Regional Labor Fluctuations: Oil Shocks, Military Spending, and Other Driving Forces

B o G o t Fo Roa Syv ed es er I nF D t Paeii r s nna cuat N 5 um M 19 ar R E G I O N A L L A B O R F LU OI S M SH I A POT O DL E FO ICTKRNI D S J D P L t a aRr o M eav u ak vi ah ma ng N I nF D t O P eii a p r s mrnnapTEc c ae at st ir taunsli te io d an ci c s Rr o i pc e i tu Imu f bnFti s e Dmetl sPiirceericis ( t a a c o th tkh w hna h haa to t u rwi mn l cc shhe epbuactl dg w th a o a R I uu ar a it oet t Wv tah wFD whoa ceu w i.l bo

R L EF L G U AB C IO TU OI S M SH I A POT O DL E FO ICTKRIND S J D P L t a aRr o M eav u ak viah ma ng A b st W q th c u oo v dna fo tatr s ntri r mota ib i u n r an ee gmm fr 1a9tp tp 19 Olro st l o ro f o y eiglut ymme U e ha al ca co p - i .S go alr n c o aw nocaay tvonba olumier p - bu oei p s har be th l shraic si 19oea Be oct mnna a a o oi bp m r th eo uf xrtihv p p p re r eftl hm oteng aennse en e (i r l d i i eemi (i i ni m d gni s de t fdmfe ioif nitustnussroev p o o an ( th r roe p af ii rila li oicaacl ren cidr coo adut cauc c o n s u m in Ou s p e orth cos tuo c o r t j rea rveiiemg re u n t th ea o hmm cw p Bai roo lt a B meonli ofourd ey tr mmep ou b s pa i the c scon lo j ir mpre f n o ei g elqi c aptub-oavtat f l f i th a o $ to $ iro m( 5 i 1 9 m c9aTo6u,rme ea8 1,c o j2std c i m t rt a l fo th eb C mwiora hEaarr e sc ti pott onacso ec d s a ps e b id ocjo c lm tcauo ee l r rroo n s40ano li smedaa tive W f a u s n r y te pin e mam n osrsmeppe s eglo N s - w e i ih ghni o p n o sav oeclm co rci o tiacwa rteha monilv b - h a g i th ea r p qm sav euTh oe se aihmpoa tatsivishl- t t s s o d pt ( ar r oeega e acu o a tl mcaacvons.gl wina eitaruochoi a u g n e g m r p l e o ga y me S u t n ra et r t e smh epe espf se- ly pgoi Nle oy m o p ian w b g se i oth d re e taoproaqm tw ut btir emcikreeil u n r b ien a m l atp i l ac go y nm me K r ec oi se y d y e g wh xme cbapoi u oc ile fenrdleonn em di *L I o nF D t u F eiRi n Br Dea v nUnges oaChiosandeana taNn iv io M A a C hoW a rmi o vce up d c k o f h nVaormse R aorwa(t d a a iN c s Toi B cDna B fuBEa l Ra e Ebsa C rrtEn sa M ench H J Z ano s pi o aaone c v r i u ktmiak Oycc n ie1nmeco elre ci cl o “ I E Sn E t eVne c utrOc Pr Sln oer’ una b t toehrDoenori iora o E an th D o nSe W t Op B et aa Lalil Kha rf s rgat an dtm tao e Mm W fpo D e lo Dee pr o cr p awey a gnpaon rmi He feertrB en fo N p c a W o V r wfASs n c t raoyuar t aiminv ra a tePerma K fo i s e r nd D gn d etao r cantr u ava suktu f esesttetdnoU N S F a an ctoh U. tD u o iEe i nT vi p i teodnnnep aar s at t r e o th a s an s p no buoi na rnh stt vite oi t B eobufolhoGo il rp o th F R S eo o aneo p yad swi ts FeertherRste ser Sy oc

1.I nt Re f i ec fo a llu a p a t a k in f e re sc jo aco po W f o l ma ou a ad o un o r f i s re re Fi w qu t co o v d f t r mo i u r a em nle I t p w c n me o re ec di t c f b a t s o as o re f S w e lut c o c l j a t c o re l un t n g p o g a se f l f T w ch t d p b w re l ma gr t l o i t w o t va d L a a by iso s t d f b r IR f t in a h v a su b the followinglist: regional f pr a u te glu f t o r s a c gr d t hi o U re iefx p a l f g i t po c o g Eu e on i a c un o p sh a t i g di e a r r a on cu i U de sp f h u a r t ti a ma o po in i U e i a r t ed di sh a U re r e v v r t vo a c ac ex ]) on re c R o t to in r w b As e a ( B a S ( 19 Ba (1 Bl a K ( B a R ( D a Wh (1 E a S ( Ei ( H a K (1 Me a So (1 R a T ( a T ( 1

f l w d u ne ec eth t s v tu t sv a ai hion t psetal die o on d p a r i d e a e r g mla am olu g iv pn fan re e lolu mp A ce pr o o s i t r d i t b o g a s t pr Fu t re di i i m r a d o ca a l in t a sp i t o b l a i T s na o i im t pree t al u i r t re d wh t di is a in s o s r di re ef o s t t in s o d C sh t t sp o in st o d g r t r d i wa U n an c e u m t a p i l lpi o i ym za T d i t b a icf i r l f p a r o ca a wo wed re m o w f a t fa af re a re f o c a j T s a c o t e f de t qaud o r r e u a l f p ne t r di ar W di me a b r o di w e u t q t re i a in w t t mp qu d d r T re d o “d es f tis w c f i t fo c at 1 Sp a t va i mi ex a m b D o De ( a Na Ae a S Ad ( p co aw t f i t r 2 Ch i t n o mi pe i a r ( t p ar f ex f b cl 3 F i t p o o in lu w r di i t m o o 2

se i nd 4 Sp a t va i t ma o s t t in s o de a pr b r me o t cr gi d i eq re 5 T n ef o na di – oon t o p a m e sh – t d a xp r e if b o r d i in m No eif o t s p a b o ex r mo c i t s o m p r ch t re a m o e De co a a b cl a co a i t p u m a i di o ec p b w f t o d —e o p s – h p l spr a d f b r f N w f lus un on a e r t mi e sh t I e t xpo di l a o a a im dr f b pa e o c a o r l ma f lu O es ef o t di a o ch o r d na de f a l p d s t c a o f 1 t 1 f a 5 st a t Di o Co P s t r o sh pa o a sm n o s o o r u T l p a la s o me dr f i o s u o s a r c o re hf ar lu W i t j dy o re nuvn e a l f p i o t ch art ad p t b t d Q br w g st t D a he r t t e g a s 3

fo W co a d Bl a K ( t m r t r gi ec d h in s un is r f r s e i y su t t i esp T a l e t l t im o t sh s u o t di o em a r r t o re d i un r U ifB a K w a de ev t t ra o t mi r v b c o s a w f t t mi re i mi b e t r o t B o L S ( me o s em rta t t b C P Su ( me op W es t c i t o na go ex o c l c j a re l un u s em a u me a a va o ec s B o t Bpe m o s e a an di r i t mr p o 3 o b s i t cr o l jo re n go p f l f i t am o 5 th 1 d T e c o j c i m t t a l f t b C em m a t c o p l u re r ur f n8e t t 2 m d p p E s ert c dco pe s a s b de j cr c es rou 4 s Sp t a f a ef in t ex rsey h n t t im o l u a ex in ne H a K ( f si a em e o l d syp Se as o o em fi b t v t a s p a r a dr fo b ag fl . F w d c e 3D J-Jaltiwanger a S (1996, cl 5 r s theories o h a 7 4

