The impact of corruption on apprehension level of immigrants: A study of the United States immigration

Buzu uko , Bilol; Lee, Byeong Wan Wo king Pape The impac o co up ion on app ehension le el o immig an s: A s udy o he Uni ed S a es immig a ion Economics Discussion Pape s, No. 2014-46 P o ided in Coope a ion wi h: Kiel Ins i u e o he Wo ld Economy – Leibniz Cen e o Resea ch on Global Economic Challenges Sugges ed Ci a ion: Buzu uko , Bilol; Lee, Byeong Wan (2014) : The impac o co up ion on app ehension le el o immig an s: A s udy o he Uni ed S a es immig a ion, Economics Discussion Pape s, No. 2014-46, Kiel Ins i u e o he Wo ld Economy (I W), Kiel This Ve sion is a ailable a : h ps://hdl.handle.ne /10419/103647 S anda d-Nu zungsbedingungen: Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen Zwecken und zum P i a geb auch gespeiche und kopie we den. Sie dü en die Dokumen e nich ü ö en liche ode komme zielle Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich machen, e eiben ode ande wei ig nu zen. So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen (insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en, gel en abweichend on diesen Nu zungsbedingungen die in de do genann en Lizenz gewäh en Nu zungs ech e. Te ms o use: Documen s in EconS o may be sa ed and copied o you pe sonal and schola ly pu poses. You a e no o copy documen s o public o comme cial pu poses, o exhibi he documen s publicly, o make hem publicly a ailable on he in e ne , o o dis ibu e o o he wise use he documen s in public. I he documen s ha e been made a ailable unde an Open Con en Licence (especially C ea i e Commons Licences), you may exe cise u he usage igh s as speci ied in he indica ed licence. h p://c ea i ecommons.o g/licenses/by/3.0/ Recei ed Oc obe 11, 2014 Accep ed as Economics Discussion Pape No embe 10, 2014 Published No embe 13, 2014 Licensed unde he C ea i e Commons License - A ibu ion 3.0 Discussion Pape No. 2014-46 | No embe 13, 2014 | h p://www.economics-ejou nal.o g/economics/discussionpape s/2014-46 The Impac o Co up ion on App ehension Le el o Immig an s: A S udy o he Uni ed S a es Immig a ion Bilol Buzu uko and Byeong Wan Lee Abs ac This pape demons a es he e ec o coun y le el co up ion on illici beha io o indi iduals in a o eign coun y. The empi ical esea ch in es iga es he p obabili y o indi iduals being app ehended o e seas due o he in luence o co up en i onmen in hei home coun ies. Using c oss-sec ional da a o empi ical analysis om 104 di e en coun ies o e he pe iod o 2009– 2011, he au ho s ocused on inding how people om a ious coun ies ac and beha e di e en ly while s a ioning ou side o hei home coun ies. Thei indings e eal some e idences ha indi iduals coming o he Uni ed S a es om co up ion- idden coun ies a e mo e likely o be app ehended han indi iduals om less co up coun ies a e. JEL F22 D73 K42 Keywo ds Immig a ion; co up ion; app ehension Au ho s Bilol Buzu uko , Depa men o Economics and Finance, Yeungnam Uni e si y, Gyongsan, 712-749, Republic o Ko ea, [email p o ec ed] Byeong Wan Lee, Depa men o Economics and Finance, Yeungnam Uni e si y, Gyongsan, 712-749, Republic o Ko ea, [email p o ec ed] Ci a ion Bilol Buzu uko and Byeong Wan Lee (2014). The Impac o Co up ion on App ehension Le el o Immig an s: A S udy o he Uni ed S a es Immig a ion. Economics Discussion Pape s, No 2014-46, Kiel Ins i u e o he Wo ld Economy. h p://www.economics-ejou nal.o g/economics/discussionpape s/2014-46 1 1 In oduc ion Co up ion is he cance o he socie y, 1 which pu s e e y possible obs acle owa ds humani y o con on wi h. Thus, i leads he na ions in o he da kness and humilia es hei igh s. The unpleasan na u e o co up ion dis o s he economic and he social li e o he coun y, which makes people su e om ha dship and injus ice. The de ini ions gi en by Wo ld Bank “The abuse o public o ice o p i a e gain” and T anspa ency In e na ional “The abuse o en us ed powe o p i a e gain” show how impo an he ole o he public o ice s and bu eauc a s a e in he socie y, who ha e he powe o make decisions on behal o he en i e na ion. One may eel cu ious why he gap among coun ies is so wide in e ms o economic pe o mance and li es yle. The majo i y o he economis s and social scien is s ag ee upon he co up ion being one o he main ac o s o he exis ence o his gap, o example, Mau o (1995) and Wei (1999) explain how co up ion nega i ely a ec s he economic g ow h. Thei indings show ha he educ ion o co up ion le el signi ican ly inc eases he economic g ow h o he coun y. The e a e dozens o excellen academic wo ks p o ing how co up ion dis o s he economy and how i leads he coun y in o mise y and po e y [e.g., Mau o (1995); Gup a e al., (1998); Tanzi (1998); TI (2008)]. The e is almos consensus ha co up ion nega i ely a ec s he coun y’s economy 2 and i s people, bu how i has an e ec on indi iduals’ ac ions and beha io while s a ioning ou side o hei home coun ies is an empi ical ques ion ha has no ound i s clea explana ion. A e he people om co up ion- idden coun ies mo e likely o be in ol ed in illegal p ac ices due o he coun y’s co up ion le el whe e illegal p ac ices as b ibe y is a common way o li e? O ? A e he people om less co up coun ies less likely o be in ol ed in illegal p ac ices due o somehow absence o co up p ac ices wi hin he coun y? The e is no pe ec answe o hese ques ions because e e y na ion has people who a e hones and dishones . Like he complexi y o measu ing he na u e o co up ion, i is also di icul o measu e he a io o hones y and dishones y o people. Acco ding o Diman e al., (2013a), hey a gue ha pe sis en co up ion in a coun y makes co up beha io a gene al a i ude among ci izens and hose emig an s om a co up ion- idden coun y may ca y some o his a i ude in o hei des ina ion coun y. 1 The s a emen o he P esiden o he Wo ld Bank James D.Wol ensohn, “People and De elopmen ”, Annual Mee ings Add ess, Oc obe 1, 1996. 2 Howe e , a numbe o e idences shows ha co up ion se es as “g easing he wheels” o he economy in pa icula ci cums ances (See, D ehe and Gassebne (2013), Meon and Sekka (2005), Meon and Weill (2010), Le (1964), Hun ing on (1968)). 2 To shed mo e ligh on his issue, we in es iga e he e ec o coun y co up ion le el (co up ion le el in a coun y o o igin) on abno mal beha io o indi iduals in a des ina ion coun y. Fo his pu pose, we ound i o be a meaning ul e o o analyze he ela ionship o app ehension le el o immig an s wi h ega d o coun y’s co up ion le el, which has no been add essed in ea lie li e a u e. 3 In ou eg ession analysis, we mainly apply se en se s o explana o y a iables hypo hesizing ha he pa icula a iables may ha e mo e gene al e ec on indi iduals’ beha io (based on hei coun y backg ound) while being ab oad, such as; co up ion (CI), weal h (GDP), human capi al (EDU), popula ion g ow h (EMPL), immig a ion s ock (IMGR), homicide a e (HMCD), and p os i u ion (PRST). 