asp.net mvc display pdf : Reader pdf reduce file size control SDK system web page winforms windows console hivexports1-part142

3.3 DataonControls
Intheprimaryspecications,Icontrolforcountryandyearxedeects,andforlogofGDPper
capita,whichisdrawnfromtheWDI.Inaddition,Iexploretherobustnessoftheresultstocontrols
foraveragetemperatureandprecipitationbythecountry-year,whicharetakenfromDell,Jonesand
Olken(2008). SummarystatisticsforthesemeasuresarealsoprovidedinTable1.
4 Results: : HIVIncidenceandExports
Thissectionpresentstheprimaryresultsinthepaper,estimatingtherelationshipbetweenexports
andnewHIVinfections. Subsection3.1belowpresentsthebasicresultsonHIVandexports,and
Subsection3.2discussesissuesofcausality.Subsection3.3discussestherobustnessofthesebasic
resultstotheassumptionsusedtogenerateHIVincidence.
4.1 PrimaryResults
Table2showtheprimaryresultsinthepaper.Columns1-3focusontheUNAIDSdata.Column1
includestheminimumsetofcontrols:laggedHIVrate,countryandyearxedeects.The
coecientonexportsispositiveandsignicant:increasingexportsincreasestheincidencerate.
Column2presentstheprimaryspecication,whichincludesacontrolforGDP.Therelationship
betweenexportsandHIVisstillpositiveandhighlysignicant,andofasimilarmagnitude.Finally,
Column3alsoincludes(withaslightlylimitedsample)controlsforaverageyearlytemperatureand
precipitation.Therelationshipremainslargeandsignicant;thecoecientactuallygetslarger,
whichisdueentirelytothechangeinthesample.Columns4-6replicatetheseregressionsusingthe
mortality-basedmeasureofincidence.AswiththeUNAIDSdata,thecoecientonexportsinthese
regressionsispositiveandsignicant;inthiscase,thecoecientisquiteabitlarger,althoughthe
standarderrorsadmitmagnitudesclosetothesizeseenwiththeUNAIDSdata.
Togiveavisualsenseofthedata,AppendixBshowsgraphsofexportsandincidence,over
time,forallofthecountriesinthesample.InsomecaseswecanseebothUNAIDSincidenceand
mortality-basedincidence;inothersonlytheUNAIDSdataisobserved.Ingeneral,thesegures
showevidenceconsistentwithwhatweseeinTable2{exportsandincidencearemovingtogether.
Thisis,ofcourse,notuniversallytrue,butdoesseemtobeevidencedinalargenumberofcountries.
InotedinthediscussioninSection2that,giventhespecicationused,wedonotexpect
theeectofexportstovarywithpreviousyear’sHIVrate.Thisisclearfromequation(5),andis
11
Reader pdf reduce file size - Compress reduce PDF size in C#.net, ASP.NET, MVC, Ajax, WinForms, WPF
C# Code & .NET API to Compress & Decompress PDF Document
pdf change page size; change font size in pdf comment box
Reader pdf reduce file size - VB.NET PDF File Compress Library: Compress reduce PDF size in vb.net, ASP.NET, MVC, Ajax, WinForms, WPF
VB.NET PDF Document Compression and Decompression Control SDK
change font size in pdf fillable form; change file size of pdf
truedespitethefactthatlaggedprevalenceisanimportantinputingeneralindrivingincidence.
However,itisneverthelessinterestingtoconsiderhowthecoecientsdieracrosscountrieswith
dierentoveralllevelsofHIV.Amongotherthings,bothmethodologiesforgeneratingHIVdataare
likelytobemorepreciseforcountrieswithhigherHIVrates,wherethedeathswillbemoreobvious
andthereismoretestingdone;thiscouldleadtostrongereectsforhigherHIVcountries.Table3
replicatesColumns2and5ofTable2,butsplitsthesampleinhalf,intohighandlowHIVcountries
basedontheiraverageHIVrateovertheentireperiod.ForbothmeasuresofHIVratethe
relationshipinhigh-HIVcountriesislargerthaninlow-HIVcountriesconsistent,perhaps,withless
measurementerrorinincidence.
