in log exports. The magnitude of each effect growsover time, suggesting that perhaps slow trade quantity responses could explain the slow growth ofregional earnings and employment effects in Figures 3. We demonstrate that this is not the caseby directly controlling for the regional change in log import and export measures when estimatingthe effect of RTRr on regional earnings growth. The specifications in Table B12 parallel those inTable 4, using the alternative change-in-log measures of trade flows for both the controls and theinstruments. In all cases, the earnings effects of liberalization grow even more when controlling forimport and export quantity growth than in the baseline specification in Panel A. The relevant Stockand Yogo (2005) critical value for the first-stage F-statistic is 21, so Panel C exhibits a potentialweak instruments problem. We therefore present two additional sets of results in Table B13, usingthe change in trade per worker measure when calculating the instruments rather than the changein log trade flow measure. The weak instrument issue is not longer present, and the effects of RTRron regional earnings still increase substantially over time.
As with the standard trade-per-worker trade flow measures considered in the main text, theanalysis using the alternative change-in-log trade flow measure rules out slow import or exportresponses as the mechanism driving the slowly growing earnings effects. This is to be expectedgiven the discussion at the beginning of this section, since the change in log trade measure doesnot well capture the effects of changing trade flows on workers’ labor market outcomes.
79
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B5: Change in Trade Per Worker vs. Change in log Trade - 1990-2000
Panel A: Imports
12
4 5
810
12
1415
17
21
222324 25
320
1000
020
000
3000
040
000
chan
ge in
impo
rts p
er w
orke
r, 19
90-2
000
-.5 0 .5 1 1.5change in log imports, 1990-2000
Omits the 4 smallest industries, with less than 150,000 employees in 1991.
Panel B: Exports
12
4
5
810
12
1415
17
21
22 23
24
25
32
050
0010
000
1500
020
000
chan
ge in
exp
orts
per
wor
ker,
1990
-200
0
-.5 0 .5 1 1.5change in log exports, 1990-2000
Omits the 4 smallest industries, with less than 150,000 employees in 1991.
Each point is an industry, with labels corresponding to Table A1. The y-axis measures the change in trade perworker initially employed in the industry (in 1991) and the x-axis measures the change in log trade. The figure omitsthe 4 smallest industries in terms of 1991 employment, which often fall well outside the scale shown. Because theyare small, they receive very little weight in the regional analysis that forms this paper’s main analysis. Marker sizeproportional to 1991 employment.
80
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B6: Change in Trade Per Worker vs. Change in log Trade - 1990-2010
Panel A: Imports
12 4
5
8
10
12
1415
17
21
222324 25
320
2000
040
000
6000
080
000
1000
00ch
ange
in im
ports
per
wor
ker,
1990
-201
0
.5 1 1.5 2 2.5change in log imports, 1990-2010
Omits the 4 smallest industries, with less than 150,000 employees in 1991.
Panel B: Exports
1
2
4
5
8
10
12
14
15
17
21
222324
25
32
020
000
4000
060
000
chan
ge in
exp
orts
per
wor
ker,
1990
-201
0
-1 0 1 2change in log exports, 1990-2010
Omits the 4 smallest industries, with less than 150,000 employees in 1991.
Each point is an industry, with labels corresponding to Table A1. The y-axis measures the change in trade perworker initially employed in the industry (in 1991) and the x-axis measures the change in log trade. The figure omitsthe 4 smallest industries in terms of 1991 employment, which often fall well outside the scale shown. Because theyare small, they receive very little weight in the regional analysis that forms this paper’s main analysis. Marker sizeproportional to 1991 employment.
81
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B7: Regional Imports and Exports, Change in log Measure - 1992-2010
-‐20
-‐15
-‐10
-‐5
0
5
10
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Imports
Exports
Liberaliza5on Post-‐liberaliza5on (chg. from 1991)
Each point reflects an individual regression coefficient, θt, following (3), where the dependent variable is the change inregional imports (blue circles) or exports using the change in log measures described in (16) and (17) in Appendix A.6.The independent variable is the regional tariff reduction (RTR), defined in (2). Note that the RTR always reflectstariff reductions from 1990-1995. All regressions include state fixed effects, but do not include pre-liberalizationtrends due to a lack of Comtrade trade data before 1989. Positive (negative) estimates imply larger increases in tradeflow in regions facing larger (smaller) tariff reductions. Vertical bar indicates that liberalization was complete by1995. Dashed lines show 95 percent confidence intervals. Standard errors adjusted for 112 mesoregion clusters.