4 To check he s eng h o explana o y a iables a ec ing he dependen a iable ( he a io o app ehended immig an ss (APPR)), we pe o med eg ession on S anda dized Va iables in sec ion V, which indica es ha WEALTH has he s onges impac on APPR, ollowing CPI, EDU, EMPL, IMGR, HMCD, and PRST espec i ely. As he main objec i e o his pape is o in es iga e he e ec o coun y’s co up ion le el on app ehension le el o immig an s, we pe o m mul i- on eg ession analysis o shed ligh on his ela ionship. Fo example, he eg ession esul s o CI in he pooled sample (Table 3) come ou o be highly signi ican wi h he expec ed posi i e signs, meaning immig an s om co up ion- idden coun ies a e mo e likely o be app ehended in he Uni ed S a es. The nume ical example shows ha i El Sal ado (CI=65) could educe i s co up ion le el o ha o Saudi A abia (CI=55), he a e age numbe o app ehended Sal ado ians in he Uni ed S a es could be educed o abou 89 people. Acco ding o Na ional Immig a ion Fo um (2013), on a e age, he US go e nmen spends o e $5 million on immig a ion de en ion expenses pe day, o daily $160 pe de ained immig an . Mo eo e , TRAC-Immig a ion (2013), epo s ha Immig a ion and Cus oms En o cemen (ICE) da a ( iscal yea 2012) es ima es ha a ound 70 pe cen o de ained immig an s spend abou one mon h in de en ion cen e s. In ac , he calcula ions e eal how he app ehension o immig an s (due o abno mal beha io ) is cos ly o he Uni ed S a es’ economy. As Diman (2014) no es “Co up ion is mo e likely o impede economic p ospe i y, as de ian beha io always causes cos s, he misalloca ion o goods and se ices and e en ually leads o a down all o ma ke p inciples.” Fu he mo e, we would like o 3 A simila app oach was applied by Fisman and Miguel (2007), hey used pa king iola ions among Uni ed Na ions diploma s li ing in New Yo k Ci y as a p oxy a iable o co up ion, which measu es home coun y co up ion no ms as an impo an p edic o o p opensi y o beha e co up ly among diploma s in a o eign coun y. 4 Please, e e o sec ion 3 o he desc ip ions o he a iables. 3 poin ou ha he choice o he Uni ed S a es o ou empi ical wo k was due o he accessible da a and he exis ence o a ious na ions in i s e i o y. The pape is o ganized as ollows: Sec ion II discuses opinions and some his o ical ac s abou he Uni ed S a es immig a ion. Sec ion III e iews some ela ed li e a u e. Sec ion IV in oduces ou main da a and a iables, and hei calcula ion me hods. The ollowing sec ion (Sec ion V) ep esen s he econome ic s a egy. The las sec ion p o ides some concluding hough s. 2 Opinions and Fac s The Uni ed S a es is a unique coun y ha hos s he mos numbe o o eign bo n-popula ion, whe e people a e eage o go wi h a ious objec i es. 5 The Uni ed S a es is he land o immig an s and has a long immig a ion his o y. The Uni ed S a es immig a ion his o y jus i ies ha some na ions a i ing o he d eamland we e mo e ulne able o pe o m co up p ac ices. By a i al o I ish in 1830-1860’s he c ime a e sha ply inc eased in he Uni ed S a es. Na i e- bo n Ame icans ega ded he I ish immig an s as subs an ial con ibu o s o high a es o c ime and paupe ism and as pe haps unsui ed o li e in Ame ican socie y, especially in he ci ies. A gene a ion la e , howe e , Ame icans e ed abou he Chinese hen wi h I alian “p oblem” [e.g. Moo e and Vedde (2000)]. These his o ical ac s ell us ha he a i als o some na ions we e indeed he eason o inc ease in c iminal p ac ices in he Uni ed S a es. Spenkuch (2011) epo s ha almos h ee qua e s o Ame icans oday belie e ha immig a ion inc eases c ime a es in he coun y; he au ho inds ha a 10 pe cen g ow h in he sha e o immig an s leads o app oxima ely 1.2 pe cen inc ease in he p ope y c ime, while he a e o iolen c imes emains essen ially una ec ed. On he o he hand, na ional s udies ha e eached he conclusion ha o eign-bo n immig an s a e less likely o commi c imes han he na i e-bo n [e.g. Immig a ion Policy Cen e (2008)]. Acco ding o he su ey esul s o Ruben and Wal e (2007), he inca ce a ion a e o na i e-bo n adul s is 2.5 imes g ea e han ha o o eign-bo n men. The o eign-bo n we e less likely o be in p ison o p ope y and assaul o enses [e.g. Bu che and Anne (1999)]. Al hough, he ac s show ha he sha e o inca ce a ed immig an s is smalle han na i es, i should no be ega ded as some hing jus i iable because e e y single illegal ac has i s nega i e e ec on economy and 5 Acco ding o he Uni ed Na ions epo “T ends in In e na ional Mig an S ock: The 20 13 Re ision,” he Uni ed S a es has he la ges numbe o immig an s (45,785,090) in he wo ld wi h he sha e o 19.8 pe cen o he o al numbe o immig an s. 4 he socie y. Acco ding o Na ional Immig a ion Fo um (2013), in iscal yea 2014, he Depa men o Homeland Secu i y (DHS) o he Uni ed S a es eques ed app oxima ely wo billion dolla s in unding immig a ion o immig a ion de en ion cen e s, which has signi ican bu den on axpaye s’ shoulde s. Mo eo e , Ruben and Wal e (2007) connec ed highe a es o immig a ion in he 1990s and 2000s wi h a na ionwide d op in c ime a es. Acco ding o hei indings, he a es o iolen c ime and p ope y c ime in he Uni ed S a es ha e declined o 34.2 pe cen and 26.4 pe cen espec i ely, e en hough he numbe o illegal immig an s has doubled o 12 million since 1994. In addi ion, Robe (2008) epo ed ha i s -gene a ion immig an s in some s a es we e 45 pe cen less likely o commi iolence han hi d-gene a ion Ame icans. Fu he mo e, Cyn hia and Elizabe h (2008) a gue ha immig a ion ei he dec eases iolen c ime a es o has no e ec . The li e a u e consis en ly inds ha immig a ion has a nega i e e ec on c ime, pa icula ly homicide a e [e.g. Jacob and Rami o (2007)]. Mo eo e , acco ding o na ional su eys, he wes e n coun y wi h he highes pe cen age o ci izens who eel immig a ion is a p oblem is he Uni ed Kingdom (62 pe cen ), ollowed by 50 pe cen in he Uni ed S a es. In mode n-day U.S. economy, mos o he an i-immig a ion g oups see immig an s as he compe i o s in he labo ma ke . Acco ding o Aa on (2011), immig an s a e he ones who ake away na i es’ jobs and lowe hei wages by o e ing p oduc i e and cheap labo . On he con a y o o he majo immig an - ecei ing coun ies, immig an s in he U.S. end o be s ongly a ached o he labo o ce and ypically expe ience low unemploymen . Ne e heless, hey a e also mo e likely o wo k in low-wage and low-s a us occupa ions. E en among highly skilled immig an s, skill unde u iliza ion is widesp ead. Bo jas (1987) eco ded ha he labo supply o immig an s had signi ican ly lowe ed he ea nings o he U.S. na i e- bo n men. A one pe cen poin inc ease in he ac ion o immig a ion in an SMSA educed he wages o less-skilled na i es by oughly 1.2 pe cen [e.g. Al onji e al., (1991)]. On he o he hand, he a i al o immig an s had posi i e and e y signi ican con ibu ion o he U.S. economy and i s image. Immig an s b ing a “b ain gain” o inno a ion and c ea i i y ha ou weighs eal o imagined cos s. Th oughou he na ion’s his o y, immig an s ha e en iched economic, in ellec ual, social, and cul u al li e in he Uni ed S a es in a numbe o undamen al espec s [e.g. Da ell (2010)]. Da ell (2010) epo ed ha i e o he eigh Ame ican ci izens who ecei ed Nobel P izes in sciences in 2009 we e immig an s and o eign-bo n o en ou pe o m na i es in e ms o excep ional con ibu ions o science. I was eco ded ha 25.3 5 pe cen o he echnology and enginee ing businesses launched in he Uni ed S a es du ing 1995- 2005 had a o eign-bo n ounde . Acco ding o he Kau man Index o En ep eneu ial Ac i i y, immig an s o e he pas decade ha e displayed a high le el o en ep eneu ial spi i . O e 2006- 2008, hey we e wice as likely as na i e-bo n o s a new businesses [e.g. Robe (2013)]. Fu he mo e, he empi ical s udies indica e ha na ions’ mi g a ion is no equal in e ms o educa ion and skills. Diman e al., (2013b) demons a e obus e idence ha co up ion is among he push ac o s o mig a ion, especially o skilled mig a ion. They a gue ha skilled indi iduals make mig a ion decisions due o lowe he e u ns o educa ion, widesp ead inequali y, he lack o social ad ancemen , and he absence o a o able wo king and li ing condi ions in co up ion- idden coun ies. Mo eo e , Haque and Jahangi (1999) epo ed ha he numbe o highly skilled emig an s om A ican coun ies wi h high co up ion le el had been signi ican ly inc easing