d t o p s h b m d f b r e a u f a t tnetr lu m i t i o t in st o de d T e c f e s ca t is t s ti co ta b s o a s me o o p t s I i t d atf t a t ev co b Ha im ( a m o a i c i po re l su t di o in b o p d e u t co o un ag di t m b c w o p sh Se l Lo R a T ( a B a C ( w f t a r i t di o s r a i l t a i i u W in t nfe a e t s t t s o de ge u in b o f ne i t p o r wo a j U pr st h o e r o s t c be un im a t ovdi o e r Th f a t ca o ex m s i t s – t b i o mi pe p co a a o p s – w f m o st ev t ne s h a g e o s u t po sh o t s ma T e o a c a b p o th t s t r o al re di a d f b sh d ag f lu A (1 a re t l ht t in w o Price shocks d ag f a i s t w clu H ( a H to (1 en i a in ex t il h l s p ha i o t ag tnit e 5

ag f I pa t ev lu s t a d d n ‘( o i t o t sh i o a un I pe w un in al di a i g a u e o th e A ar mpp o e w f t t d o u r a s v neco o t ( c o s f ef M o t va i ex b o m d f T r o t p i or a fo S 2 d t s e pl a u d a pr a stne ch o r l f i t o a s va lu co m S 3 d t dr fo a c im a o tha ti a c v at Se 4 es t ef o t d f o t r c c o ci a in u r W c sp ne t a re s a po as e o ppi a n s W a qu t su o t re m i ac f t a c va i t re c co o un fl S 5 c si V mo o t j dy o st un e a l f p p o s a m ar s v W e t c o cr l j a re l un u b t V a u re mo S ne 6 su t e a c 2 B fe o re l fl 2 S u a em tam ne W t in s a re u a c t m o u m ( t ci u r a ( t i neun r d a t nu o u in ( cl ned b t n o j 6

co b t U sy T d e f 1 t 1 f i u a f 1 t 1 f ci u ne T a we co di b t t un m In u i a a cne o p dwm f c u t U pr o t va st GT l o c un i c i a es c o pe se j re o t e f U b o pr l f ex Y w w a w i s a g in a l l t b e ov f U b s t t t b u i t Un c de I l o t d i i un t t ciof b t t me i l t pe O t 1 t 1 p t m c b t t s u r me i t.a onne o t c a l t . T me co r t . w w e t p a 1 W a co t em me t h i c d en T Bu o L St ( es em u a l s o n bu e o T B f st r na t n o p p s t pe w t j i t cos s a c t T B me ex s em ag a p h e a mi p T Cu Po S ( er p a b m o 5 d f A a H a un p t 1 G b in i t c t cl n h b e i a J t ‘ co b t U pr h e l m h e in a h m m o e cr de b t U l li o e s S G ( f a de d is TB (1 B a C ( a V ( a t P ‘ rg be in a t un r a t n l 7

I a ci e b o a m ho mps A t na l t r o B t C em r f 8 i 1 t 91 i 1 w m o t in oc d t 1 A t s l t me c be e morr b o t pBl a C me i . o t 1 t 1 p 2 T ma a pe o re c Fo Ma (1 w u a s v c m t g t fa t d va i un r a t a s i t b c at Lk=-’)+’tas+ U re t t u r ( e ne mpr i s s i p t. T ~ t ca fa t s ag l d a s a c si t mo i na a s im un r T a t re d i am l m in if t p o t U s a o as o t ec en t i di a s i t na u r S di nie t f a l t r s o ch sl o t t l t pe un di G co l a ca mo un di a a f l de a su d w di e a is s T e w w he r t a “r cy a c i t m a b t t e T d t d re do is g r t a o o wo a a i o j t t o t n- t a t u 8

I d i t f b t r df ie i T psra g tpe t u unage t rts biaa o t a m ( – p i t f dj o t ecun d e i o y b ch i l f li pa a b t m o wo a j T co a o n r di – r b ~ i eq ( – ge co va i t r c c o g u d eo ne if T s m ( a ( d n g ta or c a t a g re t be t t t s o v i s u r a i T m i ns u f as t r te i o e so o va i st l fl a f su t p o t re c co Es o t m a i T 1 Ac t c ( t s de o t ci u r i 2 T n net r r s d d o t s f e ( t y ev f e ( a t r c co ( I i ev t a t co p a i r i a f s u r va cc T s ne d o t r c co o t ci un r ( i 1 p p a i va eq 7 o t t va T va o t s a y f e am t 3 a 4 r o t t v i ecs u r Re f t in u r a p t a l ner f e c e t re c co W ssp a y f e c c b ig T 1 s cl t t re c c o w w f i a m so o va i s un r An im me o t t i t pe o t r c c F b t ci a i omun r t e o p i 9

a 0 T i 3 o t bun di i eou i o y Co 3 a 4 o T 1 r t r o f t m t t a me o t e m r B o t p C e mlo t re c co o em r fl a f 3 o t t v a t s a t s o y aer T e f p o . i gr p o re mo i t eer mpr t i t u r ne T re f t o me o t e mps a s d fe pa a t p t po in p a w t B e lo me T re m at F t r s d o t e m r i n 9 ppe p u lo t B e me a co t o 5 pe p u t b C m S o t st de o t es s f e i m t f t la u t B em me T t r c c i m m p a i va m g u ert B m T re in t pe di a s i t i a sM f m l pe i t r c c o m u ra Mi re ne a p m r i M d be h de re m na M a u am s s p W w r t m u Ma 1 e- s p o e o (~ d t 3 percent f t ci un r a t 1 p f t in u r T es o nped t . a . f t c a in u r re su ne m l p i t re c co Ev i i di t di p r c f re f ef i s pa 10

o m o em a cc f m o t cr v i B b e m ra Fu p a m i lo f o s re sh ge m m pe re i t m o e ( t B co a un se t i t l o e B o t po f c em co a oen l e i i da t r so o B em d t d i a t m an p oe e r amn m s rp i i t rl e sgdi o spt ra ymen 3 T dr fo b re fl I t se w de t d f b r u a e pl f a w ch s lu im a o c a t se va i t me 3 De a N e xp P co aw L se o st w u D p c a a a m o r de e . W su t D xp f w d o N p c t aw t ob a m c m o “ omc a b st T p co a d e a g d o st t v t a pr c le f es t one o f p o re e T s va r cowta s i n m e co w a h xup s di a p c i t sp d o t W t st is v i p c a a ex w re t st em a un o a g C (1 M a So ( M e a ( M ( a Ho a Kn (1 11

I ap st su b t ev a s d b Fi 1p r mi p co a p c i t U S f 1 t 1 i 1 do ( o t G i d f p c e A di b Cx ( a M e pae ( t a f m ep i t be o mi c a o t p W pe ( t in i de ex as w t o o t C W ( t 19 ( t es a su s b o d e a w t Vi w ( t 1 ss( t re o d c u Ca a Re ( t 1 a ( t d c a w t e o t C W ( t 1 F 1a s p c a b N w re a on t s o D co a t p i 1 T f ep i na d ex w a w p sh i t sp d o ex a il is b m i r co aw f se s i F 2 C g r t t U a a wh f t s i de ex t t p a b o m sy Ge re co r a f o t c o t V w a M g f t s ast h w s t b i t m Ca a Wa exid d i r c aw i t p pe e a n os ex d o b m t o I a pa ty o s s ne I h s c a le t t e p o ~ s Cohr i c e v h a va co a ~ c Fi 1 a 2 p t co st v i p c a p ca T pa o va a i a co t p a n se d su t t va ev i e w r t s 1