82
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B12: Slow Response of Imports or Exports, Change in log Measure - 1995, 2000, 2005, 2010(Part 1 of 2)
Change in log Formal Earnings Premia: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel A: Main specificationRegional tariff reduction (RTR) -0.096 -0.529*** -1.294*** -1.594***
(0.120) (0.141) (0.139) (0.169)
Panel B: Controls for trade quantity shocks (OLS)Regional tariff reduction (RTR) -0.142 -0.581*** -1.476*** -1.887***
(0.124) (0.142) (0.151) (0.212)Import quantity shock (change in log)
Export quantity shock (change in log)
Panel C: Latin America IVRegional tariff reduction (RTR) -0.236 -0.608*** -1.479*** -2.158***
(0.145) (0.150) (0.343) (0.579)Import quantity shock (change in log)
Export quantity shock (change in log)
First-‐stage F (Kleibergen-‐Paap)
Panel D: Colombia IV Regional tariff reduction (RTR) -0.227 -0.570*** -1.316*** -1.985***
(0.146) (0.160) (0.306) (0.509)Import quantity shock (change in log)
Export quantity shock (change in log)
First-‐stage F (Kleibergen-‐Paap)
Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
(0.032)61.45
-0.052(0.039)19.71
-0.044(0.049)-0.059*
0.025**(0.012)-0.009(0.008)
-0.009(0.062)
Negative coefficient estimates for the regional tariff reduction (RTR) imply larger declines in formal earnings inregions facing larger tariff reductions. Panel A replicates the earnings results in columns (3) and (6) of Table 1 andin Figure 3. Panels B-D control for regional import and export quantity shocks, calculated using the change in logtrade flows; see Appendix A.6 for details. These panels stack the data across years, allowing the effect of RTRr tovary over time but fixing the import and export quantity coefficients over time. We instrument for the potentiallyendogenous import and export shocks using regional measures of commodity price growth from Adao (2015) andwith regional trade flows for other countries. We consider the combination of Argentina, Chile, Colombia, Paraguay,Peru, and Uruguay (“Latin America”) Colombia alone. In each case, we measure imports and exports between thesecountries and the rest of the world, excluding Brazil. This gives us 2 endogenous variables and 57 instruments (= 3instruments × 19 years). First-stage Kleinbergen-Paap F statistics are shown, for comparison to the Stock and Yogo(2005) critical value of 21 to reject 5 percent bias relative to OLS. Standard errors (in parentheses) adjusted for 112mesoregion clusters. Efficiency weighted by the inverse of the squared standard error of the estimated change in logformal earnings premium. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
83
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Table B13: Slow Response of Imports or Exports, Change in log Measure - 1995, 2000, 2005, 2010(Part 2 of 2)
Change in log Formal Earnings Premia: 1991-1995 1991-2000 1991-2005 1991-2010(1) (2) (3) (4)
Panel E: Latin America IV (change in trade per worker)Regional tariff reduction (RTR) -0.063 -0.487*** -1.143*** -1.364**
(0.147) (0.156) (0.333) (0.561)Import quantity shock (change in log)
Export quantity shock (change in log)
First-‐stage F (Kleibergen-‐Paap)
Panel F: Colombia IV (change in trade per worker)Regional tariff reduction (RTR) -0.096 -0.529*** -1.294*** -1.594***
(0.120) (0.141) (0.139) (0.169)Import quantity shock (change in log)
Export quantity shock (change in log)
First-‐stage F (Kleibergen-‐Paap)
Formal earnings pre-trend (86-90) ✓ ✓ ✓ ✓State fixed effects (26) ✓ ✓ ✓ ✓
(0.055)-0.026(0.026)61.58
-0.022(0.050)0.005
(0.031)31.44
-0.045
See Table B12 for general notes. Because the instrument in Panel C of Table B12 was marginally weak, Panels Eand F present versions using instruments based on the change in imports per worker, while the trade quantity shocksare calculated using the change in log. Both specifications reject the weak instrument concern.
