The next two terms allow earnings to vary within and following
schooling. First, eit is an indicator for whether the individual
is enrolled in college during period t. Finally, f(·) is
a function of credits (sit) and credentials (cit) received following
college enrollment. Under the identifying assumption
that – conditional on the observable characteristics in
xit and unobservable individual-specific effects, time-trends,
and quarter-specific effects – the unobserved error term εit
is uncorrelated with schooling choices and the vector of parameters,
τ, contains estimates of the causal effect of educational
attainment on labor market outcomes. My main
parametrization of f(sit, cit, τ) includes indicators for credential
receipt and a linear term in credits received interacted
with an indicator for not having earned a credential:
f(·) = τ1sit1[c ∈ {Dc} = 0] +
c
τ cDc, where Dc indicates receipt
of specific credentials.18 With this specification, I can
test whether college drop-outs benefit from earning additional
credits and whether earnings and employment gains
are larger for particular credentials. I also explore whether
my findings are robust to different assumptions over the
function form of f(sit, cit, τ), which are described in the following
section.
The next two terms allow earnings to vary within and followingschooling. First, eit is an indicator for whether the individualis enrolled in college during period t. Finally, f(·) isa function of credits (sit) and credentials (cit) received followingcollege enrollment. Under the identifying assumptionthat – conditional on the observable characteristics inxit and unobservable individual-specific effects, time-trends,and quarter-specific effects – the unobserved error term εitis uncorrelated with schooling choices and the vector of parameters,τ, contains estimates of the causal effect of educationalattainment on labor market outcomes. My mainparametrization of f(sit, cit, τ) includes indicators for credentialreceipt and a linear term in credits received interactedwith an indicator for not having earned a credential:f(·) = τ1sit1[c ∈ {Dc} = 0] +cτ cDc, where Dc indicates receiptof specific credentials.18 With this specification, I cantest whether college drop-outs benefit from earning additionalcredits and whether earnings and employment gainsare larger for particular credentials. I also explore whethermy findings are robust to different assumptions over thefunction form of f(sit, cit, τ), which are described in the followingsection.
การแปล กรุณารอสักครู่..