I ec fo M ex a sy s o t de o m co aw st su t co M e a ( d a ex t po ev i t s di o U m e i g de R i t 2 c d em t a “ o m i pl t n t t U gond t m c a t l o re w de l m oc r l t t l ( 4 O ch i t b de t c s a p r e al n me t i M ( s t r o C a t p pr m ge i t aw o d co H c ( 2 t “ Pe d n i ca di d c ( o t b f po pu W t ex p i m r t p fo a i t al o su a t r co a t sp o aw t m di a t ta o a t t s o d o i me o Co n t tnaf o a t d r Ba o mi pe Co aw co ro h o d s a c o D pe ac f m o t r W m r v i t c co o mi e b t r o D em xp t p T r ex w sw a t na l a co s v Fi 3 p t r o d b D p tomt U p f 1 t 1 T r de s i t f s y a t K W f i t e 1 a t r slu b 1 a 1 a t U ex mi ac i So A T r d s a 1 a t m o t 1 De r hra t e o t C W i t mi 19 1

T o t s pl i F 3 p s i a t ma a va o ch i t st r o D e t p t P i t m ab c i t D e mpr o i 1 1 a 1 T cr va o c i t r p i 1 a 1 B su me in t l st m i t r o D e t po pr o i tmpf h o t s Ch i t ag n o U m p a c e w re t re ec co a t ag de d m o t s va An ev tha s t d a t ba o mi pe re a l m o p b p M ( p 13 pr se ex a h a c t p f p a im r i b s a c de No M d po t a p tr f r re re b we a i me o Co n t a pr tr f d r t w g r t a e b i o re m ndI m p c a s st a f y de r o i e st a d t de re o es b un t t e o e c o re u ne Mi e re xp O r m se t m c eg a a t b o mi p W l C d t c fer t – 1 t t i r p c p co aw ( p N f s s a w l I d t c i t D e m r p lo IO A a D a in t m a c a t s d t o ch a mo l a m m v o t 1

Re ec h l re t m c a g % ef o o s t p pi d a l m c P co a re ty pu l a m f o s I ad p co a a p su t f o i o st 1 T ca t sp i a s w w m s o a d pe o o mi e a o p c axp a D p ch i o st t b t t s C g r I p w l C de t c i p c co a t o s i t Ce re a w l l d th i t D em r i t o st i t re 3 O p sh O p sh f pr i ex f a e f t a a ha o pr s ua in t r o o p s i re cy B a S ( f e fal i e o p sh t re g r o c g s p d t 1 llBolton(19(j6) Markusenetal (1991) :Lndlvlayer(1991) contain useful descriptions of the 7 s pr a s uev o t e a s bc f o s T t b o o kn c d o su om a n a lz a n C re N E ( M M N R V M d At ( NY E N C ( I M O W W N C ( M S At IA( O, D F G MD E S Ce ( K MS W S C ( L O M ( C I M N UT a P ( C H N O l f ex Ha ( 1 L ( D ( H ( 7 D a H (1 a t s b aMl ( 1

a 19 B a H (1 S ( a B a B ( e t t ad c o a o p i aon l t b d a st i a hi un m b o di i t i c 1 T r p o o in dr d t OP s o 1 7 a 19 a t de t n i p l b 1 T d i c pe pr f 1 t 1 w pr q t p a t “t o sh o t po p A di b H ( 1 m mo i t re p o o e i t “p p r e ex t t ag ec a ce t r e I o o p s a s innt co W h t t st i o oyp p c d o i “ fo o i T ca t in b nd o s a a s i m w pr a fo L G d t n g r o e i in i du y - F t ex w u t f te d t ( mi ( c ( t coon raa p u (T ( se ( fi in a r e ( ( t ( g m ( pr me ( m ve a ( o ma i T f s in l a on i g o o m T m de br o m r o p v ant t r o IA a H u s i t: f 1 t enp t i o ‘ 1 o p de o s em S l m b a t p o “ rs re t o es u g s so p d f t 19 pe H f t o av r s esh f b a m vo gr o t pe B a B p e o t i o a o p in o g s pr 1

m in t o p ans i l t e co h W c se k in t a l ont ex e s h fa c o sh l f g d i t w o a o p i A s e w c prx m ( 3 t c pe t f e a m ve ( 3 t ca t la F e in w ca o na ti re o t f G = / + 6C - t3 i 9a101LRt–1+-vat, i = ~,3 ..., ~o, ( w O i t gr r o t re p o c p a C i t ch i apa di l o U d c a T c ( a O s t re o na em g i i zt co a la o p sh W es t re u a U d f 1 t 1 E o t O a o c ex t ex pr oe M e i s i re t a in ig i o p b t e c a n t f o i a i m c si dnd f z T e im o o pr o em g i m l i co m a T t i t o f l in W ma t i l a co S 2 (P a C Pr S 3 ( C a G Pr a S 3 ( Ma t c No t “ h fa co ef T t in d n a m t t s v i t O me ( b de sh a w d f f c IG o t di i B ( c 5 w c d ( ‘ co e a . t c xp c aon p . t l aw p . t t a C e t wic i e s me

I o p m an mr v e i t i ae la oat f tta mah“o wim aic nu c a t e g or W i th e n n s r t aot i e e e it nsd tmproa msaiopo w s mt o i a e e itt n sma e ( d pb- sSu dhaea flloe us 1 OIL~t= – ~(i92001LRt + 02101LRt_1)S28t. ( 2 I i im t n t w co t t s f S b l i t be de C y T pr r p c c i e sh w i m mslp snc i s i c om W h t O co po ywp st u m me I o wo a o p in c a s un r i s w hi em s de t in m a a b a i cr i o pr O ap a f no r t o p c a t re o se ch i st i po 3 In m a no re t n b c T dr fo co t f h t ad t t m m re re i ev s a w a pde d b O t a ex t t U bu c T t e t n b c a d b s d w e t c ais no st r t n bu cy u t me de a H m v a o p sh a un t ex t l o im d f b n cy w no st ef H fo N a T ( a Ba (1 w co a ad ex v t c d a st i t se t na b c d b o “ 1