84
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.11 Overall Employment
Table B14 shows the effect of liberalization on overall regional employment, including both formallyand informally employed workers. We use Census data to capture informally employed individuals,and control for 1980-1991 outcome pre-trends. These pre-trends may be subject to mechanicalendogeneity, since both the pre-trends and the dependent variables include the 1991 outcome level,so we instrument for the 1980-1991 pre-trends using the 1980 dependent variable level (columns (2)and (5)) or the 1980-1970 change (columns (3) and (6)). The estimates vary substantially acrossspecifications and all but one are insignificantly different from zero. These results provide littleevidence in favor of overall employment as a potential source of agglomeration economies.
Table B14: Regional log Overall Employment - 2000, 2010
Change in log Overall Employment: OLS IV IV OLS IV IV
1980 level 1970-80 chg. 1980 level 1970-80 chg.(1) (2) (3) (4) (5) (6)
Regional tariff reduction (RTR) 0.203 -0.007 -0.096 0.657** 0.301 0.051(0.209) (0.136) (0.192) (0.314) (0.220) (0.342)
Employment pre-trend (80-91) 0.329** 0.531*** 0.616*** 0.538*** 0.885*** 1.130***(0.136) (0.076) (0.133) (0.202) (0.155) (0.272)
State fixed effects (26) ✓ ✓ ✓ ✓ ✓ ✓
R-squared 0.563 0.499 0.433 0.574 0.491 0.3311st stage F (Kleibergen-Paap) 28.04 8.451 26.04 8.064
1991-2000 1991-2010
Positive (negative) coefficient estimates for the regional tariff reduction imply larger increases (decreases) in overallemployment in regions facing larger tariff reductions. Outcomes calculated using Census data. 405 microregionobservations. Efficiency weighted by the inverse of the squared standard error of the dependent variable estimate.Pre-trends computed for 1980-1991. To address potential mechanical endogeneity due to pre-trend and dependentvairable overlap in 1991, columns (2) and (5) use the 1980 dependent variable level as an instrument for the pre-trendand columns (3) and (6) use the 1970-1980 change as an instrument. Standard errors (in parentheses) adjusted for112 mesoregion clusters. *** Significant at the 1 percent, ** 5 percent, * 10 percent level.
85
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.12 Capital Adjustment Confidence Intervals
Figures B8 - B10 show the capital adjustment profiles in Figure 6, including 95-percent confidenceintervals, which were omitted from Figure 6 for clarity.
86
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B8: Capital Adjustment Quantification - Low ζ - 1992-2010
-‐3.0
-‐2.5
-‐2.0
-‐1.5
-‐1.0
-‐0.5
0.0
0.5
1.0
1.5
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Pre-‐liberaliza6on (chg. from 1986) Liberaliza6on Post-‐liberaliza6on
(chg. from 1991)
Capital (establishments) adjustment, ζ = 0.152
Each point reflects an individual regression coefficient, θt, following (3). The dependent variable is capital’s con-tribution to overall adjustment, using the number of regional formal establishments as a proxy for regional capital.This figure shows the profile using the low estimate of ζ = 0.152. The independent variable is the regional tariffreduction (RTR), defined in (2). Note that the RTR always reflects tariff reductions from 1990-1995. All regressionsinclude state fixed effects, and post-liberalization regressions control for the 1986-1990 outcome pre-trend. Negativeestimates imply larger declines in the number of establishments in regions facing larger tariff reductions. Verticalbar indicates that liberalization was complete by 1995. Dashed lines show 95 percent confidence intervals. Standarderrors adjusted for 112 mesoregion clusters.
87
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B9: Capital Adjustment Quantification - Mid ζ - 1992-2010
-‐3.0
-‐2.5
-‐2.0
-‐1.5
-‐1.0
-‐0.5
0.0
0.5
1.0
1.5
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Pre-‐liberaliza6on (chg. from 1986) Liberaliza6on Post-‐liberaliza6on
(chg. from 1991)
Capital (establishments) adjustment, ζ = 0.349
Each point reflects an individual regression coefficient, θt, following (3). The dependent variable is capital’s con-tribution to overall adjustment, using the number of regional formal establishments as a proxy for regional capital.This figure shows the profile using the low estimate of ζ = 0.349. The independent variable is the regional tariffreduction (RTR), defined in (2). Note that the RTR always reflects tariff reductions from 1990-1995. All regressionsinclude state fixed effects, and post-liberalization regressions control for the 1986-1990 outcome pre-trend. Negativeestimates imply larger declines in the number of establishments in regions facing larger tariff reductions. Verticalbar indicates that liberalization was complete by 1995. Dashed lines show 95 percent confidence intervals. Standarderrors adjusted for 112 mesoregion clusters.
