T HE C E R NT E C ANNOT H OL D : P AT RNS OF T E P OL ION ARIZAT - PowerPoint PPT Presentation

T HE C E R NT E C ANNOT H OL D : P AT RNS OF T E P OL ION ARIZAT IN N IGE RIA F . Cle me nti 1 , A. L . Da b a le n 2 , V. Mo lini 2 , a nd F . Sc he ttino 3 1 Unive rsity o f Mac e rata, Mac e rata, I taly 2 Wo rld Bank, Washing to n DC, USA 3 Se c o nd Unive rsity o f Naple s, Naple s, I taly

O VE W RVIE • I ION NT RODUCT  Nige r ia: T he GIP T r ilogy and Be yond  Aim of the Wor k • D AT A AND M E OGY T HODOL  T he Ne e d for Compar able Data  T he Nige r ian House hold Consumption Data  E nabling Data Compar ison for Nige r ia  T he Re lative Distr ibution for Ine quality Que stions  T he Re lative Distr ibution: Some Advantage s • R S E SUL T  Change s in the Nige r ian Consumption Distr ibution • C ONCL UDING R MARKS E • R S E F E RE NCE • A CKNOWL S E DGME NT UNU-WI DE R Co nfe re nc e o n "I ne q ua lity ‒ Me a sure me nt, T re nds, I mpa c ts, a nd Po lic ie s", He lsinki (F inla nd), 5 Se pte mb e r 2014 2

I ION NT RODUCT UNU-WI DE R Co nfe re nc e o n "I ne q ua lity ‒ Me a sure me nt, T re nds, I mpa c ts, a nd Po lic ie s", He lsinki (F inla nd), 5 Se pte mb e r 2014 3

Nige r ia: T he GIP T r ilogy and Be yond • Re c e nt a na lysis o n c o nsumptio n pa tte rns in Nig e ria (e .g . Wo rld Ba nk, 2013) se e ms to sug g e st a n inc re a se in “I ”-ne q ua lity tha t c o uld ha ve o ffse t the b e ne fits fro m susta ine d “G”-ro wth in te rms o f “P”-o ve rty re duc tio n. • I ne q ua lity inc re a se is, ho we ve r, just o ne a spe c t o f the who le pro b le m; o ur hypo the sis is tha t Nig e ria is a lso g o ing thro ug h a pro c e ss o f e c o no mic po larizatio n . • Bro a dly spe a king , the no tio n o f po la riza tio n is c o nc e rne d with the disa ppe a ra nc e o f the middle c la ss (e .g . F o ste r a nd Wo lfso n, 1992, a nd Wo lfso n, 1994, 1997); it c a n a lso b e re g a rde d a s the “c luste ring ” o f a po pula tio n a ro und two o r mo re po le s o f the inc o me / c o nsumptio n distrib utio n, whic h mig ht g ive rise to so c ia l c o nflic ts a nd te nsio ns (E ste b a n a nd Ra y, 1999, 2008, 2011). • Nig e ria re pre se nts a n inte re sting c a se fo r unde rta king a po la riza tio n a na lysis: GDP pe r c a pita ha s ste a dily g ro wn in the la st de c a de a nd Nig e ria is like ly to b e c o me the b ig g e st Afric a n e c o no my, b ut ye t c le a r sig ns o f c o nso lida tio n o f a na tio na l middle c la ss a re limite d; mo re o ve r, the c o untry is inc re a sing ly a ffe c te d b y sub -re g io na l c o nflic ts. UNU-WI DE R Co nfe re nc e o n "I ne q ua lity ‒ Me a sure me nt, T re nds, I mpa c ts, a nd Po lic ie s", He lsinki (F inla nd), 5 Se pte mb e r 2014 4

Aim of the Wor k • Studie s o n po la riza tio n in Nig e ria a re still fe w (e .g . Ara a r, 2008, Awo ye mi a nd Ara a r, 2009, a nd Awo ye mi e t a l., 2010) a nd the limite d a va ila b ility o f c o mpa ra b le da ta se rie s no t to o c lo se in time ha s hinde re d a n inve stig a tio n o f the lo ng -run pa tte rns – the pro c e ss o f po la riza tio n is inde e d g e ne ra lly slo w a nd sig nific a nt c ha ng e s c a n b e de te c te d o nly o ve r lo ng pe rio ds. • T he pre se nt study is inno va tive unde r a t le a st thre e a spe c ts: 1) unlike pre vio us studie s, ra the r tha n just c o mputing a nd c o mpa ring po la riza tio n indic e s, we a pply a no n-pa ra me tric to o l (the “re la tive distrib utio n”) to e xplo re po la riza tio n a lo ng the e ntire distrib utio n; 2) sinc e the re la tive distrib utio n a na lysis re q uire s a t le a st two c o mpa ra b le surve y ro unds, we use “surve y-to -surve y” imputa tio n te c hniq ue s to pro duc e fully c o mpa ra b le distrib utio ns; in this wa y we o ve rc o me the la c k o f c o mpa ra b le surve ys tha t limite d the sc o pe o f pre vio us wo rk. 3) sinc e Nig e ria is hig hly he te ro g e ne o us a c ro ss ma c ro re g io ns, we a lso a im a t do c ume nting sub -na tio na l pa tte rns o f po la riza tio n; the fle xib ility o f the re la tive distrib utio n a ppro a c h a llo ws inde e d a n a c c ura te a na lysis a t the ma c ro -re g io na l le ve l to o . UNU-WI DE R Co nfe re nc e o n "I ne q ua lity ‒ Me a sure me nt, T re nds, I mpa c ts, a nd Po lic ie s", He lsinki (F inla nd), 5 Se pte mb e r 2014 5