I l S f e s w co pe a “p n c e d is o e b i t na em mp g nt r r v i t r ( w st me oegin em s 1 M = ~ S ( 2 T s na c va t p u t e o t amt s s a ch i t st o mo p t w d n m d W e M t co ne w s un 3 A sh w s ll Pr re in t s t t i s o d p a r i ag u mo b o nfe i t p o r ca wo t j IL (1 D ( D a H ( T s l im t s t t st o d w s p a r i s u f L mene t ilu o s ‘L d b c a me o t di ono em g a i L g R a T ( a Br a C ( s t t d o s re ac in h m ex p f a u . W im t i a t s l b in i s r a t IT st h a l hi i re e dc f oT ( n ‘ ol Fo a Na (1 1 T st su t s re w b b t a m l t re pe sh t in fo S r d m a o c ce pr su b e di mpm b s re re m im a se t n a p s t t de a o l a o f o pr ll 1

n l w as e t sme t o ta pi a istt mhaa l dovnebt oiy is f e s L Rzt d t t s r ac f i at i a t t a l R d 1 W ~o t w di of industry-level excess t ag s re re w we g b st em s 7 1 1 STDISP~t= ~Si~t(Rat– Rt)2 . ( [2=1 W ex STDISP~t t c po w st un 4 T ef o t dr f o un B r ( 2 eg P A i T 2 re t es s o co o c a l va o e i va i o ba un nd re T i pe va a s b t re s de s t t s g t es cu e o a o st de i i c a l va A r in s a y f eeg s t t c s pr i a t f d t nrf c c o s u f P B nre t lsut f t j n h t cu a la v o a in v h n e p 1 a ag 9 re we co 1an va l c ‘ f eq va u C d f sir i g m m i c t a pon ut ra F s a t z Ex f C a ~ i t ci un re l 1 c b tr t t wi l o ex p T l l f C p re t f t a e o l c a xps B ( W m a si ad f t l i ca t c o j c i s 5 20

I Co t re f t ci un r re F a f d fo t c s h t an s aoe t n h o n e i re a c s l T eson r ig m f a u st de s v co a t d f O p s h l ef co i s t t co r o t un s c b t lk va A s i t t t s o co o c a l va o O i 0 pe p T co e o c a l va o O a t de va i a p un r t a u st de ch i O o n o pe p T o i va p sm r a dr f b r un f Re f t in un r a b s b t e ef o t dr f a ma s O c t m a s de o t in u r a a m sne T m d pe t t mi va f w t co s a s a s i F- st hno in t c a sil v o C a M a jo s ig E t b re ( 3 a vaF 4 a 5 O es re h T w U i t u r i s s a t =nei a y e a i a s ef t ve X co a re (i l o t d v ~ i t es co ve a q i a r B o ( w c o t t o ex t a h w t b re m e t r c co o u va T f t nef o t a l 2

an t v o c a dr r i imtoi r i csys c e as p Tom s-sgitier fo o t va i t re c co f i s F t s o br w re t re o t ex f t c u r o W p de t re c co rae ad f s a y f ef i U = U – a – ~ L a d t u s d o t re c co a t t r .,.–~ 11 1 at = ( T7J2 , ( p 5 w ~ d th u he me n r T,cy n wc e er Woe reot gti o ma pitg “a d o t re c co is T as t p o t m i ac fer t m a t va o a de t c “p di or r r~ 11 1’ o’t P — — ( - XJ?)2 , ( p S w X eq t un m o t es r c c Cl t be o u de o t ob v i t d f a t es c W co OIIQVersioone O va o t l de va a as “d v u t m i g p T f ve o a h re a t o ael ss o t m T se ve qu t mo ca t e un v i t o me dr fo Fi 4 di t t p o o a a a T 3r r i T ac di i t re c co o un r f a 2 ....

st de o l t o pe p i 1 t a 2 p po i 1 I o w t d o s di i l m t va c o t U ob on v o t d v i ( t pr di o t re c co c t a d T 3 s t a di i mo l o a a m v o t t pr di T s co b t t s i . T dy pr di r f a s d o r . p po i se w sc yer a . p p i a 1 I av . pe p a c t 1 p p f ac d B t me o me dis f a f a o o t av c va i t r c roc o u f Fi 4 s t tlu d pr d c m o t t v ac di o t l arha s A a d pr d r sh i t af o t mis o p s 1 19 a 1 T s re i t l p o 3ta a t c o t f dr fo t mo i W co a f e b z o t es c o t o d f oeB t c i c t co o dy ov o d I s t t t T c in t a f dr f m a ap co t e t l a t v o a o p s a t arb d f b re u f i t ofne tlub t a l o d a i t va is T as t mo ab t e wi t v i t r c co o w p fiu a a m F e s s 2

w f ca a me o a vo i t r c c [x T 11 g = ( (lo) r,st — ~r,s)2 I t=l a a me o pr vo [x T 11 ~P = xs@ – x.@2 . (11) s t=l T va o O f t 5 s a s i F 5 b t h o t w po o t b T d po o t b re [ – o t i a d b pl a t w b in o pr a a d b b t z l in u pr T pr v @ a g u t m i ge pr o t l dnet va T f s t pr va o w vo l u w w iat v W t i a te t o ( o o 5 st t f vde n s a d pa t t pr er T 4 re su st f o a a a a b i a l p t fo o t c o t in d on f O a o p sh em a t s m im d f T a e i f re u f i hf ne s s lua M I a O T mi va m s co t r u f t i t “a s b l co i s s F e o i l f C a Wa w in on t p c a h b a im so o un fl i t s o i l Zl ac va o t l de va a u a a ~ a v c f al e st 24

f Al t Di o Co a H W a N M h b su t la t av s t t in s o d j b t r l va f @ el ’[ 4.3 S (T 5 pi T 5re t re o a re sp v f t c a a mi pe va T fa co o s a o e w s e sp va b t st de o t co o me e CONR i s id t C Ch i t b o mi pe g t s e i a t a i co t d ow e T co s o . a ti es a c t z i b re T co o n o ef a t sp ef su t m b d h i b hi lo ef o re ec I co t sp ef o p c a a l a s s I t ci u ig r re t e ne c s o CONR i t a l a t c s f t o orv C T a f va t fi in t ~ c m ex i s st h a bi e t e p capitaow ex T c o re hl m b i n arc i bl b i i w t r t fe o t p c d F s a h e o f b t boxp o t p c s z t re l l ma a l c ap b C r t b U st t ar ex m esi c t t i m t a ow me 25

I Se e ty l so bxp c a a t i c er va a ra i t l o l B f i e i t C a CONR va a me o ex b t m e co i pr m sm i t re v s a a m li t re w t r t t s a t r c o t ti l i m l t a o a t r l T w s t m e un t re eal sp e o c a b q t ex a in o meua e i a t f a o R o i T 5 hi n etg d o i te r s ef i e t l ec e o c vaa A ef ( 6 sy St r bu c m t in a r f o p s [ K a Lo (1 pr t t re o r a t p a n o p sh i ro sy I t c o m o s o t we ch o th su e i a a p f S o re d a di b t “a a “ ch th w o s a ag a [ D ( L (1 M (1 a D a Ha (1 I o s a t m be t de a a di o f i a i t a f i t pr o r f i t n u eae a d i o p i a ne ef o ag o a em T t e t o s in ag ac th al a w a a c p i h la n ef t p de W in w a o t s a im i T 6 b al n a p v o O t e 26