o e yea s because o poo wage policies ha ha e p omp ed mig a ion o alen . By examining he ela ionship be ween co up ion and he emig a ion a e o hose wi h high, medium and low le els o educa ional a ainmen , Coo ay and Schneide (2014) ound ha co up ion inc eases he emig a ion a e o hose wi h high le els o educa ional a ainmen , whe eas he emig a ion a e o hose wi h middle and low le els o educa ional a ainmen , inc eases a ini ial le els o co up ion and hen dec eases beyond a ce ain poin due o income inequali y, which educes hei abili y o emig a e beyond a ce ain poin . Acco ding o A iu and Squiccia ini (2013), highly skilled people a e mo e likely o mo e ab oad (ou low) when he co up ion le el in he coun y is high, mo eo e , his phenomena esul s in less in low o highly skilled immig an s om ou side. 6 People’s ision o he wo ld has b oadened wi h he ad en o global media such as ele ision and he In e ne . Those hinking abou going elsewhe e can see wha he al e na i es a e and appea o ha e ewe inhibi ions abou ese ling, especially when condi ions in hei home coun y a e no e y a o able o economic o poli ical easons [e.g. Da ell (2010)]. The absence o basic needs and jus ice in he coun y se es as he main push ac o o he immig a ion and he exis ence o hose missing needs and jus ice se es as he main pull ac o o he immig a ion. Thus, i is no ob ious ha people om co up coun ies y o look o absen oppo uni ies in o he coun ies whe e i is easy o ind. 6 These indings se e as a suppo o ou indings (explaining he nega i e ela ion be ween li e acy a e and app ehension le el) in sec ion ou . 6 3 Li e a u e Re iew We ha e ound e y sca ce li e a u e ha s udies illici beha io o immig an s in a des ina ion coun y due o co up en i onmen in he coun y o o igin (Diman e al., (2013a), Fisman and Miguel (2007), Alesina e al., (2013)). Diman e al., (2013a) epo e y in o ma i e and obus e idences ha co up ion migh mig a e o he des ina ion coun y along wi h he indi iduals emig a ing om co up ion- idden coun ies. They applied a comp ehensi e da ase consis ing o annual se ies on mig a ion lows and s ocks in o OECD coun ies om 207 coun ies o o igin o he pe iod 1975-2011. Thei eg ession esul s o pooled sample (including all coun ies) show insigni ican ou comes, hough he esul s u n ou o be weakly signi ican wi h he expec ed sing only when he immig a ion a iable is lagged by i e pe iods. Howe e , he eg ession esul s o he speci ied sample (including coun ies wi h a le el o co up ion ha is highe han he a e age) show e y signi ican ou comes when i is lagged by one pe iod, mo eo e , he coe icien alue a ec ing co up ion inc eases wi h inc easing lag s uc u e. The esul s o speci ied sample indica e ha immig a ion om highly co up coun ies inc eases he co up ion le el in he des ina ion coun y. The au ho s conclude ha gene al mig a ion does no ha e a signi ican e ec on he des ina ion coun y’s co up ion le el, while immig a ion om co up ion- idden coun ies signi ican ly inc eases he co up ion le el in he des ina ion coun y. Base on hei indings hey no e ha a pe sis en co up ion in a coun y makes co up beha io a gene al a i ude among ci izens and ha emig an s om a co up ion- idden coun y may ca y some o his a i ude in o hei des ina ion coun y. Fu he mo e, Fisman and Miguel (2007) applied e y in e es ing app oach o measu e he co up beha io o o eign indi iduals in New Yo k Ci y. As a p oxy a iable o illici beha io o indi iduals, hey used pa king beha io (unpaid pa king iola ions) 7 o Uni ed Na ions o icials in Manha an, who had diploma ic immuni y om pa king en o cemen ac ions un il 2002. Conside ing he ac o ha ing diploma ic immuni y o Uni ed Na ions diploma s, he au ho s hypo hesize he unlaw ul pa king ac ions o diploma s as he cul u al no ms, which indica e co up beha io o diploma s. Thus, hey in e p e diploma s’ beha io as e lec ing hei unde lying p opensi y o b eak ules o p i a e gain when en o cemen is no a conside a ion. Thei indings showed ha diploma s om co up ion- idden coun ies 7 F om No embe 1997 o he end o 2002 in New Yo k Ci y, diploma s accumula ed o e 150,000 unpaid pa king icke s, esul ing in ou s anding ines o mo e han $18 million. 7 accumula ed signi ican ly mo e unpaid pa king iola ions be o e he enac men o con isca ing diploma ic license pla es o iola o s in 2002. A e en o cemen au ho i ies acqui ed he igh o con isca e diploma ic license pla es o iola o s, he unpaid iola ions o Uni ed Na ions o icials sha ply dec eased. 8 The au ho s conside cul u al no ms and legal en o cemen as he key ac o s in de e mining he co up ion decisions o go e nmen o icials. 9 They ind a s ong posi i e co ela ion be ween he numbe o diploma ic pa king iola ions and he co up ion le el o home coun y, which sugges s ha co up ion no ms in he coun y o o igin a e an impo an p edic o o inclina ion o beha e co up ly among diploma s. Mo eo e , hei indings p o e ha diploma s om low-co up ion coun ies beha e ema kably well e en in he absence o legal consequences, whe eas hose om high-co up ion coun ies commi many iola ions. Mo eo e , Alesina e al., (2013) examines second-gene a ion immig an s’ ( om di e en cul u al backg ounds) gende ea men a i udes, who we e bo n and aised in he Uni ed S a es and Eu ope. They ind ha immig an s’ his o ical backg ound (cul u al belie s and no ms), which has de eloped due o in luence o ins i u ions, policies and ma ke s in he coun y o o igin, is associa ed wi h unequal gende ea men e en hough hey ace he same labo ma ke , ins i u ions, and policies. Thei indings gi e e idence ha unlike ins i u ions, policies and ma ke s, cul u al no ms and belie s a e in e nal o he indi idual. E en hough, he indi iduals emain hei ex e nal (co up ) en i onmen behind hei belie s and alues mo e along wi h hem no ma e whe e hey go. The analysis o Alesina e al., (2013)’s wo k p o ides addi ional e idence ha immig an s migh expo some o hei co up beha io in o he des ina ion coun y. Though he e is limi ed numbe o li e a u e ha s udies possible mig a ion o co up beha io o indi iduals in o he des ina ion coun y, hey p o ide aluable and obus e idences ha co up ion is impo ed in o he hos coun y along wi h he a i al o some co up immig an s. Based on e iewed li e a u e, we ound ha ou app oach in s udying co up beha io o immig an s somehow ollows he me hodologies applied in ea lie li e a u e [Diman e al., (2013a), Fisman and Miguel (2007)]. Howe e , ou app oach has signi ican di e ences om wo a ailable sou ces. Diman e al., (2013a)’s analysis (mig a ion o co up beha io ) p o ide 8 The pa king iola ions d opped by o e 98 pe cen a e en o cemen was in oduced in 2002. 