Im­pact Se­ries: 05–21

Post-COVID19 resilience:
Lessons from the China Trade Shock for Europeʼs economies

Lilia Ruslanova
Postdoctoral fellow at the Department of Economics
at the University of Zurich

Mathias Hoffmann
Professor of Economics at the University of Zurich, member
of the URPP FinReg and head of the Globalization of Real
Estate Network at University of Zurich.

Outline

When we talk about the economic consequences of COVID-19, the question arises about the strength and duration of the changes triggered by the pandemic. Currently, all indications are that COVID-19 is likely to change the structure of the world economy permanently. Evidence suggests that the pandemic is a permanent reallocation shock – one that cannot be mitigated with standard fiscal and monetary stimulus. What are the structural reforms and mechanisms that would allow economies to adapt and to effectively reallocate resources in response to such shocks? Which features made economies resilient to major reallocation shocks in the past? To answer these questions, we take a closer look at the China Trade Shock and its impact on the US economy. We then draw lessons for Europe today.

Full Article

COVID-19 is like­ly to change the struc­ture of the world econ­o­my per­ma­nent­ly. Bar­rero, Bloom, and Davis (2020a, 2020b) pro­vide ev­i­dence that the pan­dem­ic is a per­ma­nent re­al­lo­ca­tion shock in the sense that shifts in work­ing arrange­ments, con­sumer spend­ing pat­terns, and busi­ness prac­tices in­duced by the pan­dem­ic will not ful­ly re­verse. This re­al­lo­ca­tion start­ed long be­fore the pan­dem­ic but has clear­ly been ac­cel­er­at­ed by it. Pagano, Wag­n­er, and Zech­n­er (2020a, 2020b) show that stocks of firms in sec­tors that are re­silient to so­cial dis­tanc­ing out­per­formed less re­silient stocks par­tic­u­lar­ly strong­ly dur­ing the COVID-19 pan­dem­ic but that this out­per­for­mance pre­dates the pan­dem­ic by sev­er­al years. Carstens (2020) ar­gues that stan­dard fis­cal and mon­e­tary stim­u­lus will not be suf­fi­cient to deal with this “great re­al­lo­ca­tion”. Rather, the op­ti­mal pol­i­cy re­quires struc­tur­al re­forms and mech­a­nisms that would al­low economies to adapt and to ef­fec­tive­ly re­al­lo­cate re­sources in re­sponse to such shocks.

Against this back­drop, it seems nat­ur­al to ask which fea­tures made economies re­silient to ma­jor re­al­lo­ca­tion shocks in the past. One such shock is the so-called Chi­na trade shock (CTS). Au­tor, Dorn, and Han­son (2013) show that US la­bor mar­ket re­gions with man­u­fac­tur­ing in­dus­tries, which were par­tic­u­lar­ly ex­posed to com­pe­ti­tion from cheap­er Chi­nese im­ports, also ex­pe­ri­enced the biggest de­cline in man­u­fac­tur­ing em­ploy­ment and wages be­tween 1991 and 2007.

The la­bor mar­ket ef­fects of asym­met­ric trade shocks are re­gion­al­ly con­cen­trat­ed and per­sis­tent be­cause the ge­o­graph­i­cal mo­bil­i­ty of la­bor is gen­er­al­ly quite lim­it­ed (Faber, Sar­to, and Tabelli­ni (2019; No­towidig­do 2020; Blan­chard and Katz 1992)). In the ab­sence of la­bor mo­bil­i­ty, the mo­bil­i­ty of cap­i­tal should there­fore play a par­tic­u­lar­ly promi­nent role for the ad­just­ment to asym­met­ric shocks in a mon­e­tary union (Mundell (1961)).