D AT A AND M E OGY T HODOL UNU-WI DE R Co nfe re nc e o n "I ne q ua lity ‒ Me a sure me nt, T re nds, I mpa c ts, a nd Po lic ie s", He lsinki (F inla nd), 5 Se pte mb e r 2014 6

T he Ne e d for Compar able Data • E xc luding c a se s o f sudde n sho c ks, me a sure s suc h a s ine q ua lity, po ve rty o r po la riza tio n te nd to mo ve slo wly in time ; c o mpa riso n o f suc h me a sure s, c o mpute d o n surve ys re la tive ly dista nt in time , c a pture s the re fo re mo re a c c ura te ly the e ffe c ts o f struc tura l mo dific a tio ns in the distrib utio n o f a we lfa re va ria b le . • Co mpa riso ns o ve r a lo ng time spa n, ho we ve r, c a n b e ma de diffic ult – if no t impo ssib le – b y c ha ng e s in da ta c o lle c tio n me tho do lo g y (T a ro zzi, 2007). • I n pa rtic ula r, the re is inc re a sing e vide nc e tha t q ue stio nna ire s re visio ns c a n a ffe c t re spo nde nts’ a nswe r in re le va nt wa ys (e .g . De a to n a nd Gro sh, 2000). • Othe r c ha ng e s suc h a s the switc h fro m a dia ry-b a se d c o lle c tio n to a re c a ll- b a se d c o lle c tio n c a n dra ma tic a lly a lte r a g g re g a te fo o d c o nsumptio n e xpe nditure s, a re le va nt c o mpo ne nt o f to ta l e xpe nditure s in ma ny de ve lo ping c o untrie s (e .g . Be e g le e t a l., 2010, a nd Ahme d e t a l., 2014). • F o r o ur spe c ific c a se , sinc e we a do pt a me tho do lo g y b a se d o n c o mpa riso n o f two distrib utio ns, it b e c o me s c ruc ia l using distrib utio ns suffic ie ntly dista nt in time in o rde r to se e sig nific a nt diffe re nc e s; the ne e d fo r c o mpa ra b le da ta re q uire s thus to b e dire c tly a ddre sse d. UNU-WI DE R Co nfe re nc e o n "I ne q ua lity ‒ Me a sure me nt, T re nds, I mpa c ts, a nd Po lic ie s", He lsinki (F inla nd), 5 Se pte mb e r 2014 7

T he Nige r ian House hold Consumption Data • T he Nig e ria ’ s Na tio na l Bure a u o f Sta tistic s (NBS) ha s g e ne ra lly use d the Na tio na l L iving Sta nda rd Surve y (NL SS) 2003/ 2004 a nd the Ha rmo nize d Na tio na l L iving Sta nda rd Surve y (HNL SS) 2009/ 2010 to mo nito r pro g re ss in po ve rty re duc tio n o ve r the la st de c a de . • T he NBS a lso c o nduc ts o the r ho use ho ld surve ys, mo st no ta b ly the Ge ne ra l Ho use ho ld Surve y (GHS) c ro ss se c tio n a nd pa ne l. • T he GHS c ro ss se c tio n is a surve y o f 22,000 ho use ho lds c a rrie d o ut pe rio dic a lly thro ug ho ut the c o untry; da ta o n c o nsumptio n a re c o lle c te d b y a sking the ho use ho ld a b o ut b ro a d c a te g o rie s o f c o nsume d ite ms in the la st mo nth: fo o d , he a lthc a re , sc ho o l, a nd so fo rth; a va ila b le da ta se ts inc lude 6 ro unds, fro m 2004/ 2005 to 2010/ 2011. • T he GHS pa ne l is a ra ndo mly se le c te d sub -sa mple fro m GHS c ro ss se c tio n c o nsisting o f 5,000 ho use ho lds a nd c o ve ring the pe rio ds 2010/ 2011 (Wa ve 1) a nd 2012/ 2013 (Wa ve 2); it is re pre se nta tive a t na tio na l a nd zo na l (g e o - po litic a l) le ve ls; in e ve ry pa ne l wa ve , ho use ho lds a re inte rvie we d two time s: o nc e in the “po st-pla nting ” pe rio d (fro m Aug ust to No ve mb e r) a nd o nc e in the “po st-ha rve sting ” pe rio d (fro m F e b rua ry to April). UNU-WI DE R Co nfe re nc e o n "I ne q ua lity ‒ Me a sure me nt, T re nds, I mpa c ts, a nd Po lic ie s", He lsinki (F inla nd), 5 Se pte mb e r 2014 8