9 The e a e housands o go e nmen o icials om 149 coun ies a ound he wo ld, s a ioning in New Yo k Ci y. 14 g ow h, HMCD – homicide a e, and PRST – p os i u ion. 18 The eade s a e ad ised o e e o Appendix 1 o he desc ip ion o addi ional a iables in sub- eg ession analysis. Fu he mo e, o a oid he eg ession speci ica ion e o in ou eg ession model, we applied Ramsey’s RESET Tes , 19 o check whe he he included a iables belong o he model o no . The F- alues o all o he included a iables a e signi ican in accep able signi icance le els, meaning ou model is p ope ly speci ied. We do no ind any signi ican eason o he applica ion o simul aneous-equa ion models in ou eg ession analysis, because he cause-and- e ec ela ionship be ween he app ehension le el o immig an s and he included explana o y a iables is unidi ec ional, meaning he app ehension o le el o immig an s may no cause endogenei y bias. 5.3 Reg ession on S anda dized Va iables To s a wi h, i would be insigh ul i we check he impac o he a ious explana o y a iables on he dependen a iable (APPR). Fo his pu pose (Table 2), we pe o m eg ession on S anda dized Va iables o check which explana o y a iables ha e s onge impac on he dependen a iable. The s anda dized a iables we e de i ed using he o mula 20 o S anda dized Va iables. Since, he s anda dized a iables a e equal on basis (i does no ma e in wha uni he dependen and independen a iables a e measu ed), one can di ec ly compa e he coe icien s ob ained om he OLS. The e o e, he coe icien s can be used as a measu e o ela i e s eng h o he explana o y a iables; he la ge coe icien s a e mo e ela i e o explain he dependen a iable. The coe icien s o s anda dized a iables we e de i ed by unning he ollowing eg ession: + 21 (5) Whe e, a iables wi h “*” sign ep esen s anda dized a iables. “The ad an a ge o using s anda dized a iables, o s anda diza ion pu s all a iables on equal oo ing because all s anda dized a iables ha e ze o means and uni a iances.” 22 18 No e, since he igh ans o ma ion o he da a imp o es he empi ical esul s, he a iables we e ans o med in such a way ha hey a e app op ia e o use o he log model. 19 Please, e e o Appendix 6 o he eg ession speci ica ion e o es (Ramsey’s RESET Tes ). 20 To de i e he s anda dized a iables, one should sub ac he mean alue o he a iable om i s indi idual alues and di ide he di e ence by he s anda d de ia ion o ha a iable. 21 No e: he eg ession on s anda dized a iables does no include he in e cep because i is always equals o ze o. 22 Damoda N. Guja a i “Basic s Econome ics” ou h edi ion © The McG aw−Hill Companies, 2004 page# 215. 15 Table 2 Reg ession on S anda dized Va iables Dependen a iable: APPR [2-1] [2-2] [2-3] [2-4] [2-5] [2-6] [2-7] CI 0.299329*** 0.272967*** 0.301906*** 0.298661*** 0.290285*** 0.294821*** 0.252016*** (0.0012) (0.003) (0.0012) (0.0013) (0.002) (0.0016) (0.0083) WEALTH -0.37631*** -0.37371*** -0.37557*** -0.37986*** -0.37434*** -0.36875*** -0.35994*** (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0002) (0.0002) EDU 0.268662*** 0.269476*** 0.266281*** 0.272033*** 0.26948*** 0.26603*** 0.260337** (0.0078) (0.007) (0.0088) (0.0075) (0.0079) (0.0089) (0.0103) EMPL 0.263868*** 0.229855*** 0.271728*** 0.260666*** 0.257994*** 0.24618** 0.213634** (0.0021) (0.008) (0.0023) (0.0026) (0.003) (0.0131) (0.0389) IMGR -0.24743*** -0.24457*** -0.24559*** -0.24218** -0.24419*** -0.23588** -0.21067** (0.0082) (0.0082) (0.0091) (0.0105) (0.0095) (0.0175) (0.0333) HMCD 0.223696** 0.187997** 0.220981** 0.227826** 0.212766** 0.226491** 0.150274 (0.0106) (0.0335) (0.0123) (0.0101) (0.0185) (0.0104) (0.1106) PRST 0.165963** 0.174699** 0.162378** 0.157431* 0.164454** 0.167893** 0.155539* (0.0333) (0.0239) (0.0397) (0.0508) (0.0357) (0.0325) (0.0536) ETH - 0.144653* - - - - 0.216299** (0.073) 0.0214 LING - - -0.02717 - - - -0.1198 (0.7238) 0.1781 RLG - - - -0.03305 - - -0.03104 (0.6619) 0.6882 THEFT - - - - 0.041986 - 0.059681 (0.6068) 0.4645 ALCH - - - - - -0.0349 -0.04474 (0.72) (0.6496) R-sq 0.520012 0.535899 0.520639 0.520973 0.521342 0.520658 0.550151 Adj R-sq 0.490322 0.502058 0.485686 0.486044 0.486439 0.485706 0.496365 D- s a 1.985874 2.023138 1.992343 1.980617 1.964469 1.983892 2.048469 No es: *, **, and *** indica e signi icance le els a 10%, 5%, and 1%. Numbe s in pa en heses a e he p alues. The s anda dized alues we e ob ained om logged alues o he a iables. The ank o explana o y a iables by he s eng h o hei impac on he dependen a iable shows ([2-1], Table 2) ha he WEALTH explains APPR he mos , ollowing CPI, EDU, EMPL, IMGR, HMCD, and PRST espec i ely. As we can see om he eg ession esul s, wi h all o he a iables held cons an , one s anda d de ia ion inc ease in he s anda dized WEALTH leads, on a e age, o app oxima ely -0.37631 s anda d de ia ion dec ease in he s anda dized APPR. Simila ly, holding he o he a iables cons an , one s anda d de ia ion inc ease in he s anda dized CI, on a e age, leads o oughly 0.299329 s anda d de ia ion inc ease in he s anda dized APPR. F om he esul s, we can conclude ha co up ion indeed has s ong impac on he immig an s’ app ehension le el in he Uni ed S a es. 16 As you no iced om he esul s, he EDU has posi i e sign and i is highly signi ican in 1% le el, meaning immig an s coming om coun ies wi h highe li e acy a e a e mo e likely o be app ehended. Logically, i is agains once expec a ions because li e a e people a e in ac less in ol ed in c iminal p ac ices han illi e a e people. Fo una ely, we ound somehow con incing solu ion o his puzzle. As we no ed ea lie Diman e al., (2013a) epo s ha he a io o highly skilled immig an s om co up ion- idden coun ies is highe han coun ies wi h lowe co up ion. Acco ding o a s udy by A iu and Squiccia ini (2013), hei esul s also indica e ha highly skilled people a e mo e likely o mo e ab oad i hei o igin coun y is highly co up . - These indings explain he eason why people coming o he Uni ed S a es om coun ies wi h highe li e acy a e a e mo e likely o be app ehended. Acco ding o hese ac s, he immig an s isi ing he Uni ed S a es om coun ies wi h highe li e acy a e ela i ely less educa ed and ha e ewe skills han he immig an s om co up ion- idden coun ies. The in e p e a ion o he esul s ells us ha holding he o he a iables cons an , one s anda d de ia ion inc ease in he s anda dized EDU, on a e age, leads o app oxima ely 0.268662 s anda d de ia ion inc ease in he s anda dized APPR. The popula ion g ow h (EMPL) also demons a es e y signi ican impac on he change in app ehension le el. Holding he o he a iables cons an , one s anda d de ia ion inc ease in he s anda dized EMPL (signi ican in 1% le el), on a e age, inc eases he s anda dized APPR by 0.263868 s anda d de ia ion. As o he in e na ional immig an s ock (IMGR), i has nega i e sign in 1% signi icance le el, indica ing wi h all o he a iables held cons an , one s anda d de ia ion inc ease in he s anda dized IMGR leads, on a e age, o app oxima ely -0.24743 s anda d de ia ion dec ease in he s anda dized APPR. The homicide a e (HMCD) shows ha i has posi i e ela ion wi h he APPR in 5% signi icance le el, holding he o he a iables cons an , one s anda d de ia ion inc ease in he s anda dized HMCD, on a e age, leads o oughly 0.223696 s anda d de ia ion inc ease in he s anda dized APPR. The p os i u ion (PRST) has ela i ely less s eng h on in luencing he app ehension le el among he explana o y a iables. I is signi ican in 5% le el wi h posi i e sign, holding he o he a iables cons an , one s anda d de ia ion inc ease in he s anda dized PRST, on a e age, leads o 0.165963 s anda d de ia ion inc ease in app ehension le el. The sub- eg ession models demons a e simila esul s as model [2-1]. In Model [2-2], he E hnic F ac ionaliza ion (ETH) is signi ican in 10% le el, indica ing holding he o he a iables 17 cons an , one s anda d de ia ion inc ease in ETH, on a e age, leads o 0.144653 s anda d de ia ion inc ease in he app ehension le el. The es o sub- a iables applied in sub- eg essions do no show any signi ican esul s in any accep able signi icance le els. Model [2-7] eg esses he explana o y a iables wi h he all emaining sub- a iables, he esul s a e p e y simila o he mos ou comes in he Model [2-1] excep HMCD which appea s o be insigni ican . 