Oneimportantissueistowhatextentindividualcountriesaredrivingtheseresults;thisis
likelytobeamoresignicantissueintheanalysiswiththemortalitydata,wherewehaveonly
elevencountries.AppendixTable1reportscoecientsfromregressionsoftheforminColumns2
and5ofTable2excludingeachcountryinthesample. Eveninthecaseofthemortality-baseddata,
thecoecientsandsignicancedonotseemtobeverysensitivetotheexclusionofanyofthe
countriesinthesample.
Finally,beforeturningtodiscusscausality,weconsiderseveralauxiliarymeasuresoftrade.
Table4reportsresultsfromregressionsofHIVontheothertrademeasures{valueandvolumefor
majorexports,andimports. PanelAshowtheresultsfortheUNAIDSdata;PanelBshowsthe
resultsusingthemortality-baseddata.Thesecoecientsareconsistentlypositive,andforthemost
partatleastmarginallysignicant.TheyarelessprecisethantheresultsusingtheWDIdata,
perhapsduetothefactthatwhenlookingatmajorexportsonlywemissasubstantialportionof
exportactivity,perhapstheportionthatismostvariable. Nevertheless,thefactthatthese
coecientsarepositiveandinmanycasessignicantdoessuggesttheresultsarenot\special"tothe
WDIexportdata.
Therelationshipwithexportsestimatedhereisfairlylargeinmagnitudeand,hence,the
overalleectofchangesinexportsonHIVmaybebig.BasedonthecoecientsinTable2,Icreate
asimplecounterfactual:ifexportsinallyearsinthesamplewerelowerby25%,whateectwould
thathaveonHIVinfections?Althoughtheexactvaluedependsonwhichmeasureofexportsisused,
onaveragethecounterfactualsuggeststhata25%declineinexportswouldmeanareductionof
between20%and50%inthenumberofnewinfections.
12
VB.NET Image: Image Resizer Control SDK to Resize Picture & Photo
easy and developers can shrink or reduce source image NET Image SDK supported image file formats, including & profession imaging controls, PDF document, image
best pdf compressor online; pdf change font size in textbox
C# Image: Zoom Image and Document Page in C#.NET Web Viewer
jpeg), gif, bmp (bitmap), tiff / multi-page tiff, PDF, etc. the web viewer will instantly enlarge or reduce the source file until the file size fits the
adjust size of pdf file; best way to compress pdf files
4.2 Identication
Theresultsabovesuggestthatthereisareduced-formrelationshipbetweenHIVandexports.This
leavesopenthequestionofwhethertherelationshiphereiscausal. Thatis,ifthereisanincreasein
exports,willthatleadtoanincreaseinHIV?Thisissueis,ofcourse,distinctfromthequestionof
whatmechanismdrivesthisrelationship,whichisdiscussedinthenextsection.
Arstissuewithidenticationisthepossibilityofreversecausality:higherHIVcausing
higherexports. Thisseemsveryunlikely.First,inthetimeframeweareconsidering{variationsat
theyearlylevel{itishardtoseehowmorenewinfections(whichwouldbeasymptomatic)could
aecttheeconomy.Second,eventotheextentthatHIVmightaectexportsinthelongrun(for
example,bydebilitatingtheworkforce)therelationshipwouldlikelybeintheotherdirection:more
HIVwouldmeanlowerexports,nothigher.Forbothofthesereasonsreversecausalityshouldnotbe
amajorconcern.
Asimilartiming-relatedlogicappliestotheoriesthattheresultsaredrivenbya
relationshipbetweensomeothercauseofdeathandexports.Forexample,ifroadaccidentsarean
importantsourceofmortality,theymightgoupduringperiodswithmoreexports,whenthereis
moretrucking.However,thiscannot drivetheresults,becausewhatweareestimatingisthe
relationshipbetweennewHIVinfectionsthisyearandexports,notHIVdeathsandexports.
