Assignment 3: PROGRESA conditional cash transfers

Assignment three: PROGRESA conditional cash transfers and
secondary faculty enrollment
October 31, 2022
The info come from Bobonis and Finan (2009, Evaluate of Economics and Statistics) who regarded
on the impact of PROGRESA conditional cash transfers on faculty enrollment and the position of peer
results. Knowledge was collected in 1997 (baseline) and by 1998 (endline/submit=1) just some villages
(deal with=1) had acquired the transfers. Remedy throughout villages was randomized. Inside every vil lages, solely households with a low sufficient welfare index have been thought-about poor and due to this fact eligible
(eligible=1) to obtain transfers; different households have been deemed ineligible (even within the remedy
villages). The info embrace the next variables for college kids and heads of households (collected
at baseline except in any other case specified):
folio – family ID
numero – particular person ID
yr – yr
village – village ID
sc – faculty enrollment
eligible – PROGRESA eligible
welf are – eligibility welfare index (decrease for the poor)
submit – =1 for 1998, =zero for 1997
deal with – PROGRESA remedy village (=1), management (=zero)
gender – (=1 Male, =zero if Feminine)
indig – (=1 indigenous, =zero in any other case)
hohedu – Head of family education attainment
hohsex – Head of family gender (=1 Male, =zero Feminine)
hohage – Head of family age
f amn – family measurement
distsec – minimal distance to secondary faculty
mindist – minimal distance between the village and huge city heart
meanvillsc – common faculty enrollment of eligible youngsters within the village/yr
1
(a) What can we are saying in regards to the determinants of secondary faculty enrollment primarily based on the
(cross-sectional) relationships within the information (for 1997)? [2pts]
(b) At what stage must you cluster the usual errors in (a)? [1pt]
(c) Which traits are systematically (statistically) totally different between eligible (poor) and
ineligible (“wealthy”) households? [2pts]
(d) Present proof that the across-village randomized remedy was actually randomized [2pts]
(e) Present an estimate of the typical remedy impact (ATE) of PROGRESA transfers (on
faculty enrollment) for the eligible youngsters and the ATE for the ineligible youngsters utilizing the professional gram randomization and interpret the outcomes [2pts]
(f) At what stage ought to we cluster the usual errors in elements (d) and (e)? [1pt]
(g) We are able to additionally estimate the ATE of village-level PROGRESA remedy utilizing a distinction in-difference estimator – do that for the eligible youngsters and focus on how these estimates relate to
your estimates in (e) in addition to what we study general traits in education [2pts]
(h) Suppose we solely had information from remedy villages. Use the eligible and ineligible youngsters
to estimate a mean remedy impact utilizing a difference-in-difference specification. Focus on the
potential bias from this estimate. [2pts]
(i) Suppose we wished to estimate the causal impact of the village-level faculty enrollment fee
of eligible youngsters on faculty enrollment of ineligible youngsters. Present an OLS and a IV (2SLS)
estimate of this relationship [hint: you can use the randomized treatment as the instrument, and
can calculate the IV coefficient in excel using the reduced form and first-stage coefficients], focus on
whether or not the IV circumstances are glad and interpret the outcomes. [2pts]
(j) Suppose we solely had information from remedy villages within the post-period (1998). Noting that
eligibility for this system is expounded to the welfare (poverty) index, present as near a causal
estimate as one can estimate given the info limitations and focus on the instinct for why your
method is superior to easily regressing faculty enrollment on program eligibility and obtainable
management variables [hint: think in terms of the logic of a regression discontinuity design] [4pts]
2
three