In a re­cent pa­per (Hoff­mann and Rus­lano­va 2021), we ex­am­ine how dif­fer­ences in fi­nan­cial in­te­gra­tion across lo­cal economies (states and com­mut­ing zones) in the Unit­ed States af­fect­ed sec­toral re­al­lo­ca­tion af­ter the CTS. To cap­ture lo­cal fi­nan­cial in­te­gra­tion, we ex­ploit the wave of state-lev­el bank­ing dereg­u­la­tion that swept through the US from the 1970’s un­til the ear­ly 1990’s. Since states dereg­u­lat­ed in dif­fer­ent years (Kroszn­er and Stra­han 1999), there was con­sid­er­able vari­a­tion at the state lev­el in the de­gree of bank­ing lib­er­al­iza­tion un­til lo­cal economies were hit by the CTS in the ear­ly 1990’s. More specif­i­cal­ly, states that opened their bank­ing mar­kets for out-of-state banks ear­li­er had a stronger pres­ence of coun­try­wide banks and ‒ as we show ‒ a more elas­tic sup­ply of bank cred­it to house­holds. We de­vel­op a sim­ple mod­el of a small open econ­o­my with hous­ing and a trad­able sec­tor (man­u­fac­tur­ing) to ar­gue that bank­ing in­te­gra­tion fa­cil­i­tates and speeds up the re­al­lo­ca­tion be­tween these two sec­tors af­ter man­u­fac­tur­ing is hit by a terms of trade shock (i.e. the CTS). In our mod­el, this hap­pens be­cause fi­nan­cial in­te­gra­tion al­lows house­holds to smooth con­sump­tion which sta­bi­lizes their de­mand for the non-trad­able good (hous­ing). This in turn sta­bi­lizes hous­ing prices and wages in the hous­ing sec­tor. Giv­en the shock to man­u­fac­tur­ing, high­er house prices (and there­fore: a high­er mar­gin­al prod­uct of la­bor in the hous­ing sec­tor) speeds up re­al­lo­ca­tion of la­bor away from the im­port-ex­posed man­u­fac­tur­ing sec­tor to­wards the hous­ing sec­tor.

Banking integration facilitates and speeds up the reallocation between housing and manufacturing after manufacturing is hit by a terms of trade shock

Our em­pir­i­cal re­sults ‒ sum­ma­rized in Fig­ure 1 and 2 ‒ line up with these pre­dic­tions. We clas­si­fy states into two groups: ear­ly lib­er­al­iz­ers are states that opened their bank­ing mar­kets for banks from oth­er states be­fore 1985 and are there­fore rel­a­tive­ly more fi­nan­cial­ly in­te­grat­ed. Con­verse­ly, states that opened their bank­ing mar­kets only af­ter 1985 are clas­si­fied as late lib­er­al­iz­ers.

As is ap­par­ent from the re­gres­sion lines, a giv­en in­crease in im­port ex­po­sure over 1991-2007 gen­er­al­ly leads to big­ger de­cline in wage, real es­tate em­ploy­ment and house price growth for late dereg­u­la­tion states than for ear­ly-dereg­u­la­tion states (the or­ange re­gres­sion line is falling more steeply while the blue line is flat or in­creas­ing). How­ev­er, a giv­en in­crease in im­port ex­po­sure gen­er­al­ly de­creas­es the man­u­fac­tur­ing share more for ear­ly-dereg­u­la­tion states (the blue re­gres­sion line is falling more steeply). This sug­gests that ear­ly dereg­u­la­tion (i.e. fi­nan­cial ope­ness) not only shields a state’s wider econ­o­my (hous­ing and real es­tate mar­kets, gen­er­al wage growth) from the im­pact of im­port com­pe­ti­tion in man­u­fac­tur­ing. It ac­tu­al­ly seems to ease re­al­lo­ca­tion of la­bor — ear­ly dereg­u­la­tion is as­so­ci­at­ed with a stronger de­cline in man­u­fac­tur­ing and a stronger in­crease in real-es­tate re­lat­ed em­ploy­ment.