I s i t u re Sep a ne c b m f m e s w a a C xpC a I t h a e T ev i T 6 su t t o as p b t t st t a as o o a mi s I llt c u r r t es ef o ne eg c i o a r c aw a ne t a l a t e o p c T a i e m dr f mi pe ch T e f O p t a m a i ths c u y i n r m e a vea shm ma g pil i ettsy irlnoe u r re T dene f s i c w t v t t im o o p o r ac w t a m o b “ a ‘ ch La W t ev i T 6 f t as pr t h t r s re i P B r t m o t e i w T h o e c s o po a ne voe i r a co s l e o f t mi pe va i t c un r a f t o s va i t in un re T h o i d l o po a ne v is i r f t c s v i b r a f t mi pe evg i t c u r eg 5. T j dy o u em a nepa 5 A P V M w O S a C A T i t j dy o re nvl m v w s a V m t in O C logem ( t c u r (UN), a t l f p r ( Es ar w a t O 27

an CON va t t V s co b B a K ( W i re sp ef o C nci a m s o b I l o o ea d o ex w id a st is V b i t f ca or o t r in (OIL, COeNd, EMP, UN, PART). Re OIL a C i t or d n a t r W es t V o p d f t 5 s o t 1 t 1 p W in t l o e va a f s a y f e a i co s c a st F A oea A i A A d a co s o im re fu co t t e V a t ca or sp a I t m t w r a t t r o l ma va t OIL a C im Fi 6 s t dy re t a u s d OIL s T p ef o t B em m i a d o 0 p a t y n y l em s l 0 p b i i l T p i o u i ar o sl o n0e pe p o y a t i U re t i in lne i s y T e o p a s a p T ef o a ne u esr d s t C s i Fi 7 a q s b s i mua T p r i t B em me i a 0 p d t y a t i T re su t B e o t r kmo m a t “d ad me i re t r s T i f b th su o pe a an u rc n t r eia i m e cmsm t iuspslpt pa Q a ei va V msE a u t ordering (0~-L, cON~ ~~x? EMP, UN, PART, STDISP, MIL) y r s t t o d i t t 28 . : , , , .

employment response. This pattern is especially pronounced for the longer term response, but even the impact effects are dominated by employment adjustments, suggesting a rapid migration response. To investigate the sensitivity of these conclusions to the employment measure, we re-estimate the VAR using the CPS employment measure in place of the BLS measure. Figures 8 and 9 display the BLS and CPS employment response functions implied by the two VAR systems. According to Figure 8, the impact effect of an oil price increase on the two employment measures is nearly idential, but the longer term responses differ greatly. In particular, the longer term CPS employment response is about one-half larger than its impact response and about one-half larger than the corresponding longer term BLS employment response. Evidently, longer term migration responses are considerably larger than the initial employment impact, and this additional effect involves employment losses in sectors not covered by the BLS measure. Figure 9 also illustrates important differences between the impulse response patterns for the BLS and CPS employment measures,but of a quite different nature. The impact effect on CPS employment isvirtually zero, and the CPS response function lieseverywhere above the BLS response function. This result indicates that the negative BLS employment response to a decline in military contracts ispartly offset by employment increases in BLS uncovered sectors. As a corollary, the implied migration responsesto military expenditure shocks are smaller than suggested by VAR models that focus on the BLS employment measure. In summary, Figures 6-9 support the following conclusions. First, most of the impact effects of negative regional shocks show up as declines in the regional cycle component of unemployment and participation rates. That is, the migration of jobs and workers does not occur rapidly enough to immediately dissipate the local unemployment effects of local 29

shocks. Second, the longer term imprint of regional shocks largely involves changes in the regional distribution of employment and the work force, rather than persistent differences in the regional cycle component of unemployment and participation rates. The migration of jobs and workers dissipates the local unemployment effects of local shocks within a few years. Third, the longer term employment losses (and migration outflows) induced by negative regional shocks greatly exceed the initial employment losses. In other words, regional shocks impart a certain momentum to regional employment growth that persists for several years. Fourth, apparent responses of the BLS employment measure to oil and military impulses partly reflect effects on the distribution of employment between BLS covered and uncovered sectors. 5.2 Incorporating regional spillovers into the VAR model Section 4 finds large spillover effects of military contract awards on unemployment rates. Motivated by thisfinding, weintroduce spillovereffectsinto our VAR frameworkand explain how to calculate impulse response functions that account for them. We then apply the expanded VAR model to calculate thecost of creatingjobs andreducing unemployment. The CONR values are linked to the CON values by a set of linear identities, so we need not expand the dimension of the VAR to accommodate spillovers. Instead, we simply introduce the current and two lagsof CONR on the right side of each regression as an exogenous forcing variable. The main issue is how to appropriately calculate impulse response functions for military expenditlmx. Indiscussingthis matter, itwillbe convenient to alter notation slightly as indicated below. Let POP(RIS’) and DEF’(RIs’) denote population and real contract awards, respectively, in region R, less the corresponding value for state s’. Then, assuming that yearto-year population changes are small, we can approximate CONR for state s’ at t as 30

follows: (12) 1 DEFt(s) POPt(s) DPOEPt(sAF) PtOP(t(RsIS)’) ~ A DEFt(RIs’, s) 1 [ Popt(s) 1 POP JRIS’) + PoPt(s) POPt(Rls’) [ Hence, the impact of an impulse to CONt(s) on CONRt(s’) equals the coefficient is invariant to the size of the CON impulse IPOPt(s)/POPt(Rls’ )]. (This coefficient under the approximations used in (12).) Summing over s’ ER, s’ # s, delivers ~O~~(s) ~[1/POPt(Rls’)]. s’#s Taking the average of this expression for the states in the region gives spILLt(R) = ; ~ F’OP ( , S) ~[1/p@’t(Rk’)]. (13) SER s’#s Thus, to capture spillover effects in the VAR impulseresponseanalysis,weshockthe CONTRterm by SPILL times the size of the CON impulse. To capture the combined effects of own-state and spillover effects. we simultaneously introduce a unit CON shock and a CONR shock of SPILL units. According to (13), wehave potentially differenttime-varying valuesof SPILLt(R) for each region. However, simple algebra shows that equal-sized states within a region implies SPILL = 1. We maintain this value for WILL in our cost-of-job calculations below. 31

I 5.3 Military spending and the cost of jobs Recent studies by government agencies, congressional coalitions and private research organizations reach vastly different conclusions about the regional and economywide implications of the post-Cold War defense cutbacks. In her study based on county-level data, Guthrie (1993) estimates that a $1 million decline in prime contract awards lowers employment by 9 to 50 workers.24 We use our unemployment regression and VAR models to estimate the cost, in terms of national government purchasesfrom local firms,of creating local jobs and reducing local unemployment. We express these costs in present value terms using annual discount rates of 3, 5 and 10 percent.25The resulting calculations deliver estimates of the present value cost of creating one job-year and of reducing unemployment by one person-year. Table 7reports the results. When the BLS employment measureisused, the estimates indicate that, depending on the discount rate used,savingonejob requiresdefensecontract awards of $59,030 to $92,297 [column (2)]. As discussed earlier, the impact of defense spending on employment is smaller when the CPS measure is used; this translates into a range of cost-per-job estimates of $166,259 to $201,653 [column (3)]. Our regressionresultsalsopoint to considerable spillovereffectsfrom military contract awards to adjoining states in the Census division.26 Hence, an additional dollar in defense zAOtherstudies include the Committee for”Economic Development (1991), the Northeast- Midwest Congressional Coalition (1991), the Congressional Budget Office (1992), and Schmidt and Kosiak (1992). 2ST0adjust for the lag of expenditures relative to contract awards, we multiPIY the ‘aw job cost estimates by .6+ [.3/(1 + r)] + [.1/(1 + r)2], where r is the discount rate. This adjustment is in line with Bolton’s (1966) discussion and reduces the cost estimates by 1.5 to 4.5 percent. 260ur reported results are fOrunemployment rates, but significant spillover effects arise 32