AdditionalmortalityfromothercausesinanygivenyearwillbeattributedtoHIVinfectionsan
averageof10yearsinthepast,not toHIVinfectionsinthecurrentyear.Ifanything,thistypeof
storyislikelytodampentheresults.
Asecondissueliesinomittedvariablebias.Whenconsideringthispossibility,itis
importanttonoterstthattheregressionsTable2includecountryandyearxedeects.The
inclusionofthesecontrolsmeansthatanyconfoundingfactorwillhavetoalsovarywithinacountry
overtime. Wecantakethisastepfurtherbyestimatingtheregressionsusingthesexedeectsplus
country-specictrends.Insomesenseusingsuchextensivecontrolswithalimitedsamplesizemay
beaconcern,andmakeitsomewhathardertointerpretmagnitudesandunderstandwherethe
identicationiscomingfrom.However,totheextentthattheresultsholdqualitativelywiththese
trendsincluded,thisexcludesalargesetofpossibleconfounds.Forexample,onepossible
confoundingfactorischangesingovernment: ifanewgovernmentcomesinthatcaresalotabout
theeconomybutnotabouthealthwecouldseeanincreaseinexportsbutadecreaseinHIV
prevention.However,thesepolicychangesareunlikelytohappenquickly,andincluding
13
How to C#: Special Effects
filter will be applied to the image to reduce the noise. LinearStretch. Level the pixel between the black point and white point. Magnify. Double the image size.
pdf compressor; reader pdf reduce file size
VB.NET Image: Compress & Decompress Document Image; RasterEdge .
reduce Word document size according to specific requirements in VB.NET; Scanned PDF encoding and decoding: compress a large size PDF document file for easier
pdf text box font size; can pdf files be compressed
country-specictrendsshouldcapturethem.
Columns1and2ofTable5reporttherelationshipbetweenexportsandthetwomeasures
ofHIVincidence,controllingforcountry-specictrends. TheseregressionexcludeGDPcontrols{
thedataissimplynotsucienttoestimatealloftheseparameterstogether.Moreover,forthemost
part,reportedGDPmovessmoothlyupwardovertimeinthesecountriesandisthereforelargely
capturedbytheadditionalcontrols. Thecoecientsremainpositiveandstatisticallysignicantfor
bothHIVmeasures.Thislimitsthesetofomittedvariableswhichcouldbedrivingtheresult.
Asasecondargumentagainstomittedvariablesdrivingtheresults,Itakeadvantageofan
instrumentalvariablesstrategy,instrumentingforexportswithworldcommodityprices.The
countriesusedhereexportprimarilycommodities,sotheirexportsareheavilyin uencedby
commodityprices,buttheyaresucientlysmallproducersthatthepriceislargelydeterminedby
eventsoutsideoftheirborders.Theexclusionrestriction{thatworldcommoditypricesdonotdrive
HIVinfectionsforreasonsotherthanthroughtheireectonexports{thereforeseemsreasonable.
Toperformtheinstrumentalvariableanalysis,Igenerateapricemeasureforeachcountrythatisa
weightedaverageoftheworldpricesforeachexportmajor,withtheweightsdeterminedbythe
shareofthetotalexportsaccountedforbyeachcommodity.Historicalcommoditypricesaredrawn
fromtheIMF’sInternationalFinancialStatisticsdatabase(year1980-2001). Ifocusonlyonmajor
exports,sinceidentifyingcommoditypricesforallexportsisnotpossible. Toadjustforthefactthat
thesemajorexportsmakeupalargershareoftotalexportvalueinsomecountriesthanothers,these
regressionsareweightedatthecountry-levelbytheshareofexportsaccountedforbythecommodity
price.