88
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B10: Capital Adjustment Quantification - High ζ - 1992-2010
-‐3.0
-‐2.5
-‐2.0
-‐1.5
-‐1.0
-‐0.5
0.0
0.5
1.0
1.5
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Pre-‐liberaliza6on (chg. from 1986) Liberaliza6on Post-‐liberaliza6on
(chg. from 1991)
Capital (establishments) adjustment, ζ = 0.545
Each point reflects an individual regression coefficient, θt, following (3). The dependent variable is capital’s con-tribution to overall adjustment, using the number of regional formal establishments as a proxy for regional capital.This figure shows the profile using the low estimate of ζ = 0.545. The independent variable is the regional tariffreduction (RTR), defined in (2). Note that the RTR always reflects tariff reductions from 1990-1995. All regressionsinclude state fixed effects, and post-liberalization regressions control for the 1986-1990 outcome pre-trend. Negativeestimates imply larger declines in the number of establishments in regions facing larger tariff reductions. Verticalbar indicates that liberalization was complete by 1995. Dashed lines show 95 percent confidence intervals. Standarderrors adjusted for 112 mesoregion clusters.
89
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
B.13 Exit by Establishment Size
Here we examine the relationship between establishment exit and RTRr, separately by initialestablishment size. We run the following specification at the establishment-year level, using thesample of all active establishments in 1991.
Exitirt =6∑
k=1
βkt Sizeki ·RTRr +
6∑k=1
φkSizeki + γtNT i + ϑtPreExit
1986−1990r + εirt (22)
where Sizeki is an indicator for whether establishment i fell into size bin k in 1991, NT i is anindicator for establishments in the nontradable sector, and PreExit1986−1990r is a pre-trend controlfor the share of regional establishments in 1986 that shut down between 1986 and 1990.
Figure B11 plots the βkt coefficients, with the relevant initial employment bin definitions shownon the right side. Although there is some variation across establishment sizes, with more exit amonglarger establishments than smaller establishments, it is clear that exit rates increased throughoutthe size distribution for establishments whose regions faced larger tariff declines.
90
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
Figure B11: Regional log Cumulative Formal Establishment Exit, by Establishment Size - 1992-2010
-‐0.1
0.1
0.3
0.5
0.7
0.9
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Liberaliza6on Post-‐liberaliza6on (chg. from 1991)
1-‐4
5-‐9
10-‐19
100+
20-‐49
50-‐99
Plots the βkt coefficients in (22), estimated using the sample of all active establishments in 1991. The size range,indexed by k, is reported at the right side of each profile. Vertical bar indicates that liberalization was complete by1995.
91
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
C Model
C.1 Baseline Model
This section generalizes the specific-factors model of regional economies from Kovak (2013) to allowfor changes in regional productivity (agglomeration economies) and changes in labor and capitalinputs.
The economy consists of many regions, indexed by r, which may produce goods in many in-dustries, indexed by i. Production in each industry uses Cobb-Douglas technology with constantreturns to scale and three inputs: labor, a fixed industry-specific factor, and capital. Labor, Lr,is assumed to be perfectly mobile between industries within a region. The industry-specific factor,Tri, is usable only in its respective region and industry and is fixed over time. Capital, Kri, isusable only in its respective region and industry but may change over time. Output of industry iin region r is
Yri = AriL1−ϕiri
(T ζiriK
1−ζiri
)ϕi, (C1)
where ϕi, ζi ∈ (0, 1). To allow for the possibility of agglomeration economies and factor adjustment,we allow Ari, Lr, and Kri to change over time. Goods and factor markets are perfectly competitive,and producers face exogenous prices Pi, common across regions and fixed by world prices and tariffs.
Consider a particular region r, and suppress the region subscript. Let aLi, aT i, and aKi be therespective amounts of labor, specific factor, and capital used in producing one unit of Yi. Regionalfactor market clearing implies ∑
i
aLiYi = L, (C2)
aT iYi = Ti ∀i, (C3)
aKiYi = Ki ∀i. (C4)
Perfect competition implies that the price equals factor payments,
aLiw + aT isi + aKiRi = Pi ∀i, (C5)
where w is the wage, si is the specific-factor price, and Ri is the price of capital. Define x as theproportional change in x, and differentiate (C5).