While Fig­ure 1 em­pha­sizes the long-run dif­fer­ences be­tween the two groups, Fig­ure 2 shows es­ti­mates of the dy­nam­ic re­spons­es of dif­fer­ent vari­ables to the CTS. The up­shot of the two fig­ures is the same: For a giv­en ex­po­sure to Chi­nese im­ports fi­nan­cial­ly more open lo­cal economies saw a swifter re­al­lo­ca­tion of la­bor from the im­port-ex­posed man­u­fac­tur­ing sec­tor into the non-trad­able (hous­ing) sec­tor, with more pro­nounced de­clines in the man­u­fac­tur­ing em­ploy­ment but low­er de­clines in the real es­tate em­ploy­ment. Thus, fi­nan­cial in­te­gra­tion spurred re­al­lo­ca­tion in fi­nan­cial­ly in­te­grat­ed states while in late-lib­er­al­ized states the ail­ing man­u­fac­tur­ing sec­tor de­clined more slow­ly ‒ at the ex­pense of over­all em­ploy­ment. Con­sis­tent with the mech­a­nism in our mod­el, wages and non-trad­able prices, in par­tic­u­lar hous­ing prices, re­mained rel­a­tive­ly sta­ble in fi­nan­cial­ly more open states. Anal­o­gous re­sults also hold for the growth rates of state av­er­age in­come and con­sump­tion per capi­ta. House­hold’s abil­i­ty to bor­row in or­der to smooth con­sump­tion is key for re­al­lo­ca­tion be­cause it keeps de­mand for hous­ing and house prices up. Con­sis­tent with this pre­dic­tion, we see that house­hold bor­row­ing in­creased more in fi­nan­cial­ly in­te­grat­ed states.

In the pa­per, we fur­ther cor­rob­o­rate our find­ings on com­mut­ing-zone lev­el data. In ad­di­tion, we pro­vide bank-coun­ty lev­el ev­i­dence based on the Home Mort­gage Dis­clo­sure Act (HMDA) data base. This ev­i­dence shows that ge­o­graph­i­cal­ly in­te­grat­ed banks played a cru­cial role in ac­com­mo­dat­ing the ad­di­tion­al cred­it de­mand of house­holds that was in­duced by their de­sire to smooth con­sump­tion af­ter the CTS.

Financial integration spurred reallocation in financially integrated states while in late-liberalized states the failing manufacturing sector declined more slowly – at the expense of overall employment.

Summary

Our re­sults hold some lessons for Eu­ro­pean pol­i­cy­mak­ers in the post-pan­dem­ic world. The ef­fects of the pan­dem­ic re­al­lo­ca­tion shock will be very un­even with­in ‒ and even more so be­tween ‒ mem­ber coun­tries of the Eu­ro­pean Mon­e­tary Union (EMU). La­bor mo­bil­i­ty and cap­i­tal mo­bil­i­ty should both be im­por­tant ad­just­ment mech­a­nisms. How­ev­er, la­bor mo­bil­i­ty in Eu­rope re­mains much low­er than in the Unit­ed States (House, Proeb­st­ing, and Tesar 2018), while bank­ing and cap­i­tal mar­ket union ‒ the ma­jor Eu­ro­pean projects of fi­nan­cial in­te­gra­tion in the last decade ‒ also re­main woe­ful­ly in­com­plete. The lack of gen­uine cross-bor­der bank­ing in­te­gra­tion in the Eu­ro­pean Mon­e­tary Union (EMU) has long been iden­ti­fied as a prime rea­son for why risk shar­ing among EMU coun­tries is so low and gen­er­al­ly not re­silient dur­ing ma­jor crises (Draghi 2018; Hoff­mann et al. 2019a, 2019b). Our re­sults show that ac­cess to fi­nance for firms and in par­tic­u­lar for house­holds will be key to ease the great re­al­lo­ca­tion along. But Eu­rope still does not have an in­te­grat­ed re­tail bank­ing mar­ket. Com­plet­ing the bank­ing union while mov­ing for­ward on cap­i­tal mar­kets union will there­fore be cru­cial in mak­ing the EMU re­silient against the chal­lenges of the re­al­lo­ca­tion trig­gered by the COVID-19 shock and to sim­i­lar shocks in the fu­ture. The risk shar­ing mech­a­nisms in the EMU so far have been sur­pris­ing­ly re­silient dur­ing the cur­rent cri­sis (Gio­van­ni­ni, Horn, and Mon­gel­li 2021). But Eu­rope’s ex­pe­ri­ence from the great fi­nan­cial cri­sis of the pre­vi­ous decade shows that we should not take this for grant­ed