Columns3and4ofTable5showIVregressionswhereexportsareinstrumentedwith
commodityprices(theinstrumentsarecommoditypricesinthecurrentyear,previousyearandyear
beforethat,tocapturelags). ForbothmeasuresofHIVthecoecientonexportsispositiveand
statisticallysignicant,andsimilarinmagnitudetowhatweseeintheOLSestimates. Itshouldbe
noted,however,thattheF-statisticontheinstrumentsintherststageisrelativelysmallinboth
regressions,sotheseresultsshouldbetakenwithcaution.
Finally,wecanexplorethecausalityissuesbroadlybylookingatwhetherfuture exports
drivecurrentHIVrate.Iftrue,thiswouldsuggesteithersomeomittedvariablein uenceor,in
general,callintoquestionthespecicationanddata. Toexplorethis,IregresseachmeasureofHIV
oncurrentexportsandfutureexports,withleadsofdierentlengths(2-10years).
6
Figures2aand
6
Ibeginwithtwoyears,ratherthanoneyear,sincenoiseinthedataissuchthatinfectionsfromnextyearmightwell
14
View Images & Documents in Web Image Viewer | Online Tutorials
page document or image file, like Word, PDF or TIFF to help developers to decrease and reduce current zooming Reset the Size of Currently Viewed File via btnFit
best pdf compression tool; 300 dpi pdf file size
VB.NET Image: How to Process & Edit Image Using VB.NET Image
Compact rich image editing functions into several small-size libraries that are VB.NET programmers the API to scale source image file (reduce or enlarge image
change page size pdf acrobat; pdf file compression
2bgraphcoecientsonpresentandfutureexportsfromtheseregressions.ForbothHIVmeasures
thecoecientoncurrentexportsisconsistentlypositiveandsignicantinallspecications.Wenote
thatthecoecientvariessome,particularlyinthecaseoftheUNAIDSdata,simplybecausethe
samplebecomesmorerestricted(i.e.toearlyyears)whenweincludeverylongforwardleads. In
Figure2a,focusingontheUNAIDSdata,thecoecientonfutureexportsissmallandinsignicant
atallleads.InFigure2b,focusingonthedeathdata,thisislargelytrueaswell,althoughitlooks
likethecoecientonexportstwoyearsoutissignicant,albeitsmaller.Thismaybeduetoissues
ofspill-overin\allocation"ofHIVinfectionsacrossyears,orduetotheuseofthesmoothing
parameters,whicheectivelyforcesomerelationshipinincidenceratesacrossyears. Overall,
however,theseguresarestronglysupportiveoftheresults. Exportsthisyearmatter;exportsinthe
futuredonot.
7
AlloftheseanalysesprovidesupportfortheclaimthattherelationshipbetweenHIVand
exportsiscausal,atleastinthesensethatifexportsincrease,weexpecttoseeHIVincrease.The
nextsubsectiondiscussestherobustnessofthisresulttotheassumptionswhichgenerateincidence;I
thenmovetodiscusswhatmechanismsmightdrivethisrelationship.
4.3 RobustnesstoAssumptionsGeneratingHIVData
OneimportantconcernwiththeregressionsaboveisthattherelationshipbetweenHIVandexports
isbeingdrivenbytheassumptionsthatgenerateincidence,ratherthanbytheunderlying
relationship. ForboththeUNAIDSdataand,especially,themortalitydata,assumptionsgointo
convertingeithertheprevalencedata(forUNAIDS)ordataondeaths(forthemortality-based
estimates)intoincidencerates.InthissectionIbrie yconsiderhowtherelationshipchangesifIvary
twoofthecentralassumptions:thetimepathtodeathfrominfectionandtheepidemicstartdate.
Toaddresstheissueoftimetodeath,IrstgenerateboththeUNAIDSincidenceandthe
mortality-basedincidenceusingtwoalterativepathsoftimetodeath.Oneofthealternativepaths
featuresafastertimetodeath: thisfastertimepathassumesthateveryonehasthetimetodeath
proleoftheoldestagegroup. Theotheralternativepathmodelstimetodeathas atterbutwith
similarspeed:8%ofpeopledieeveryyearbetween5and14yearsafterinfection,ratherthanthe
morepeakedshapeobservedinFigure1(bothalternativepathsareillustratedinAppendixFigure
1).
beattributedtothisyear,makingthisalesspowerfultestatveryshortleads.