(1− ϕi)w + ϕiζisi + ϕi(1− ζi)Ri = Pi + Ai ∀i, (C6)
which uses the fact that, from cost minimization,
(1− ϕi)aLi + ϕiζiaT i + ϕi(1− ζi)aKi = −Ai ∀i. (C7)
Differentiate the factor market clearing conditions.∑i
λi(aLi + Yi) = L, (C8)
Yi = −aT i ∀i, (C9)
Yi = Ki − aKi ∀i, (C10)
92
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
where λi ≡ LiL is the share of regional labor allocated to industry i, and we use the fact that Ti = 0.
With Cobb-Douglas production, the elasticity of substitution is one, so
aKi − aT i = si − Ri, (C11)
aLi − aKi = Ri − w, (C12)
Combining (C8), (C10), and (C12) yields∑i
λiRi − w = L−∑i
λiKi. (C13)
Combine (C9), (C10), and (C11) to yield
si = Ri + Ki ∀i. (C14)
Plug this into (C6) and simplify.
Ri =Pi + Ai − (1− ϕi)w − ϕiζiKi
ϕi(C15)
Finally, plug this into (C13), solve for w, and restore regional subscripts to yield the equilibriumrelationship for regional wage changes, equation (7) in the main text.
wr =∑i
βriPi +∑i
βriAri − δr
(Lr −
∑i
λri(1− ζi)Kri
)(C16)
where βri ≡λri
1ϕi∑
j λrj1ϕj
and δr ≡1∑
j λrj1ϕj
.
C.2 Agglomeration Economies
As discussed in the main text, when examining agglomeration economies and quantifying the long-run effects of slow capital adjustment and agglomeration, we assume perfectly mobile capital in thelong run (Rr = R ∀r), and identical technology across industries (ϕi = ϕ ∀i and ζi = ζ ∀i). Theassumption of perfectly mobile capital allows us to substitute out the change in capital, Kri for thechange in its price, R, which is constant across industries and regions.
Start with the labor market clearing condition in (C8), substitute in the specific-factors clearingcondition in (C9) and the Cobb-Douglas conditions in (C11) and (C12) to yield∑
i
λisi − w = L. (C17)
Rearrange the zero-profit condition in (C6) to solve for si,
si =1
ϕζ(Pi + Ai)−
ϕ(1− ζ)
ϕζR− 1− ϕ
ϕζw, (C18)
93
Trade Liberalization and Regional Dynamics Dix-Carneiro and Kovak
and plug it into (C17). Solving for w and restoring regional subscripts yields the following expres-sion.
wr =1
1− ϕ(1− ζ)
∑i
βri(Pi + Ari)−ϕζ
1− ϕ(1− ζ)Lr −
ϕ(1− ζ)
1− ϕ(1− ζ)R (C19)
where βri ≡λri
1ϕ∑
j λrj1ϕ
= λLri
To incorporate agglomeration economies, we assume a constant elasticity agglomeration func-tion, (9), and a constant labor supply elasticity, (10). Substituting these into (C19) and simplifyingyields the following expression for the regional wage change, equation (11) in the main text, whichwe use to estimate the agglomeration elasticity, κ.
wr =η
η[1− ϕ(1− ζ)]− κ+ ϕζ
∑i
βriPi −ϕ(1− ζ)η
η[1− ϕ(1− ζ)]− κ+ ϕζR (C20)
We also use an alternative employment-based approach to estimate κ. Start by noting thatemployment in a region × industry pair is given by Lri ≡ aLriYri. Differentiating this definition andplugging in the specific-factor market clearing condition, (C9), and the Cobb-Douglas substitutionconditions, (C11) and (C12), we have
Li = si − w. (C21)
Substitute in si from (C18) and simplify.
Li =1
ϕζ(Pi + Ai)−
1− ϕ(1− ζ)
ϕζw − ϕ(1− ζ)
ϕζR (C22)
Plug in the labor supply and agglomeration equations, (10) and (9).
Li =1
ϕζPi −
η[1− ϕ(1− ζ)]− κηϕζ
w − ϕ(1− ζ)
ϕζR (C23)
Finally, plug in the equilibrium wage change in (C20), combine terms, and restore regional sub-scripts to yield equation (12) in the main text.
Lri =1
ϕζPi −
1
ϕζ· η[1− ϕ(1− ζ)]− κη[1− ϕ(1− ζ)]− κ+ ϕζ
∑i
βriPi −ϕ(1− ζ)
η[1− ϕ(1− ζ)]− κ+ ϕζR (C24)
94