spending to a state createsjobs not only in that state but in adjoining states in the census division. If thejob creation in other statesistakeninto account, the cost-per-job estimates decline markedly. Our VAR model with spillover effects implies that it costs an estimated 34,000 to 56,0001982 dollars to buy one job-year in the BLS covered sector. 6. Conclusions The story behind regional labor marketfluctuations in the postwar U.S. economy has a large cast of players: oil price shocks, military contract awards, the basing of military personnel, other national shocks with uneven effects among regions, other shocks that influence the cross-industry dispersion of demand within regions – all play important roles in at least some episodes or certain aspects of the story. But, since the early 1970s, oil price shocks have been the leading actor in the story – the most important driving force behind region-specific fluctuations in unemployment rates and employment growth. Beyond the magnitude and abruptness of oil price movements, the explanation for their pronounced regional cycle effects has three essential elements: (i) regions differ in industry mix, (ii) industries differ in sensitivity to movements in the relative price of oil, and (iii) the reallocation of productive factors across industries and regions is costly and time consuming. For example, Michigan and Indiana – states with a large concentration of employment in Transportation Equipment and Primary Metals – experience relatively high (low) unemployment rates in the aftermath of an oil price increase (decrease). This regional unemployment response tends to persist for several years. The dominant equilibrating mechansim that brings regional unemployment rates back into alignment is the net migration of people and workersbetween states. Oil shocks affect the spatial structure of factor demand through their impact on the for employent as well. 33

industry structure of demand. Some other events operate more directly on the spatial structure of demand. In this regard, we document clear roles for military contract awards and the basing of military personnel. Military basing decisions play a relatively important role asdriving forces behind unemployment fluctuations in Alaska, Hawaiiand the District of Columbia. Contract awards play a relatively important role in Delaware, Connecticut and Washington. Our preferred estimates, which account for spillover effects across state boundaries, imply a cost of local job creation (through national government purchases from local firms) equal to 34 to 56 thousand 1982 dollars per job-year based on BLS employment figuresand about twice as much based on CPS employment figures. The cost of local unemployment reductions are am order of magnitude larger. The story of regional labor market fluctuations contains important clues about the nature of aggregate business cycle fluctuations. We mention two. First, the spatial dispersion in labor market tightness varies notably over time. Between 1959 and 1992, the cross-state standard deviation of civilian unemployment rates (conditioning out state and year fixed effects) ranges from 0.8 to 2.3 percentage points. This measure of dispersion in the regional cycle component of state unemployment rates displays a clear pattern of countercyclical movements in relationship to the national business cycle. Our regression models account for much of the cyclical variation in the spatial dispersion of labor market tightness since 1973, primarily through the estimated effects of oil shocks. Second, wefind asymmetric unemployment responsesto positive and negative regional shocks. Negativeshocks – whether involving oil prices, military basing or contract awards– have a greatereffect than equal-sized positive shocks. This evidence implies that shocks to the spatial structure of demand (e.g., a reallocation of government contract awards) cause short-run increases in aggregate unemployment. Our evidence of asymmetry is similar to Hooker and Knetter’s (1996) findin~ that declines in military contract awards cause 34

larger responses in state-level employment than equal-sized increases. The evidence of asymmetry in studies of regional fluctuations is complementary to findings of aymmetric aggregate responses to oil price ups and downs in Mork (1989), Hamilton (1996), Hooker (1996) and Davis and Haltiwanger (1996,1997). 35

References Asdrubali, Pierfederico, Bent E. Sorensenand Oved Yosha (1996) “Channels of Interstate Risksharing: US 1963-1990.” Quarterly Journal of Economics, 111,no. 4,1081-1110. Barre, Robert J. and Xavier Sala-i-Martin (1991) “Convergence Across States and Regions” Brookings Papers on Economic Activity, no. 1, 107-182. Barre, Robert J. and Xavier Sala-i-Martin (1992). “Convergence,” Journal of Polztzcal Economy, 100, no. 2 (April), 223-251. Bartik, Timothy J. (1991) Who benefits from state and local economic development pohcies? W.E. Upjohn Institute for Employment Research. Kalamazoo, Michigan. Bauer, Paul W. and Susan M. Byrne (1991). “The sectoral and regional effects of oil shocks: who’s over abarrel?” Economic Commentary, FederalReserve Bank of Cleveland, July 15, 1991. Blanchard, Olivier and Lawrence F. Katz (1992). “Regional evolutions” Brookings Papers on Economic Activity, no. 1, 1-61. Blank, Rebecca and David Card (1989) “Recent Trends in Insured and Uninsured Unemployment During the 1980’s,” Industrial Relations Section, Princeton University. Bolton, Roger (1966). Defense purchases and regional growth. The Brookings Institution. Washington, D.C. Borjas, George J. and Valerie A. Ramey (1995) “Foreign Competition, Market Power, and Wage Inequality,” Quarterly Journal of Economics, 110,no. 4(November), 1075-1110. Brainard, S. Lael and David M. Cutler (1993) “Sectoral Shifts and Cyclical Unemployment Reconsidered,” Quarterly Journal of Economics, 108, no. 1, 219-243. Brown, S.P.A. and John K. Hill (1988). ‘bLfJweroil prices and state unemployment” Contemporary Policy Issues, VI, July 1988, 60-68. Burtless, Gary (1983). “Why is insured ummployment so low?” Brookings Papers on Economic Activity, no. 1, 225-249. Committee for Economic Development (1991) “The Economy and National Defense: Adjusting to Military Cutbacks in the Post-Cold War Era,” Congressional Budget Office (1992) “The Economic Effectsof Reduced Defense Spending,” Washington, DC. 36

Crump, JeffreyR. (1989) “The SpatialDistribution of Military Spending in the United States 1941-1985,” Growth and Change, Summer, 50-62. Davis, Steven J. (1986) Allocative Disturbances, Aggregate Disturbances, and Unemployment Rate Fluctuations, Brown University Ph.D. Dissertation. Davis, Steven J. (1987) “Fluctuations in the Pace of Labor Reallocation,” Carnegze- Rochester Conference Series on Public Policy, 27, 335-402. Davis, Steve J. and John Haltiwanger (1990) “Gross Job Creation and Destruction: Macroeconomic Evidence and Macroeconomic Implications,” NBER Macroeconomics Annual, 5, 123-168. Davis, Steven J. and John Haltiwanger (1996) “Driving Forcesand Employment Fluctuations,” NBER working paper no. 5775. Davis, Steven J. and John Haltiwanger (1997) “Sectoral Job Creation and Destruction Responses to Oil Price Changes and Other Shocks,” working paper, University of Chicago. Davis, Steven J. , John Haltiwanger and Scott Schuh (1996) Job Creation and Destruction. Cambridge, Massachusetts: MIT Press. DiPasquale, Denise and William C. Wheaton (1996) UrbanEconomics andRealEstate Markets. Englewood Cliffs, New Jersey: Prentice Hall. Eberts, Randall W. and Joe A. Stone (1992). Wage and employment adjustment in local Zabormarkets, W.E. Upjohn Institute for Employment Research. Kalamazoo, Michigan. Eichengreen, Barry (1990) “Currency Union,’>Eco7LomicPolicy (April 1990), 118-187. Forrest, D. and B. Naisbitt (1988) ‘(The Cyclical Sensitivity of Regional Unemployment and Unemployment Differentials.” Ik@onal Studies, 22, 149-153. Green, Gloria (1971) “Measuring Total and State Insured Unemployment,” Monthly Labor Review (June), 37-48. Guthrie, Susan (1993) “Defense spending and employment: evidence from regional data,” Harvard University working paper. Hamilton, James, (1983) “Oil and the Macroeconomy since World War II,” Journal of Political Economy, 91, 221-248. Hamilton, James (1985) “Historical Causes of Postwar Oil Shocks and Recessions,” 37