7
Thecoecients here comefrommultiple regressions (onefor eachlead); the results lookverysimilar ifwerun a
singleregressionwithleadsofdierentlengthsincluded.
15
VB.NET Image: How to Zoom Web Images in Visual Basic .NET Imaging
and also some file types like PDF and multi out" functionality allows VB developers to easily reduce the size of web image or document file being displayed
best online pdf compressor; pdf file size limit
C# Word: How to Compress, Decompress Word in C#.NET Projects
Efficiently reduce Microsoft Office Word document size using C# code; to compress the Word document file to a & profession imaging controls, PDF document, image
change font size in pdf text box; adjust size of pdf
Inadditiontothesetwoalternatives,itmaybeimportanttoconsiderhowtheavailability
oftreatmentcouldchangetheseresults. Asnotedinthedatasection,treatmentlevelsbefore2002
wereextremelylow{UNAIDSestimatesonly0.2%ofinfectedindividualsinSub-SaharanAfrica
werereceivingtreatmentby2002. Thissuggeststhat,forthemortality-baseddata,whichisallfrom
thepre-2002period,thischangeisnotimportant.TheUNAIDSdata,however,goesthrough2007,
atwhichpoint10%ofHIV-positiveindividualswerereceivingtreatment(UNAIDS,2008). To
explorethis,asathirdalternativeIassumethestandardtimetodeathpriorto2002andthen
assumenodeathsafterthat. Thisisobviouslyanextremechange,butitshouldgiveaboundonhow
thischangecouldaecttheresults.
TherstthreerowsofTable6showthecoecientsonexportsintheprimaryspecication
(correspondingtocolumns2and5ofTable2,withcountryandyearxedeectsandacontrolfor
GDP).Thedierentassumptionsontimetodeathmakealmostnodierencefortheresultswiththe
UNAIDSdata. Thisisre ectiveofthefactthat,forthemostpart,incidenceinyeartinthisdatais
thedierencebetweenprevalenceinyeartandyeart 1.Themagnitudeoftheadjustmentsfor
deathratesaresmallsimplybecauseinevenlateintheepidemicpeoplewhoaredyingwereinfected
earlyintheepidemic,andearlyinfectionratesarelower. Forthemortality-baseddatatheresults
diermoreinmagnitude,whichre ectsthefactthatthistimetodeathisamuchmoreimportant
inputtothecalculationwhenwerelyondeathratesasastartingpoint. However,evenwiththe
dierentpaths,theresultsarepositiveandsignicant(infact,theseresultsarelargerinmagnitude
thantheprimaryresults).
Ialsoconsiderhowtheresultschangeifwechangetheassumed\startdate"ofthe
epidemicbyfouryearsineitherdirection(eitherassumetheepidemicstarts4yearsearlieror4years
later).TheseestimatesareshowninthethirdandfourthrowsofTable6.Again,thesechanges
makelittledierencetotheresults,foreithertheUNAIDSdataorthemortality-baseddata.
Overall,theresultsinTable6indicatethat,whiletheexact magnitudeoftherelationship
betweenHIVandexportscanbesomewhataectedbytheassumptionsaboutgeneratingincidence,
thegeneralconclusionsofthissectionarerobusttochangesintheimportantelementsofthedata
generation.
8
8
Itis worthnotingthatthereareanumber ofotherassumptions that gointogeneratingthemortality-baseddata,
whichdonotaecttheUNAIDSdata. Forexample,duetonoiseinthedeathdatawedosomesmoothingoveryears,
requiringthechoiceof aparticularsmoothingparameter. . Oster r (2008)discusses in moredetailtherobustness of the
estimatestotheseassumptions.