5.4 Tes s o biasness o he esul s The eg ession esul s o he models ha e been igo ously checked o any possible bias ou comes, using all he a ailable es ools (EVIEWS 5, e c). To check whe he he esiduals a e no mal dis ibu ed, we applied Ja cue-Be a es . As he esul s 23 indica e he esiduals a e no mally dis ibu ed in each model ( he p- alues o Ja cue-Be a es a e highe han 5% in each model, meaning ou esiduals a e no mally dis ibu ed). To check whe he he models su e om He e oskedas ici y p oblem, we used W hi e’s He e oskedas ici y Tes , o una ely, ( he Obs*R-sq and i s p- alues a e highe han 5% o each model, meaning ou models do no su e om he e oscedas ici y p oblem) he es esul s showed he homoscedas ici y o he a iances. Since he collinea i y p oblem is almos una oidable obs acle in empi ical wo ks, we we e somehow able o a oid mul icollinea i y p oblem among he explana o y a iables by ans o ming some a iables in o dummy a iables. 24 The high collinea i y be ween he explana o y a iables causes la ge a iances and co a iances, which makes he eg ession di icul o p ecisely es ima e. In Appendix 5, we discussed he a iance-in la ing ac o (VIF), o check whe he he models a e a ec ed by mul icollinea i y. 5.5 Mul iple Reg ession Analysis We apply he O dina y Leas Squa es (OLS) me hod o es ima ing he eg ession analysis. Since he e is no signi ican eason o belie e ha immig an s’ app ehension le el may cause he endogenei y bias, we do no ex end ou empi ical wo k by applying he Two S age Leas Squa es (TSLS) me hod. In ou models, he dependen a iable APPR is exp essed as a linea unc ion o he explana o y a iables. Ou assump ion is ha he cause-and-e ec ela ionship be ween app ehension le el and he included explana o y a iables is unidi ec ional. The explana o y a iables a e he cause and he dependen a iable (APPR) is he e ec . As we discussed abo e, 23 Please, e e o Appendix 4 o he esiduals es esul s. 24 Fo example, he ans o ma ion o Homicide Ra e da a in o dummy a iables signi ican ly dec eased he collinea i y p oblem be ween CI and MHCD. 18 acco ding o he MWD es , he log-linea model is he igh model o conduc ing ou eg ession analysis. Table 3 Resul s o log-linea model (pooled sample) Dependen a iable: LOG(APPR) [3-1] [3-2] [3-3] [3-4] [3-5] [3-6] [3-7] LOG(CI) 0.421175*** 0.384082*** 0.424802*** 0.420237*** 0.408451*** 0.414833*** 0.354604*** (0.0013) (0.0032) (0.0012) (0.0013) (0.0022) (0.0017) (0.0086) LOG(WEALTH) -0.15817*** -0.15708*** -0.15786*** -0.15966*** -0.15734*** -0.15499*** -0.15129*** (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0002) (0.0003) LOG(EDU) 0.690154*** 0.692245*** 0.684038*** 0.698813*** 0.692257*** 0.683393*** 0.668767** (0.0081) (0.0073) (0.0092) (0.0078) (0.0082) (0.0093) (0.0107) LOG(EMPL) 2.657087*** 2.314584*** 2.736236*** 2.62484*** 2.597931*** 2.478973** 2.15124** (0.0023) (0.0083) (0.0025) (0.0028) (0.0032) (0.0136) (0.0400) LOG(IMGR) -0.13518*** -0.13362*** -0.13417*** -0.13231** -0.13341*** -0.12887** -0.1151** (0.0085) (0.0086) (0.0095) (0.0109) (0.0099) (0.0181) (0.0343) LOG(HMCD) 0.189539** 0.159291** 0.187239** 0.193038** 0.180278** 0.191907** 0.127328 (0.0111) (0.0344) (0.0128) (0.0105) (0.0191) (0.0108) (0.1126) LOG(PRST) 0.128586** 0.135355** 0.125809** 0.121976* 0.127418** 0.130082** 0.12051* (0.0342) (0.0247) (0.0407) (0.0520) (0.0367) (0.0334) (0.0550) LOG(ETH) 0.111138* 0.166184** (0.0745) (0.0221) LOG(LING) -0.021 -0.09259 (0.7252) (0.1805) LOG(RLG) -0.02546 -0.0239 (0.6635) (0.6898) LOG(THEFT) 0.034629 0.049223 (0.6087) (0.4670) LOG(ALCH) -0.02679 -0.03435 (0.7214) (0.6514) Cons an -9.9261** -8.31739* -10.2669** -9.76608** -9.64104** -9.06388* -7.36236 (0.0319) (0.0734) (0.0307) (0.0361) (0.0392) (0.0831) (0.1724) 0.520012 0.535899 0.520639 0.520973 0.521342 0.520658 0.550151 Adjus ed 0.485013 0.496817 0.480272 0.480634 0.481034 0.480292 0.490831 D -s a 1.985874 2.023138 1.992343 1.980617 1.964469 1.983892 2.048469 F –s a 14.85788*** 13.71211*** 12.89758*** 12.91482*** 12.93393*** 12.89853*** 9.274186*** (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) Obs. 104 104 104 104 104 104 104 No es: *, **, and *** indica e signi icance le els a 10%, 5%, and 1%. Numbe s in pa en heses a e he p alues. In Table 3, he es ima ion esul s o he log-linea eg ession o he pooled sample show ha almos all he included explana o y a iables u n ou o be signi ican in 1-10% le els wi h he expec ed signs in each model. As he esul s show [3-1], holding he o he a iables cons an , he elas ici y o APPR wi h espec o CI is abou 0.42, sugges ing ha i co up ion le el goes up by 1 pe cen , on a e age, he a io o he app ehended immig an s goes up by abou 0.42 pe cen . In models [3-1]-[3-7], he a e age coe icien o CI is 0.404026, meaning by holding he o he a iables cons an , i he mean co up ion le el inc eases by 1 pe cen , on a e age, he mean a io 19 o app ehended immig an s goes up by abou 0.40 pe cen . Le us conside he ollowing nume ical example, i El Sal ado (CI=65) educes i s co up ion le el o ha o Saudi A abia (CI=55), he a e age numbe o app ehended Sal ado ians in he Uni ed S a es could be educed o abou 89 people. These esul s suppo ou expec a ion ha co up ion le el in a home coun y has signi ican e ec on i s people’s unlaw ul p ac ices while being ab oad. Table 4 Resul s o log-linea model (Less co up coun ies) Depenen a iable: LOG(APPR) [4-1] [4-2] [4-3] [4-4] [4-5] [4-6] [4-7] C -16.7193** -13.8246* -15.594** -17.3186** -16.7828** -15.925* -10.6152 (0.0206) (0.0581) (0.0376) (0.017) (0.0221) (0.053) (0.2181) LOG(CI) 0.352684** 0.321066* 0.334916* 0.358434** 0.349498** 0.348519** 0.304901* (0.0355) (0.053) (0.0507) (0.0327) (0.0432) (0.0412) (0.0751) LOG(WEALTH) -0.1883*** -0.17898*** -0.18868*** -0.18637*** -0.18903*** -0.18452*** -0.16143** (0.0019) (0.0028) (0.002) (0.0021) (0.0022) (0.0038) (0.0113) LOG(EDU) 0.796914 0.744753 0.75121 0.908754 0.797504 0.817878 0.924872 (0.198) (0.2221) (0.2317) (0.1486) (0.2029) (0.1973) (0.1447) LOG(EMPL) 4.260627*** 3.66388*** 4.081091*** 4.303791*** 4.277486*** 4.067957** 2.837968 (0.0017) (0.0078) (0.0035) (0.0016) (0.002) (0.014) (0.1022) LOG(IMGR) -0.26624*** -0.25777*** -0.27766*** -0.27248*** -0.26657*** -0.26147*** -0.27083*** (0.0028) (0.0033) (0.0026) (0.0023) (0.0031) (0.0048) (0.0051) LOG(HMCD) 0.172883 0.164249 0.17618 0.20389 0.166769 0.176416 0.262323 (0.1991) (0.2155) (0.1946) (0.1399) (0.2672) (0.1986) (0.1039) LOG(PRST) 0.106069 0.104894 0.103379 0.08667 0.107185 0.105246 0.056244 (0.1912) (0.1894) (0.2067) (0.2955) (0.1962) (0.2001) (0.5093) LOG(ETH) 0.131613 0.174987* (0.1285) (0.0809) LOG(LING) -- 0.052961 0.070007 (0.5601) (0.5424) LOG(RLG) -0.08806 -0.17122* (0.2934) (0.0757) LOG(THEFT) -- 0.011879 -0.05637 (0.9236) (0.6757) LOG(ALCH) -- -0.02443 -0.05019 (0.8359) (0.6685) 0.587327 0.609155 0.59061 0.597907 0.587416 0.587743 0.641692 Adjus ed 0.521674 0.53644 0.514445 0.523099 0.510656 0.511044 0.531444 D -s a 2.226939 2.210325 2.257447 2.164559 2.228306 2.236727 2.140215 F –s a 8.945978*** 8.377261*** 7.754292*** 7.992549*** 7.652644*** 7.662977*** 5.820419*** (0.000001) (0.000001) (0.000002) (0.000002) (0.000003) (0.000003) (0.000012) Obs 52 52 52 52 52 52 52 No es: *, **, and *** indica e signi icance le els a 10%, 5%, and 1%. Numbe s in pa en heses a e he p alues. To be mo e speci ic, i would be ai i we di ide he o al sample obse a ions in wo g oups, Less Co up Coun ies and Mo e Co up Coun ies, because ea ing di e en coun ies in one g oup may no p esen some impo an ou comes. 