Energy Journal, 6, 97-116. Hamilton, JamesD. (1996) “This IsWhat Happened to the Oil Price - Macroeconomy Relationship,” Journal of Monetary Economics, 38, no. 2 (October), 215-220. . Hooker, Mark (1996) “What Happened to the Oil Price - Macroeconomy Relationship?” Journal of Monetary Economics, 38, no. 2 (October), 195-213. Hooker, Mark and Michael M. Knetter (1996) “The Effects of Military Spending on Economic Activity: Evidence from State Procurement Spending,” Working Paper, Dartmouth College, June. Lilien, David (1982) “Sectoral Shiftsand Cyclical Unemployment,” Journal of Political Economy, 90, 777-793. Loungani, Prakash (1986). “Oil price shocks and the dispersion hypothesis,” Review of Economics and Statistics, (August), 536-39. Loungani, Prakash, Mark Rush and William Tave (1990) “Stock Market Dispersion and Unemployment,” Journal of Monetary Economics, 25, no. 3, 367-388. Markusen, Ann, Peter Hall, Scott Campbell and Sabina Deitrick (1991) The Rise of the Gunbelt: The Military Remapping of IndustrialAmerica. New York: Oxford University Press. Mayer, Kenneth R. (1991) The Political Economy of Defense Contracting. New Haven: Yale University Press. Mork, Knut A. (1989) “Oil and the Macroeconomy When Prices Go Up and Down: An Extension of Hamilton’s Results,” Journal of Political Economy, 97, no. 3 (June), 740-744. Mork, Knut (1994) “Business Cycles and Oil Markets,” The Energy Journal, 15 (special issue), 15-38. Marston, Stephen T. (1985) “TWOViews of’ the Geographic Distribution of Unemployment,” Quarterly Journal of Economics, 100, no. 1, 57-79. Mehay, Stpehen and Loren Solnick (1990) “Defense spending and state economic growth,” Journal of Regional Science, 30, no. 4,477-487. Neumann, George and Robert Topel (1992) ‘(Employment Risk, Diversification, and Unemployment,“ Quarterly Journal of Economics, 38

Northeast-Midwest Congressional Coalition (1991) “Defenseless: Declining Dollars for the Northeast-Midwest Region,” July. Schmidt, Conrad P. and Steven Kosiak (1992) “Potential Impact of Defense Spending Reductior~ on the Defense Industrial Labor Force by State,” Defense Budget Project, March. Schmidt, Ronald H. (1989). “Natural resources and regional growth,” Economic Review, Federal Reserve Bank of San Francisco (Fall), 3-19. Thirwall, A.P. (1966) “Regional Unemployment as a Cyclical Phenomenon,” Scottish Journal of Political Economy, 13, 205-219. Topel, Robert H. and Sherwin Rosen (1988) “Housing Investment in the United States,” Journal of Politcal Economy, 96, no 4, (August), 718-740. Topel, Robert H. (1994) “Wage Inequality and Regional Labor Market Performance in the U.S” in Labor Markets and Economic Performance: Europe, Japan, and the USA, edited by Toshiaki Tachibanaki. New York: St. Martin)s Press. Vroman, Wayne (1991) “The Decline in Unemployment Insurance Claims Activity in the 1980’s,” unpublished. Wegner, Merrill (1991) “Defenseless: Declining Military Dollars for the Northeast- Midwest Region” Northeast-Midwest Congressional Coalition, July. 39

I I I 1 1 m I \ \ I I I I I I I I I I I Q I u I u I I L a) I I I I I I I I I I / 1 I I I / / I t I I E 1 U-L \ \ 5b \ \ C \ “a \ \ \ ) a) I L 3 m Cn m “c I 0 0 0 0 0 0 0 0 0 0 0 m e PI N + 4

. . h s .... \ .... ‘. s m . . -..” ... -.-.. ,, . . .. ... , .. ‘. m . . . ‘. .. .- ----- .... -. - . .” . . . . .-.> .. , , . - .... l . w . , ., , * A l.- . .( > .. ... .. .0 , . C b . . , , .....--” 5 ..-..-” c) .. ... . . 1 “ . ” .- . -- .. ..-” . . .-...... ..., .. . . -.. . v m . . % - m . . 0 . P . , t t ...-- ----- , , . * -.. ”- > . ..- Q : .. ‘. ,{ . . “ , ,,, . , ‘......... ------ .. --” --------- ..-.-”- ------ “---- ------- ----------- Is .’ ,’ ..-..’ ---- , .... . . : ; . , .’ . ,“ , ..“ ‘. , . ,, , L .’ “.’1 =$ ,. , ~.-- - ” --- k ..2 . . . . “ .. 0 . . . . . 0

- . m . m . 0 . 0 0 0 0 t ! ! ! e a) m / ‘\ \ / o I“. m I ‘ ‘> \ I : \; / \. \’. u =.”~, /, > L u ‘ -- - - / / f\ / \ / / / 1 I w al - \ a) ‘ \ \ G 1 G \ m //’ cm I (n u) 0 / w m / b c ( : \’\ u \\ //’ u c a E / I .I A / “’”” 0 / \ (’ ‘n E u a) (n c 0 % c1 ti 5 ti a) L ~ .— IL

1 1 I * * n i, , I , , , . . Q I E . c1) . % c . > , 3 , a) . < , . C5 0 5 , s 9- C , 0 . 9- U 5 Q co ln i I I t l — . ..

, t ! I [ I I , * I I I , I ! , 3 I * o , ! Q * [ * o 0 * , I . , , - E * * , , m , . ~ g I 1 I — — ———— -

9 5 I I i . i 0 I , 0 . I i , I . . . 0 . I . I . . s 0 , I ~ I I z I E c I I 1 I I i I I * C9 N T- 0 F 8 y , t

6 Employment,Unemployment,Participation response to an oil price increase 0.4 ............................................................................................................................... . . . c 8 w m o ?K M .~ w w w w w w x 1...................................................................................................................................................,..4...,.................................................................................. -0.2““ .....................................................................................................................................................................................................................................”.........,”.... -o.4- ............................................................................................................................. ................................................................................................ ...................................................................................................................................................... ‘ -1 I I I I I I i I I I 1 2 3 4 5 6 7 8 9 10 year -=- employment + unemploymentrate* participationrate I

Figure 7 . Employment,Unemployment,Participation response a , o.1-- 0 .................................................................................................................................................................................................................................................... -0.1- -oo2- .................................. -o.3- ........................................................... ....................................................... .......................................................................................................................... -o.4- -0.5 I I 1 c I I 1 I I I 1 2 3 4 5 6 7 8 9 10 year --- omploymont + unomploymonrtato* participationrate . . I

u) g) :: : :

. Fi 9 A comparison of two employment measures Response to a fall in defense contracts (1 -0.05.. ... .......... .......................... \ — * — — . 0 - 0 - 9.- -. . ,* ”———. ” ” ” . - . -—. - .. — -— -0.1 ............ . . . . . . . . . . . . . .. . . .. . . . . . . . . . .. . . . . . . . . .. .. . . .. .. . \ ... . . . . .. .. . . . . . . . . . . . . ... . . . . . . .. . . . .. . .... . .. . .... . . . . . .......”...........-.............—..............-......O.......-..”.....-...”...”.......-..........—.-o..u.—O...—.............O........ ..-., *.. .-. . . . ”. ... . . . ” .- -0.15.......7 —. - ” . .. -0.2................ ............................................... l.............................................................................. . . ...........”*,”,.,o.*”oeo”””—” . . . . . — . . “. . .. . . ”. .. \ -0.25 ............................................................................................................................ .......O............,,...O .....”.”**.....**,****,O**,O .“..”C*”-”,—”C-*.OC,.. \ -0.3 ............................ ......................................................................... ” O . e e Q.e.. ooQ”**— -0.35’ ....... . ............................... ................ ... .. ,-. " - - . o,*_* " . ,....”..-” " -—*-.* **-—-*-. - ** -0.4” “ - - ” ” ” ” “ ” K. .. ..-. —. . ... . ” -... .-. ”. .—- .. —* . ..0 .. . 0 ” . . . . . . .” -0.45” year I-W nonagr.employment+ cMfianemployment I . .