16
5 Mechanism: : TruckingandPeople-Movement
Asdiscussed,thereareanumberofmechanismsbywhichexportscouldaectHIVinfections.One
channelisthroughtheincreasedmovementofpeople. Moreexportsmeanbothmoreproduction,
whichmayusemigrantorothertemporarylabor,andmoretruckingisnecessarytomovethings
fromtheplaceofproductiontocitiesorports.Asecondchannelisthroughincome;ifmoreexports
meansmoreincome,andsexisanormalgood,thiscouldalsodrivetheeect.InthissectionI
providesomepreliminaryevidencesuggestingthattruckingandtransitmayplayaroleinmediating
thisrelationship.However,fullyseparatingoutwhatmechanismsareimportant,ormakingastrong
causalargumentthattruckingormovementofpeopleisakeyelementdrivingtherelationship,is
beyondthescopeoftheanalysisinthispaper.
TruckersandSex;TrucksandExports Therstargumenthereissimplytonotethat
truckershavemoresexthanindividualsinthegeneralpopulation,andmoreexportsmeansmore
trucking. Therelationshipbetweentruckersandsexiswellknowninexistingworkand,infact,
existingevidencepointstoanumberofwaysinwhichtruckingandHIVarelinked.First,truck
driversandothermigrants(i.e. thosewhospendtimelivingortravelingawayfromhome)tendto
havemoresexualpartnersthantheaverageinthepopulation(Lurieetal,2003a;Breweretal,1998;
BrockerhoandBiddlecom,1999;Anaretal,1997;Anar,1993;Orubuloye,Caldwelland
Caldwell,1993). Second,thesexualpartnershipsthesepeoplehaveawayfromhometendtobe
higherriskthanthosetheyhaveathome,largelybecausetheirpartnersaremorelikelytobe
infected:forexample,theyaremorelikelytobebargirlsorcommercialsexworkers(Orubuloyeet
al,1993).Finally,thepartners(forthemostpart,wives)ofthosewhotravelmaybemorelikelyto
haveadditionalsexualpartnerswhiletheirspousesareaway(Lurieetal,2003b). Thelevelofrisky
behaviorissignicant;inasurveyinNigeria,Orubuloyeetal(1993)ndthatthetruckdriversin
thesamplehaveanaverageof12non-maritalpartnersayear.
Thisexistingevidencesuggestsalinkbetweentruckingandsex. PanelAofTable7
demonstratesthesecondstep:moreexportsmeanmoretrucking. Thisanalysisusesdatafromthe
WorldDevelopmentIndicatorsandregressestheweightofgoodstruckedwithinayearonexports
thatyear,withcountryandyearxedeectsandacontrolforGDP.Column1includesonlythe
countersinoursampleforwhichthisdataisavailable;thisisunfortunatelyaverysmallset(just24
country-years). However,thecoecientispositiveandstronglysignicant. Column2includesall
lowandlower-middleincomecountrieswithavailabledataand,inthislargersample,alsoshowsa
17
signicantpositiverelationship.
Takentogether,thesetwofactssuggestarolefortruckinginmediatingthisrelationship.If
havingmoreexportsmeanstherearemoretrucks,andtruckershavemoresexwhentheyareaway
fromhomethennot,thesetogetherimplythatanincreaseinexportsshouldincreaseriskysex.Itis,
ofcourse,importanttonotethatthisdoesnotprovetheconnection,andtherearecertainlymodels
ofbehaviorwhichwouldnotbeconsistentwiththis.Forexample,ifthemarginal truckdriverdoes
nothavealotofriskysex,thentheincreasedtruckingwouldnotobviouslyincreaserisk.
VariationsacrossCountries: RoadsAsperhapsamoredirecttest,wecantakeadvantageof
variationacrosscountriesinoursampleinroadavailability. Totheextentthatanimportant
mechanismhereistruckingormovementofpeople,wewouldexpecttheeectofexportstobe
strongerinareaswhereincreasesineconomicactivitywillhavealargereectonthesevariables.