20 The eg ession esul s o Less Co up Coun ies (Table 4), sugges simila ou comes wi h he Pooled Sample analysis, while some a iables u n ou o be insigni ican . In models [4-1]-[4-7], he a e age elas ici y o APPR wi h espec o CI is abou 0.34, by holding he o he a iables cons an , i he mean co up ion le el goes up by 1 pe cen , on a e age, he mean a io o app ehended immig an s inc eases by abou 0.34 pe cen . Fo ins ance, i Bahamas (CI=28) educes i s co up ion le el o ha o Singapo e (CI=8), he a e age numbe o app ehended Bahamians in he Uni ed S a es could be educed o abou 60 people. Table 5 Resul s o log-linea model (Mo e co up coun ies) Depenen a iable: LOG(APPR) 5-1 5-2 5-3 5-4 5-5 5-6 5-7 C -0.62813 -0.76673 -0.9897 -0.56877 1.44265 -0.50964 0.213691 (0.9388) (0.9256) (0.9049) (0.9456) (0.8667) (0.9518) (0.9815) LOG(CI) -1.34697 -1.17913 -1.37922 -1.3466 -1.5237 -1.32298 -1.23945 (0.3409) (0.4103) (0.3341) (0.3466) (0.2893) (0.3676) (0.4159) LOG(WEALTH) -0.1487** -0.14933** -0.14673** -0.14986** -0.14571** -0.1482** -0.13871** (0.0155) (0.0155) (0.018) (0.021) (0.0182) (0.0178) (0.0388) LOG(EDU) 0.40401 0.424387 0.388833 0.403903 0.387984 0.401639 0.388441 (0.2022) (0.1838) (0.2252) (0.2076) (0.2231) (0.2125) (0.2385) LOG(EMPL) 2.489676* 2.329394 2.619403* 2.480244* 2.204833 2.443223 2.18493 (0.0816) (0.107) (0.0739) (0.0881) (0.1341) (0.1224) (0.2073) LOG(IMGR) -0.05473 -0.06175 -0.05425 -0.05326 -0.04995 -0.05227 -0.06322 (0.4751) (0.4251) (0.4826) (0.511) (0.5171) (0.5367) (0.4827) LOG(HMCD) 0.148444 0.112934 0.14924 0.149094 0.140707 0.149335 0.073103 (0.1413) (0.3022) (0.1426) (0.14630 (0.1662) (0.1465) (0.5328) LOG(PRST) 0.143897 0.162171 0.126913 0.141916 0.125002 0.145249 0.126364 (0.164) (0.1274) (0.2439) (0.1942) (0.2385) (0.1717) (0.2968) LOG(ETH) 0.082362 0.160822 (0.408) (0.1937) LOG(LING) -0.04677 -0.1146 (0.5995) (0.3074) LOG(RLG) -0.00586 0.014554 (0.9505) (0.8832) LOG(THEFT) 0.073715 0.071086 (0.4143) (0.4527) LOG(ALCH) -0.00828 -0.01041 (0.9423) (0.9321) R-sq 0.366914 0.377029 0.371008 0.366972 0.376764 0.366992 0.406148 Adj R-sq 0.266196 0.261128 0.253986 0.249199 0.260813 0.249223 0.223424 D s a 1.917539 1.899408 1.930379 1.91665 1.887395 1.917988 1.919749 F-s a 3.64298*** 3.253014*** 3.170413*** 3.115932*** 3.249335*** 3.116207*** 2.222741** P ob (0.003491) (0.005527) (0.006516) (0.007265) (0.005568) (0.007261) (0.029949) obs 52 52 52 52 52 52 52 No es: *, **, and *** indica e signi icance le els a 10%, 5%, and 1%. Numbe s in pa en heses a e he p alues. Howe e , he ou comes o he Mo e Co up Coun ies (Table 5) do no ep esen signi ican esul s ha suppo ou iew because he coe icien s o CI a e insigni ican wi h he opposi e 21 signs in each model while he es o he coe icien s a e lowe in signi icance o insigni ican a all. Ou esul s o Mo e Co up Coun ies g oup con adic s he indings o Diman e al., (2013a), hei analysis shows ha he im mig an s’ low om mo e co up coun ies is associa ed wi h signi ican inc ease in OECD coun ies’ co up ion le el. Howe e , ou esul s show ha he low o immig an s speci ically om mo e co up coun ies do no signi ican ly in luence o he co up ion le el o he Uni ed S a es. Based on ou indings we conclude ha Diman e al., (2013a)’s obus e idences migh be applicable o some OECD coun ies, bu no pa icula ly o he Uni ed S a es. Table 6 Resul s o log-linea model (Regional Dummies) Depenen a iable: LOG(APPR) [6-1] [6-2] [6-3] [6-4] [6-5] [6-6] [6-7] C -10.6148** -9.26189** -7.95994 -9.04243* -9.69231** -10.1333** -10.1297** (0.0188) (0.039) (0.1155) (0.066) (0.0356) (0.0288) (0.0295) LOG(CI) 0.33255** 0.419253*** 0.435099*** 0.404881*** 0.444069*** 0.415339*** 0.418138*** (0.0108) (0.0009) (0.001) (0.0026) (0.0008) (0.0015) (0.0014) LOG(WEALTH) -0.15458*** -0.12568*** -0.15461*** -0.15959*** -0.1606*** -0.15959*** -0.15583*** (0.0001) (0.0019) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) LOG(EDU) 0.735151*** 0.432118 0.688033*** 0.699995*** 0.685323*** 0.748069*** 0.641344** (0.004) (0.1069) (0.0084) (0.0077) (0.0084) (0.0053) (0.0185) LOG(EMPL) 2.884308*** 2.649748*** 2.223148** 2.475988*** 2.59819*** 2.647392*** 2.741484*** (0.0008) (0.0017) (0.0231) (0.008) (0.0028) (0.0024) (0.0019) LOG(IMGR) -0.17451*** -0.0943* -0.1195** -0.14157*** -0.13567*** -0.1346*** -0.13176** (0.001) (0.0681) (0.0265) (0.0076) (0.0081) (0.0088) (0.011) LOG(HMCD) 0.159521** 0.078275 0.167014** 0.20182** 0.189941** 0.204656*** 0.212029** (0.0293) (0.3427) (0.0325) (0.0101) (0.0107) (0.0074) (0.0108) LOG(PRST) 0.085912 0.075781 0.139196** 0.132416** 0.1245** 0.136367** 0.125438** (0.1592) (0.219) (0.0246) (0.0311) (0.0398) (0.0262) (0.04) LOG(EAP) -0.2143** (0.0121) LOG(LAC) 0.238814*** (0.0085) LOG(ECA) -0.07764 (0.3455) LOG(MENA) 0.057643 (0.592) LOG(NORTHA) 0.366648 (0.2022) LOG(SA) 0.126296 (0.3227) LOG(SSA) -0.07544 (0.5285) R-sq 0.550926 0.553955 0.524511 0.521469 0.528204 0.524955 0.522026 Adj R-sq 0.513109 0.516394 0.48447 0.481172 0.488474 0.484951 0.481775 D s a 1.982504 1.903976 1.941967 1.969022 2.003717 1.961786 1.978982 F-s a 14.56829*** 14.74789*** 13.09927*** 12.94054*** 13.29478*** 13.12262*** 12.96944*** P ob (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) obs 104 104 104 104 104 104 104 No es: *, **, and *** indica e signi icance le els a 10%, 5%, and 1%. Numbe s in pa en heses a e he p alues. 22 As we no ed ea lie , speci ica ion o coun ies in o less and mo e co up coun ies e eals some addi ional in o ma i e esul s. The compa ison o esul s shows ha ou expec a ions seem o be mo e applicable o less co up coun ies a he han mo e co up coun ies. In Table 6, we apply egional dummies o es whe he people coming om speci ic egions a e mo e likely o be app ehended o no . We di ided coun ies in o se en egional g oups acco ding o W o ld Bank’s egional di ision. 25 The eg ession esul s e eal ha wo egions (ou o se en) ha e signi ican impac on app ehension le el. The Models [6-1]-[6-2] indica e ha people om Eas Asia and Paci ic (signi ican in 5% le el) a e less likely o be app ehended in he Uni ed S a es, while people om La in Ame ica and Ca ibbean (signi ican in 1% le el) a e mo e likely o be app ehended. The nume ical compa ison also con i ms ha he a e age (2009-2011) numbe o app ehended immig an s om LAC (57,708) is abou 7.4 imes mo e han he app ehended immig an s om EAP (7,849) in ou sample. No e ha his a io (7.4) could be much highe i we include some coun ies wi h ex emely high numbe o app ehended people om LAC, which we e no included in ou coun y samples o a oid ou lie s. Table 7 Resul s o log-linea model (Income G oup Dummies)) Depenen a iable: LOG(APPR) [7-1] [7-2] [7-3] [7-4] C -8.62149*(0.0608) -9.44934**(0.0393) -9.93221**(0.0312) -9.30986**(0.0451) LOG(CI) 0.284956**(0.0494) 0.399269***(0.0021) 0.368144***(0.0066) 0.400893***(0.0023) LOG(WEALTH) -0.13686***(0.0008) -0.17039***(0.0000) -0.15805***(0.0001) -0.15374***(0.0001) LOG(EDU) 0.715029***(0.0056) 0.487671*(0.0837) 0.617364**(0.0198) 0.685926***(0.0085) LOG(EMPL) 2.425666***(0.005) 2.808886***(0.0012) 2.759594***(0.0016) 2.516919***(0.0041) LOG(IMGR) -0.10375**(0.0497) -0.14206***(0.0055) -0.13404***(0.0089) -0.12486**(0.0164) LOG(HMCD) 0.179617**(0.0147) 0.18225**(0.0137) 0.170645**(0.0238) 0.200149***(0.0078) LOG(PRST) 0.136946**(0.0228) 0.127851**(0.0335) 0.125356**(0.0384) 0.135206**(0.0267) LOG(HI) -0.177988*(0.0547) LOG(LI) -0.19166*(0.089) LOG(UMI) 0.084185(0.1944) LOG(LMI) 0.081213(0.2621) R-sq 0.538402 0.534484 0.528491 0.526356 Adj R-sq 0.49953 0.495283 0.488785 0.486471 D s a 2.005 1.955106 1.95972 2.006026 F-s a 13.85083 13.63435 13.31012 13.1966 P ob (0.00000) (0.00000) (0.00000) (0.00000) obs 104 104 104 104 No es: *, **, and *** indica e signi icance le els a 10%, 5%, and 1%. Numbe s in pa en heses a e he p alues. 25 Please, e e o Appendix 1 o he desc ip ion o he egional and income g oup a iables. 