Ta 1 V c mo om Ust = as + Yt + Ed = P&&t-l+ ~st ( ( (2) (3) (4) * Civilian Insured unemployment unemployment EmpJPop.)X 100 (CPS Emp./Pop.)X1(KI D v (1958to1992) (1958to1992) (1958~oh1992) (1949?:992) Me o d v e 5. a 3 p 3 ri en 4 u 2. 1 9 5 s d o de v t e a van ri ia O 1. 1 1 2 s d e 1. 1 1 3 s d e 1. 1 9 2 standarddeviationofregionalcycle c o m po O . . . . s d i / / P ( e rr st ( ( R R (p=O N t T 1 T v c mo i e ao obt u ab no stmpl ris nw o t p t s c da se T rla r r t erg o mo= f seap Oo re t dscn- p t z Al s a c o at u ba o er a n timp we 50

Ta 2 U r r n eegm E C ( ( e p D V c u u F l o d v e ar ir pe I -0 - S l o d v e ec ar pe - . s re p ca de c on I I M t 5 IL -0 c i s p c d t e ha a m efap te pl i s e I I s e I S s d in F ( p I I M t 5 IL 3.98 2.06 O t 2 IL 48.88 53.94 (0.000) I (0.000) M t 2} I. I S t 2 T OC I o d v e ar pe ( er o d vS e at pe I rain S Er o t E t st an .645 .452 4 * * R 0.92 0.93 Ti p us i e s er 1959 toti t 1992 T n o o b um ost 1720 er 1864

N t Ta 2 ot T r i st efi ef a y fig necf re 2 T i v han b sc b tha rd st de erpeis o t 1 19 p er Pa n i {. i t nu oto l ontd va t a i i t r F e eCO t 5 me t gxta r co rte co eg va a fi la o CO T e r i t s o t esst epco o t c oa l v an t s ert i t o aasle taw t smp S a t 1’ ile i i b “g* a 5 a n1d bn “ a “ r if 9 4 I Pa B t nu h i t t c y oa la v pono t i te v c b e fr t ra xc e ri gr

3 R C C U F C D Mean standarddeviation m a d & pq p & - “dynamicV” Q O ] ] ] ] ]

Ta 4 T R Cy C C eo U om R Fneiv gi lu W T V i th ar Me S St w t f tah ( s d acva ta ev (a st (v i s i pa ~ 1. 0 W Vi (3 Mi ( Al (2 R I ( a s d t e ct vain In (2 u 1. 0 Mi (2 W Vi ( In (2 Al ( R p s d r t e e vain di Is (2 & 2. 0 Mi (3 rn ( De (2 O ( p s d - r t e e vain di S Ca ( ‘ v d er y na C o O o m po at ] 0. 0 De (0 Co ( U (0 wa ( Mi ( a tp ] 0. 0 Al (0 D C( Ha (0 Wy ( Vi ( ot$ ] 1. 0 Mi (2 In ( N Gi ( 1 S Ca ( O (2 utp ] 1. 0 Mi (2 In ( O (1 W Vi ( N Ca ( utp ] 0. 0 Wy (0 N M ( Lo (0 W Vi (0 Fl (

Ta 5 S pi Pm A Su o E C os ( e ti ( ff ( er i p s a t r en an — c u u D V . t c s d e 1 C - c re c m c s e M t 5 I L -0 -R 0 c i r p cea cl e has m t (0 af (pl O t 2 o pr g i wi s n XL tro 0. te at0 e m m p (0 lo ( M 2 r -0 x{ s i e t m nma tpl(0 du (e- S t 2 T OC 0 0 s st m dt in a i ar t (0sp (e- P B F -t an ( l is p i a gev r n en if C t 5 ON 2 3 (0 ( C t 5 O N 2 -R 3 (0 ( M t 5 IL 3 2 (0 ( M t 5 I L 0 -R 0 (0 ( O t 2 IL 5 5 (0 ( M t 2 Ix 5 1 (0 ( S t 2 T 2 OC 3 (0 ( S er o t e t s an . ti . R 0 0 Ti p us i e s er 1 t ti1 1 t 1 1 T n o o b um ot s 1 ew

. . r i e d v 2. Thein v s d f C t T d i s d 3 4 5

5

. No t Ta 6: r i e d v tO 2} S 2). TD 2. i v m 5. F C s d 5

TA 7 S a jo s i pe c ll al ‘%%, Pa A E C o u in st i em i,t’ [b o V e st (1 D ( Co p j i ( C p j sc ( C p j ra ( o C (b o B ba (c ( es w e m me me ea pmlpsp ad 9 $ -. .- $ $ E r u .e D i c e u u s a $ $ $ $ $ $ $ $ $ 150,000 Pa C E C o u r s i u ed ntie [b o O e i Ta 2 & 5 st D ( Co p pe i ( C p pe sc ( C p p ra ( o i (b o ci bans(c ( es w u ra n u er n sp mapd em 3 $ 3$ ,61, $1 $ $ $ 10% $ $ $ 1

8 I $1 II II :. Ill ‘- . II t II ‘ i i II 1- I I “- ‘ l\i II ‘ t , ,1 I t II . . I lie I I .* ‘ I \1I 1[ “ II , I 1 \ 11 l - I / .- , 1 l :1 t .- : i I . 0 : . : .v )I 1 I I t. . I \ . I i\ I 9 \\ t w / \\ : I . \\ ‘ I \ \ // , . \ “ (t I \\ II \ ‘ L I I } . I m : l l a . O9 - O : .. l - l e 99-. *” : : ; . . l . . - : 0 : . “. - .” : ..0 “ *- : ”* . , I . I, . . I 1- ; ii h 1. b1 . . . “ , I . . I I I ? , l . f.. I . l .

Fi A t . 1 I I “- II ‘ I 1 // b 1I “- // I 1- // ‘ 1 1- II / I .- II 1/ / I I b * .0 - . :0 . :0 “ .0 ”- . e - *-: ;O . :oo I , I I I 1 I I I I I I w i I I . i I 8 I II t- I : . I * . L *mm ” . .l 0 00 l“ . : :: 0 l I .:: o i ”: : : t * l “ 1 1 { I I I I I I \ I } - l *m- - . - . .*-- -0 . l “ . - . - 0- o “ 1 - - a--::-: -= 9. I . . ‘. I I t 1 I I!. ‘1 f I , 1- 1} I l . . I I I l I t . . r. ,I . I