Thisislikelytobemoretrueincountrieswithmoreroads. Putdierently,ifweseeanequally
strongrelationshipbetweenexportsandHIVincountrieswithnoroads,orverylimitedroads,this
wouldsuggestthattruckingisnot animportantmechanisms. . Conversely,ifweseeastronger
relationshipincountrieswithextensiveroadnetworks,thiscouldpointtowardtheimportanceof
transitandmovementinmediatingtheexport-HIVrelationship.
DataonroadscomefromtheWorldDevelopmentIndicators.Iusetwomeasures:total
kilometersofroadsdividedbylandareaandtotalkilometersofpavedroads,dividedbylandarea.
Thesemeasuresdonotvarymuchovertime,soIuseaconstantvaluepercountry,theaverageof
thesevariablesovertheentireperiod.Iestimatetheprimaryspecications(countryandyearxed
eects,GDPcontrols),butincludeaninteractionbetweentheroadmeasuresandexports,aswellas
thelevelofexports(thecontrolforroadleveliscapturedinthecountryxedeect).
TheseregressionsareshowninPanelBofTable7. FocusingrstontheUNAIDSdata,in
Columns1and2,wendthatforroadsoverall,andforpavedroads,theleveleectofexportsin
theseregressionsissmallandnotsignicant,buttheinteractionislarge,positiveandsignicant. In
otherwords,therelationshipismuchlargerinareaswheretherearemoreroads,pointingtothe
importanceoftransitinthisrelationship. Theseresultsaresimilarifwesimplydividethesamplein
halfbasedonroaddensityandestimatetherelationshipforthetwohalvesofthesampleseparately
{thereisapositiveandsignicantrelationshipforcountriesinthetophalf,andnorelationshipfor
thoseinthebottom(resultsavailablefromtheauthor).Columns3and4showthisrelationshipin
themortality-baseddata;weseethesamepatterns,althoughtheresultsareonlysignicantfor
pavedroads.Thisislikelyduetothemorelimitedsetofcountriesinthesedataand,hence,the
18
morelimitedvariationinroaddensity.
VariationsacrossCountries: PortsFinally,inatestrelatedtotheabove,Icanexplore
whethertherelationshipbetweenexportsandHIVisstrongerincountriesthataremorerelianton
roadtransportation.Inparticular,weseparatethecountriesinthesampleintotwogroups:those
withports,whichwepositarelessreliantonroadsforgoodstransport,andthosewithout.Since
mostofthecommoditiesproducedbythesecountrieswilleventuallybetransportedbyshipoutof
Africa,forcountrieswithoutaportthereissignicantadditionaltruckingtimenecessarytoexport
goods. Totheextentthattruckingisamediatingfactorhere,wewouldexpecttheexport-HIV
relationshiptobelargerinareaswithnoport.
PanelCofTable7showstheseregressions,dividingthesamplesintocountrieswithaport
andthosewithout.Forbothdatasourcestherelationshipismuchlarger,andmoresignicant,for
countrieswithout aport,consistentwiththetheoryabove. . Ofcourse,therecouldbeotherfactors
whichvaryacrosscoastalandnon-coastalcountries,andplayintotheexport-HIVrelationship.This
evidenceissupportiveofaroleforatransitmechanismbut,again,notconclusiveonitsown.
Theevidenceinthissubsectioncertainlydoesnotprovethattruckingisanimportant
mechanismbehindtheHIV-exportrelationship. Alloftheevidence,however,isatleastsupportive
ofthiscase.Inadditiontosheddinglightonthismechanism,thisevidencefurthersupportsthe
claimthattherelationshipbetweenexportsandHIViscausal.Ifanomittedvariablelike
governmentpolicywasdrivingthisresultitseemsunlikelywewouldsee,forexample,adierence
betweencitieswithportsandthosewithout.
6 Application: : PrevalenceDeclinesinUganda
Theprevioussectionsprovidesevidencethat,atthemacrolevel,newHIVinfectionsandeconomic
activityarelinked. Thissectionconsiderswhetherthatlinkmaybeatleastpartiallyresponsiblefor
oneofthemorestrikingpatternsinHIVprevalenceinAfricaoverthe1990s.