23 Using he same me hodology (Table 7), we di ide coun ies in o ou income g oups, such as, high-income (HI), low-income (LI), uppe -middle-income (UMI), and lowe -middle-income (LMI) g oups o check whe he he e is any associa ion o income wi h app ehension le el o immig an s. We di ided coun ies in o income g oups based on Wo ld Bank’s income g oup di ision. As we expec ed, he eg ession esul s pe o m some in o ma i e ou comes. The coe icien s o high-income (HI) and low-income (LI) coun ies a e signi ican a 10% le el wi h nega i e sign, sugges ing people coming om hese wo income g oup coun ies a e less likely o be app ehended in he Uni ed S a es while uppe middle income (UMI) and lowe middle income (LMI) g oup coun ies do no show any accep able signi ican esul s. Based on he abo e indings, we can conclude ha co up en i onmen o a coun y has signi ican ly nega i e e ec on indi iduals’ beha io , which makes hem o pe o m hei home gained immo al expe iences in a o eign coun y, by causing hem being app ehended due o he esul s o hose expe iences. To ou bes knowledge, he e ec o co up ion in his manne has no been poin ed ou in ea lie li e a u e. 6 Conclusion This pape analyzes he app ehension le el o immig an s based on he co up ion le el o hei coun y o o igin. The empi ical esul s come ou o be e y no able, indica ing ha indi iduals coming o he Uni ed S a es om co up ion- idden coun ies a e mo e likely o be app ehended han indi iduals om less co up coun ies. Fo example, i El Sal ado (CI=65) educes he co up ion le el o ha o Saudi A abia (CI=55), he a e age numbe o app ehended Sal ado ians in he Uni ed S a es could be educed o abou 89 people. F om he conduc ed empi ical wo k, we can conclude ha he app ehension le el o o eign na ions ab oad migh signi ican ly inc ease i he co up ion le el in hei home coun ies is high. Since he esul s suppo ou expec a ions, i would be wo h men ioning ha coun ies wi h highe le el o co up ion migh imp o e/dec ease hei ci izens’ beha io owa ds co up p ac ices in a o eign coun y by cu ing he disease o co up ion wi hin he coun y. Finally, conside ing all he indings, we sugges ha coun ies’ co up ion le el is he ac ha has o be se iously con olled by each coun y whe he i is less co up o mo e. 30 Appendix 6: Reg ession Speci ica ion E o Tes Ramsey’s RESET Tes 26 Res ic ed Model Un es ic ed Model M1 log(app )= c +log(cpi) 0.226193 M2 log(app )= c +log(cpi)+log(weal h) 0.350477 M2 log(app )= c +log(cpi)+log(weal h) 0.350477 M3 log(app )= c +log(cpi)+ +log(weal h)+log(empl) 0.385022 M3 log(app )= c +log(cpi)+log(weal h)+log(empl) 0.385022 M4 log(app )= c +log(cpi)+ log(weal h)+log(empl)+log(img ) 0.404831 M4 log(app )= c +log(cpi)+log(weal h)+ +log(empl)+log(img ) 0.404831 M5 log(app )= c +log(cpi)+log(weal h)+ +log(impl)+log(img )+log(edu) 0.445123 M5 log(app )= c +log(cpi)+log(weal h)+log(impl)+ +log(img )+log(edu) 0.445123 M6 log(app )= c +log(cpi)+log(weal h)+log(impl)+ +log(img )+log(edu)+log(hmcd) 0.496938 M6 log(app )= c +log(cpi)+log(weal h))+log(impl)+ +log(img )+log(edu+log(hmcd) 0.496938 M7 log(app )= c +log(cpi)+log(weal h)+log(impl)+ +log(img )+log(edu)+log(hmcd)+log(p s ) 0.520012 ( ) ( ) ( ) Un es ic ed Model-2 s Res ic ed Model-1 F- alue 19.3260038520576*** Un es ic ed Model-2 s Res ic ed Model-3 F- alue 5.61727411387074** Un es ic ed Model-3 s Res ic ed Model-4 F- alue 3.29501536538362* Un es ic ed Model-4 s Res ic ed Model-5 F- alue 7.11620052732407*** Un es ic ed Model-5 s Res ic ed Model-6 F- alue 9.99092557179831*** Un es ic ed Model-6 s Res ic ed Model-7 F- alue 4.61491537288433** The main pu pose o applying Ramsey’s RESET Tes is o ind ou whe he he included new explana o y a iables belong o he model o no . The null hypo hesis indica es ha i he calcula ed F alue is signi ican in 1%, 5%, o 10% le els, we conclude ha he Res ic ed model is mis-speci ied, o he wise, we accep he Un es ic ed model. The eade can see om he calcula ed F alues ha hey a e all signi ican in he gi en p obabili y alues. 26 The in e es ed eade is ad ised o e e o “Damoda N. Guja a i “Basics Econome ics” ou h edi ion © The McG aw−Hill Companies, 2004 Pa II, Chap e 13, page#521-523” o mo e de ailed calcula ion me hod o Ramsey’s RESET ( eg ession sp eci ica ion e o es ) Tes . 31 Re e ences Aa on T. (2011). “The Economic In eg a ion o Immig an s in he Uni ed S a es: Long- and Sho -Te m Pe spec i es.” Mig a ion Policy Ins i u e, All Righ s Rese ed, July 2011. h ps://www.google.co.k /u l?sa= & c =j&q=&es c=s&sou ce=web&cd=1& ed=0CCoQFjAA&u l=h p%3A%2F%2Fwww.mig a ion policy.o g%2Fpubs%2FEconomicIn eg a ion.pd &ei=Lc44VMX4BMj58QWmyYDQDQ&usg=AFQjCNGhUMjwg2KFOC0Kh- JJaaz7WTneNA&sig2=w5emYKqGg1X5dXb6LSb6YA&b m=b .77161500,d.dGc&cad= j Alesina, A., De leeschauwe , A., Eas e ly, W., Ku la , S., Waczia g, R. (2003). “F ac ionaliza ion.” Jou nal o Economic G ow h, 8,155-195. h p://download.sp inge .com/s a ic/pd /535/a %253A10.1023%252FA%253A1024471506938.pd ?au h66=1413244513_6e3 398953 d399b3142174a155ed2ab&ex =.pd Alesina, A., Giulina, P., Nunn, N. (2013). “On he O igins o Gende Roles: Women and he Plough,” Qua e ly Jou nal o Economics 128(2): 469-530. h p://www.econ.no hwes e n.edu/semina s/Nemme s11/Giuliano.pd Al onji, J. G., & Da id C. (1991). “The E ec s o Immig a ion on he Labo Ma ke Ou comes o Less-Skilled Na i es,” in John M. Abowd and Richa d B. F eeman, eds., Immig a ion, T ade, and he Labo Ma ke . Chicago: Uni e si y o Chicago P ess, pp. 201-234. h p://da idca d.be keley.edu/pape s/na i es.pd A iu A., & Squiccia ini P. (2013). “The balance o b ains: co up ion and high skilled mig a ion.” Ca holic Uni e si y o Lou ain, IRES Discussion Pape No.10. h p://si es.uclou ain.be/econ/DP/IRES/2013010.pd Bo jas, G.J. (1987). “Immig an s, Mino i ies, and Labo Ma ke Compe i ion,” Indus ial and Labo Rela ions Re iew 40: 382-392. h p://www.hks.ha a d.edu/ s/gbo jas/publica ions/jou nal/ILRR1987.pd Bu che , K. F., & Anne M. P. (1999). “The Role o Depo a ion in he Inc a ce a ion o Immig an s.” NBER Wo king Pape No. 6974 h p://www.nbe .o g/chap e s/c6060.pd Bu che , K. F., & Anne M. P. (2005). “Why A e Immig an s’ Inca ce a ion Ra es So Low? E idence on Selec i e Immig a ion, De e ence, and Depo a ion,” (Chicago, IL: Fede al Rese e Bank o Chicago) Wo king Pape 2005-19. h p://www.nbe .o g/pape s/w13229.pd Coo ay, A., & Schneid e , F. (2014): “Does Co up ion P omo e Emig a ion? An Empi ical Examina ion.” IZA Discussion Pape No. 8094 h p:// p.iza.o g/dp8094.pd Cyn hia, B., & Elizabe h, V. (2008). “Beyond My hs and S e eo ypes Fac s abou Immig a ion and C ime Fac s abou Immig a ion and C ime.” Coali ion o Humane Immig an Righ s o Los Angele (CHIRLA), Oc obe 2008. h p://chi la.o g/ iles/Fac shee Immig a ionanC ime.pd Da ell, M. W. (2010). “B ain Gain – Re hinking U.S. immig a ion policy”. B ookings Ins i u ion P ess, Washing on, D.C. 2010 h p://www.b ookings.edu/ esea ch/books/2010/b aingain Diman , E. (2014). “Th e Na u e o Co up ion - An In e disciplina y Pe spec i e” (Cen e o In e na ional Economics, Wo king Pape Se ies). No. 2014-06 h p://g oups.uni-pade bo n.de/wp-wiwi/RePEc/pd /ciepap/WP79.pd Diman , E., K iege T., & Ma ga e e R. (2013a). “A C ook is a C ook . . . Bu is He S ill a C ook Ab oad?” On he E ec o Immig a ion on Des ina ion-Coun y Co up ion, Wo king Pape Se ies, No. 2013-14 h ps://www.wgu h.uni- eibu g.de/ o schung/da a/da a/wgsp_dp_2013_03 Diman , E., K iege , T., & Meie ieks, D. (2013b). “The e ec o co up ion on mig a ion, 1985-2000.” Applied Economics Le e s 20: 1270- 1274. h ps://wiwi.uni-pade bo n.de/ ileadmin/leh s uehle/depa men -4/wiwi-dep-4-ls-8/Team/Diman _K iege _Meie ieks__2013__- _The_E ec _o _Co up ion_on_Mig a ion__1985-2000.pd