Trade Liberalization and Regional Dynamics · Contributions I Literature nding substantially di...
Transcript of Trade Liberalization and Regional Dynamics · Contributions I Literature nding substantially di...
Trade Liberalization and Regional Dynamics
Rafael Dix-Carneiro1 Brian K. Kovak2
1Duke UniversityNBER and BREAD
2Carnegie Mellon University - Heinz CollegeNBER and IZA
Conference on Inequality, Globalization and MacroeconomicsUniversity of Southern California
April 29, 2017
Regional Effects of Trade
I Freer trade generates overall economic gains but is also likelyto create winners and losers (Stolper and Samuelson 1941)
I 60 years of work examining trade’s effects on workers withdifferent skills or in different industries
I Starting late 2000s: trade has vastly different effects onworkers in different local labor markets
I Effects determined by regional industry mix
I Trade’s costs and benefits are unevenly distributedgeographically, not just across industries or skills
Slow Adjustment to Trade
I No / Perfect mobility in trade theory: short- / long-runmodels
I Policy makers: what happens in the short- to medium-run?How do we get to the long run?
I Recent work on transitional dynamics of trade liberalizationI Kambourov (2009), Artuc, Chaudhuri and Mclaren (2010), Cosar
(2013), Dix-Carneiro (2014), Caliendo, Dvorkin and Parro (2015),
Helpman and Itskhoki (2015), Traiberman (2016), Bellon (2016),
and many others.I Structural estimation/calibration followed by simulation
I We document observed dynamics of adjustment followingreal-world trade liberalization
Contributions
I Literature finding substantially different effects of tradeshocks across local labor markets
I Topalova (2007), Kovak (2013), Autor, Dorn and Hanson (2013),and many others
I Effects estimated over windows of 7/10 years.
I We estimate the evolution of the local effects of the early1990s Brazilian trade liberalization
I Discrete liberalization shock → empirical labor market dynamicsI Compare existing medium-run estimates to short- and long-run
Brazilian Trade Liberalization
Tariff Changes 1990-1995
-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
Empirical Approach
I Follow Kovak (2013): trade-induced labor demand shocks
Regional Tariff Reductions
RTRr = −∑i
βrid ln(1 + τi )
βri ≡λri
1ϕi∑
j λrj1ϕj
I i : industryI τi : tariff rate in industry iI d : long difference from 1990-1995I λri : share of regional employment in industry iI ϕi : 1 minus labor share of value added
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%
90-10 gap ≈ 0.10
Data Description
Administrative Data: RAIS
I Census of Brazilian formal labor market
I Establishment-level informationI geographic location (municipality), industry, employment
I Worker-level informationI gender, age, education (9 categories), earnings
I Years: 1986-2010
Empirical Approach
Compare evolution of outcomes across regions facing larger vs.smaller tariff cuts
yrt − yr ,1991 = θtRTRr + αst + γt(yr ,1990 − yr ,1986) + εrt ,
I yrt : value of region r outcome in post-lib year t
I αst : state fixed effects
I θt : effect of liberalization on change in outcome by year t.
I RAIS analysis also includes pre-liberalization outcome trend control,(yr,1990 − yr,1986)
I OutcomesI formal employment
I formal earnings premium: µrt from ln(earnjri ) = µrt + φit + Xjtβt + ejrit
ResultsEmployment
-‐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)
ResultsEarnings
-‐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)
Dynamics in Labor Demand
I r = 1, ...,R regions and i = 1, ..., I industries
I CRS Cobb-Douglas production (shares may vary by i)
Yri = AriL1−ϕiri
(T ζi
ri K1−ζiri
)ϕi
where ϕi , ζi ∈ (0, 1)
I Lr labor, perfectly mobile between industries within regionsI Tri fixed factor (e.g., resources), specific to region and industryI Kri capital, specific to region and industry, may change through
depreciation and investment
I Ari region-industry Cobb-Douglas productivity shifter
I Competitive markets. Exogenous output price Pi
I Examine effects of changing vector of output prices (liberalization)
Dynamics in Labor Demand
I Factor market clearing, zero profits, and cost minimizationimply equilibrium relation
wr =∑i
βri Pi +∑i
βri Ari − δr
(Lr −
∑i
λri (1− ζi )Kri
)
where βri ≡λri
1ϕi∑
j λrj1ϕj
and δr ≡1∑
k λrk1ϕk
.
Dynamics in Labor DemandEvidence for Slow Capital Adjustment
Number of formal establishments proxy for regional capital stock
-‐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
Dynamics in Labor DemandEvidence for Slow Capital Adjustment
Immediate investment response, slow depreciation response
-‐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
Dynamics in Labor DemandEvidence for Slow Capital Adjustment
Immediate investment response, slow depreciation response
-‐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
Job creation and destruction defined as in Davis and Haltiwanger (1990)
Dynamics in Labor DemandEvidence for Agglomeration Economies
I Assume in the long run...I Constant elasticity agglomeration function
(Glaeser Gottlieb 2008, Kline Moretti 2014):
Ari = κLr
I Constant elasticity labor supply:
Lr =1
ηwr
I Perfectly mobile capital
Rr = R, ∀r
Dynamics in Labor DemandEvidence for Agglomeration Economies
I Imposing these assumptions yields the following expressions
wr =η
η[1 − ϕ(1 − ζ)] − κ+ ϕζ
∑i
βri Pi −ϕ(1 − ζ)η
η[1 − ϕ(1 − ζ)] − κ+ ϕζR
Lri =1
ϕζPi−
1
ϕζ·
η[1 − ϕ(1 − ζ)] − κ
η[1 − ϕ(1 − ζ)] − κ+ ϕζ
∑i
βri Pi−ϕ(1 − ζ)
η[1 − ϕ(1 − ζ)] − κ+ ϕζR
I Test for agglomeration economies
Lri = γ0 + γ1Pi + γ2RTRr + εri
γ2 < 0 ⇒ κ > 0
Dynamics in Labor DemandEvidence for Agglomeration Economies
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
γ2 < 0⇒ κ > 0, i.e. agglomeration economies are present
Dynamics in Labor DemandQuantification
I Estimating agglomeration elasticity, κ
I Estimate 1/η from Lr = (1/η)wr , with RTR as IV for wr
I Estimate κ using non-linear least squares from each of
wr =η
η[1 − ϕ(1 − ζ)] − κ+ ϕζ
∑i
βri Pi −ϕ(1 − ζ)η
η[1 − ϕ(1 − ζ)] − κ+ ϕζR
Lri =1
ϕζPi −
1
ϕζ·
η[1 − ϕ(1 − ζ)] − κ
η[1 − ϕ(1 − ζ)] − κ+ ϕζ
∑i
βri Pi −ϕ(1 − ζ)
η[1 − ϕ(1 − ζ)] − κ+ ϕζR
Bootstrap entire procedure to account for correlation between η and κ
Dynamics in Labor DemandQuantification
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)
I Estimates within range of prior literature (Melo et al. 2009)
I Kline and Moretti (2014) find 0.2
Summary
ContributionsI Empirically describe the dynamics of labor market transition in
response to trade liberalizationI Dynamic context for existing static regional resultsI Challenge conventional wisdom on equalizing migration
I Mechanism that qualitatively and quantitatively explainsgrowing effects
I Finite elasticity of regional labor supplyI Agglomeration economiesI Slow adjustment of complementary factors
I Benefits and costs from trade liberalization unevenlydistributed across space for years after liberalization