Treść książki
Przejdź do opcji czytnikaPrzejdź do nawigacjiPrzejdź do informacjiPrzejdź do stopki
EmploymenteffectsofUSminimumwagepolicy
iscomparingtheestimatesofβ1inregressions
lowerlevelofemploymentfor19-21yearold
inequation(2)withthoseobtainedfromthe
workers.Againtheestimatedminimumwage
expandedmodelinequation(3).
elasticityforthisgroupisstrengthenedconside-
rablyintheregressioncontrollingforEconomic
ln(
Y
ist
)
±
B
0
+
B
1
ln
(
MW
st
)
+
Areaspecifictimeeffects.Incontrastwiththe
+
B
2
ln(
EMP
ist
TOT
)
+
B
3
ln(
POP
ist
TOT
)
+
(3)
resultsforthoseaged14-18and19-21,theresults
incolumn5revealapositivecorrelationbetween
u
i
+
EA
*
t
t
+
8
ist
,
theminimumwageandtheemploymentof
22-24-year-oldworkers.Thisresultisstatistically
significantwithapositiveelasticityof.095in
V.Results
PanelB.Theresultsinthatpanelsuggestthat
theminimumwagehasthebiggesteffectonthe
Ourmainfindingsabouttheeffectsofthe
youngestworkers,slightlysmallerbutstillnegative
minimumwageonthelevelofemployment
effectsonemploymentfor19-21-year-olds,and
andearningsforyouthinthethreeagegroups
apositiverelationshipwithemploymentforthe
arepresentedinTable2.PanelAshowsthe
22-24-year-olds.Theminimumwageispositively
resultsfromthetraditionalpaneldatamodel
relatedtotheaveragemonthlyearningsofworkers
(Equation(2))andpanelBreportsestimatesof
inallthreeagecategories,withstatisticallysigni-
themodelwithEA-specifictimeeffects(Equa-
ficantresultsinPanelB,wherewearecontrolling
tion(3)).Becausestateminimumwagesarethe
forspatialheterogeneity.
sameforallcountieswithinonestate,theerror
Table3reportstheestimatesofminimum
termε
satisfythebasicassumption
it
inequation(2)and(3)oftendoesnot
8
it
Ź
(0,
σ
8
2
)
in
wageeffectsondynamicjobchangesforallyouth
intheGreatLakesregion,usingthepaneldata
thepaneldataset,butinsteadtheidiosyncratic
modelwithEconomicAreaspecifictimeeffects.
errortermsareprobablycorrelatedwithineach
Wefindthatbothaccessionsandseparations
state.Therefore,weuseclusteredstandarderrors
forallyoutharesubstantiallylowerincounties
fortheestimatedcoefficients,whichallowforan
withahigherminimumwagelevel.Thenegative
arbitrarypatternofcorrelationintheerrorterms
effectoftheminimumwageonaccessionsand
acrossdifferentcountieswithinthesamestate.
separationsisalsoseeninlowerturnoverrates
Alltherobuststandarderrorsinbracketsinare
injurisdictionswithhigherminimumwagele-
clusteredatthestatelevelforallregressions.
vels.Thisisconsistentwiththeevidencefound
InColumn1ofTable2,wefindthatthe
byDubeetal.(2011)andPortugalandCardoso
minimumwageisnegativelycorrelatedwith
(2006)andsuggeststhat,byraisingthewage
theteenage(14-18yearsold)employmentlevel.
andmakingjobsmorevaluabletotheirholders,
ByaddingEA-specifictimeeffects,ourmodel
theminimumwagelowersquitratesandspurs
showsabiggernegativeteenageemploymenteffect
greaterworkereffortleadingtofewerdismissals
withtheelasticityof-0.21,comparedwiththat
(ShapiroandStiglitz1984).Withagivenlevel
of-0.10basedonthetraditionalpanelmodel.
ofemploymentandalowerseparationrate,we
Thus,ourestimationoftheteenageemployment
wouldalsoexpectalowerhiringratesincethere
elasticityfallsintotherangebetween-0.1and
wouldbefewervacanciestofill.Anotherpossible
-0.3intheconsensusofnationalCPSstudiesbut
reasonforanegativeturnoverrateelasticityisthat
incontrasttoDubeetal.(2010)andAllegrettoet
ahigherminimumwageshiftstheemployment
al.(2011),whofindthattheemploymentelasticity
distributionawayfromhigh-turnover,low-wage
isindistinguishablefromzero,aftercontrolling
firmstolow-turnover,high-wageones(Dube,
forspatialheterogeneity.
2011).
Column3inTable2presentstheeffectsof
Finally,Column4presentstheestimated
theminimumwageonemploymentfor19-to-
coefficientsfortheminimumwageinregressions
21-year-oldyoungworkers.Basedonthemodel
withthenetjobgrowthrateasthedependent
withEA-specifictimeeffect,wefindthatahigher
variable.Countyjobgrowthratesaggregated
minimumwageisassociatedwithasignificantly
acrossestablishmentsforworkersaged14-18and
11