Docquie a, F. & Khalid S. (2006). “B ain d ain and inequali y ac oss na ions”. Discussion Pape Se ies. IZA DP No. 2440 h p://de 3.cep .o g/mee s/wkcn/4/4556/pape s/docquie .pd D ehe , A., & Gassebne , M. (2013). “G easing he wheels? The impac o egula ions and co up ion on i m en y.” Public Choice, June 2013, Volume 155, Issue 3-4, pp 413-432 h p://cen au . eading.ac.uk/35196/1/a %253A10.1007%252Fs11127-011-9871-2.pd Fisman, R., & Miguel, E. (2007). "Co up ion, No ms and Legal En o cemen : E idence om Diploma ic Pa king Ticke s." Jou nal o Poli ical Economy 115 (6): 1020-1048 h p://eml.be keley.edu/~emiguel/pd s/miguel_pa king.pd Gup a, S., Da oodi, H., & Te me, R. A. (1998). “Does Co up ion A ec Income Equali y and Po e y?” Washing on, DC: IMF, 1998, IMF Wo king Pape Se ies, No. 98/76 h p://www.im .o g/ex e nal/pubs/ /wp/wp9876.pd 32 Haque, N. U. & Jahangi A. (1999), ”The quali y o go e nance: second-gene a ion ci il e o m in A ica”, Jou nal o A ican Economies 8, 65- 106. h p://www.im .o g/ex e nal/pubs/ /wp/wp98164.pd Hooghe, M., Vanhou e, B., Ha dyns, W., & Bi can, T. (2010). “Unemploymen , Inequali y, Po e y and C ime: Spa ial Dis ibu ion Pa e ns o C iminal Ac s in Belgium, 2001-2006”, he B i ish Jou nal o C iminology, (2011) 51, 1–20 h p://www.socialcohesion.be/C ime%20Belgium%2031%20May%202010.pd Hun ing on, S. P. (1968). “Poli ical o de in changing socie ies.” New Ha en: Yale Uni e si y P ess, 2006 - Poli ical Science - 488 pages h p://p ojec s.iq.ha a d.edu/go 2126/ iles/hun ing on_poli ical_o de _changing_soc.pd Immig a ion Policy Cen e , (2008). “F om Anecdo es o E idence: Se ing he Reco d S aigh on Immig an s and C ime,” Ame ican Immig a ion Council. Washing on D.C. Sep embe 2008 h p://www.immig a ionpolicy.o g/si es/de aul / iles/docs/Se ing heReco dS aigh onImmig an sandC ime9-10-08.pd Jacob, I. S., & Rami o, M. (2007). “Displaced, Dispossessed o Lawless? Examining he Link Be ween E hnici y, Immig a ion, and Violence.” Agg ession and Violen Beha io 12 (2007): 564-581. h p://www.sciencedi ec .com/science/a icle/pii/S1359178907000250 Kau mann, D. (2004). “Co up ion, Go e nance and Secu i y: Challenges o he Rich Coun ies and he Wo ld”, Chap e in he Global Compe i i eness Repo 2004. h p://unpan1.un.o g/in adoc/g oups/public/documen s/un-dpadm/unpan044780.pd Le , N. H. (1964). “Economic de elopmen h ough bu eauc a ic co up ion.” Ame ican Beha io al Scien is , 8(3), 8–14. h p://abs.sagepub.com/con en /8/3/8.abs ac Mau o, P. (1995). “Co up ion and G ow h” Qua e ly Jou nal o Economics 110(3): 681-712. h p://homepage.n u.edu. w/~kslin/mac o2009/Mau o%201995.pd Mau o, P. (1997). “The E ec s o Co up ion on G ow h, In es men , and Go e nmen Expendi u e: A C oss-Coun y Analysis,” in Co up ion and he Global Economy, edi ed by Kimbe ly Ann Ellio , Washing on: Ins i u e o In e na ional Economics. h p://www.piie.com/publica ions/chap e s_p e iew/12/4iie2334.pd Meon, P.G., & Sekka , K. (2005). “Does co up ion g ease o sand he wheels o g ow h?” Public Choice, 122(1), 69–97. h p://link.sp inge .com/a icle/10.1007%2Fs11127-005-3988-0 Meon, P.G., & Weill, L. (2010). “Is co up ion an e icien g ease?” Wo ld De elopmen , 38(3), 244–259. h p://www.sciencedi ec .com/science/a icle/pii/S0305750X09001193 Moo e, S., & Vedde , R. (2000). “The Immig a ion P oblem: Then and Now,” The Independen Re iew, Vol 4, n°3,2000 h p://www.independen .o g/pd / i / i _04_3_ edde .pd Na ional Immig a ion Fo um S udy (2013). “The Ma h o Immig a ion De en ion: Runaway Cos s o Immig a ion De en ion Do No Add Up o Sensible Policies,” P ac ical Solu ions o Immig an s and o Ame ica, Augus 2013. h p://www.immig a ion o um.o g/images/uploads/ma ho immig a ionde en ion.pd Robe , J. S. (2008). “Re hinking C ime and Immig a ion”. Con ex s 7: 28-33. h p://www.law.yale.edu/documen s/pd /Clinics/Sampson_Immig a ion_Con ex s.pd Robe , W. F. (2013). “Kau man Index o En ep eneu ial Ac i i y, 1996-2012,” 2013 by Ewing Ma ion Kau man Founda ion. h p://www.kau man.o g/wha -we-do/ esea ch/kau man-index-o -en ep eneu ial-ac i i y Ruben, G. R., & Wal e , A. E. (2007). “The My h o Immig an C iminali y and he Pa adox o Assimila ion: Inca ce a ion Ra es among Na i e and Fo eign-Bo n Men” (Washing on: DC: Immig a ion Policy C en e , Ame ican Immig a ion Law Founda ion ), pp. 12-14. h p://www.immig a ionpolicy.o g/si es/de aul / iles/docs/Imm%20C iminali y%20(IPC).pd Saisana, M., & Sal elli, A., (2012). “Co up ion Pe cep ions Index 2012 S a is ical Assessmen ,” T anspa ency In e na ional. h p://publica ions.j c.ec.eu opa.eu/ eposi o y/bi s eam/111111111/27386/1/lbna25623enn.pd Se a, D. (2006). “Empi ical De e minan s o Co up ion: A Sensi i i y Analysis.” Public Choice 126(1/2): 225-256. h p://www.gp g.o g/pubs/wo kingpape s/pd s/gp g-wps-012.pd Shlei e , A. & Robe , W. V. (1993 ). “Co up ion” The Qua e ly Jou nal o Economics 108 (3), 599–617. h p://p ojec s.iq.ha a d.edu/go 2126/ iles/shlei e _and_ ishy.pd Spenkuch , J. L. (2011). “Unde s anding he Impac o Immig a ion on C ime” Uni e si y o Chicago. h p://www.kellogg.no hwes e n.edu/ acul y/spenkuch/ esea ch/immig a ion_c ime.pd Tanzi, V. (1998). “Co up ion a ound he Wo ld - Causes, Consequences, Scope, and Cu es.” IMF S a Pape s 45(5): 559-594. h p://www.im .o g/ex e nal/pubs/ /wp/wp9863.pd 33 The US Census Bu eau (2006-2011). Ame ican Communi y Su eys (ACS), Table B05006 "Place o Bi h o he Fo eign-Bo n Popula ion"; Decennial Census 2000, Summa y File 3, Table QT-P15. "Region and Coun y o A ea o Bi h o he Fo eign-Bo n Popula ion: 2000." h p://www.census.go /acs/www/da a_documen a ion/2010_ elease/ The Wo ld Bank (1997). “The Role o Wo ld Bank in Helping Coun ies Comba C o up ion,” Po e y educ ion and economic managemen , Sep embe 1997. h p://www1.wo ldbank.o g/publicsec o /an ico up /co up n/co p n.pd TRAC-Immig a ion (2013). “Legal Nonci izens Recei e Longes ICE De en ion,” June, 2013 h p:// ac.sy .edu/immig a ion/ epo s/321/ T anspa ency In e na ional (2008 ). “Po e y and co up ion” Wo king Pape Se ies, No.02, 2008 h p://www.ci icus.o g/new/media/Po e y- Co up ion-Ma ch-2008.pd T eisman, D., (2000). “The causes o co up ion: A c oss-na ional s udy.” Jou nal o Public Economics, Volume 76, pp. 3 99-457. T ends in In e na ional Mig an S ock: The 2013 Re ision h p://www.is .umich.edu/cps/pewpa/a chi e/a chi e_98/19980019.pd Wa ne , J.A. (2005). “The social cons uc ion o he c iminal alien in immig a ion law, en o cemen and p ac ice and s a is ical enume a ion: consequences o immig an s e eo yping,” Jou nal o Social and Ecological Bounda ies, 1 (2), 56–80 h p://www. amiu.edu/coas/jseb/12 iles/Wa ne .pd Wei, Sh. J. (1998). “Co up ion in Economic De elopmen : Bene icial G ease, Mino Annoyance, o Majo Obs acle?” Ha a d Uni e si y and Na ional Bu eau o Economic Resea ch. Wo king Pape Se ies No. 2048 h p://www.an i-co . u/a chi e/Co up ion%20in%20Economic%20De elopmen %20Bene icial%20G ease%20Mino %20A.pd Wol ensohn, J. D. (1996). “People and De elopmen ,” Annual M ee ings Add ess by he P esiden o he Wo ld Bank. Oc obe 1, 1996. h p://web.wo ldbank.o g/WBSITE/EXTERNAL/EXTABOUTUS/ORGANIZATION/EXTPRESIDENT/EXTPASTPRESIDENTS/PR ESIDENTEXTERNAL/0,,con en MDK:20025269~menuPK:232083~pagePK:159837~piPK:159808~ heSi ePK:227585,00.h ml Please no e: You a e mos since ely encou aged o pa icipa e in he open assessmen o his discussion pape . You can do so by ei he ecommending he pape o by pos ing you commen s. Please go o: h p://www.economics-ejou nal.o g/economics/discussionpape s/2014-46 The Edi o © Au ho (s) 2014. Licensed unde he C ea i e Commons A ibu ion 3.0.