HIVprevalencedeclinedsharplyinUgandaintheearly1990s,adeclineuniqueamong
Sub-SaharanAfricancountries(thisisapparentinboththeUNAIDSdataandinthe
mortality-basedestimates).Thisdeclineisgenerallyattributedtoanti-HIVeducationaleorts,
includingthosebytheUgandaNationalAIDSProgram(NACP)andtheABCcampaign,bothof
whichencouragedchangesinsexualbehavior:partnerreduction,\zero-grazing",abstinenceand
19
condomuse(Green,2004;Greenetal,2006;Slutkinetal,2006).
9
Amongthemoreimportant
componentsoftheeducationaleortswerein-schooleducation,communicationbyfaith-based
organizations,billboardsandotherformsofpublicadvertising(foragoodsummaryofthespecic
aspectsofABC,seeGreen,2004;Slutkinetal,2006provideasummaryofnon-ABCpartner
reductioneortslikeNACP).Recentworkhasarguedthatchangesinconcurrencyasaresultofthis
programwereimportantinthedeclineinHIV(Epstein,2007).
Ugandawasamongtherstcountriestoaddresstheepidemicheadonandthesubsequent
declineinprevalencewasseenasasignalofthesuccessofthebehavioraleducationapproach.
Subsequently,anumberofothercountriesadoptedthistypeofpreventionstrategy(Kenya,
Tanzania,Rwanda,Malawiandothers).Outsideaidgroupshavepursuedsimilarinterventions.
PEPFAR,inparticular,hasadoptedthiscampaignasamajorpartoftheiranti-HIVstrategy:
roughlyonehundredmilliondollarsoftheirfundingwasspentonABCprogramsin2004.
However,intheearly1990sUgandaalsosawamajordeclineinexports.Thiswaslikely
dueinlargeparttoadeclineincoeepricesandproduction.
10
Figure3demonstratesthatthe
declineinexportvaluelinesupcloselywiththechangesinincidenceinUganda;thisistrueinboth
theUNAIDSdataandthemortality-basedestimates.
11
GiventherelationshipestimatedinSection4,thedeclineinexportvolumeinUganda
duringthisperiodsuggeststhatHIVincidencewouldhavedeclinedsomewhatevenwithoutany
educationcampaign.UsingthecoecientestimatesoftherelationshipbetweenexportsandHIV
incidence,alongwithdataonthemagnitudeofthedeclineinexports,wecancalculatehowlargewe
wouldexpectthedeclineinincidencetobebasedontheexportdeclinealone.Wecancomparethat
withtheactualmagnitudeofthedecline,andestimatewhatshareoftheUgandan\miracle"appears
tobeexplainedbyexports.
TheresultsofthiscalculationareshowninTable8.NotethatthecoecientsIuseinthis
calculationarebasedonregressionestimateswithout Uganda,sothiscanbeseenasanoutof
sampleprediction. Theshareofthedecline(frompeaktotrough)whichisexplainedbyexports
9
Although thereisafair amountofagreementonthesuccessofeducationingeneralinUganda(althoughseeGray
et al, 2006),thereis lessagreement ontherelativesuccessofthepartner reductionelements,versus ABCspecically,
andwithintheABConthevariousmeritsofAversusBversusC;forasummaryandcitationsonthiscontroversy,see
Greenetal,2006.
10
Thischangeinexportsdidnotre ectacollapseoftheUgandaneconomyingeneral.Itwasdrivenbyaworldcoee
priceshock,whichwascausedbyfactorsoutsideofUganda,anddecreasedexports.
11
TheUNAIDS datais somewhat noisy. . This s is driven bythefact that in theantenatalclinicdatafor the 1980s,
UNAIDS reported a13% HIV rate e in 1986, but 24%in 1987and d then n 25% % in the next year. . The e increase during
this period was surely much smoother r than n this, , and d if properly represented would d likely look k much more like the
mortality-baseddata.
20
Documents you may be interested
Documents you may be interested