Topic/Lecture Paper Keywords Main Results Methodology
Lecture 2: E. Duflo (2000) Indonesia: Enrollment rates increased Date of birth and region of
Education “Schooling and constructing Per district: number of schools birth jointly determine
Labor Market schools to be constructed was exposure to the program
Consequences of (1973-74; proportional to the number of Identification assumption:
School 1978-79) primary school age children not there is no time-varying and
Construction in enrolled in school region specific effect that
Indonesia: Increase in years of education correlate with the program
Evidence from an Increase in wage rates (could be violated if changes
Unusual Policy → higher economic returns to in education returns correlate
Experiment”, education with initial level, or if other
American programs were rolled out at
Economic the same time
Review, 91(4), Remedy: control for
795-813 enrollment in 1971, control
for access to water and
sanitation programs
Cost- benefit analysis:
compare building and
maintenance cost of school
with additional wealth
generated
Lecture 3: R. Chetty, A. Salience of Manipulate price tags at a major Triple difference approach:
Taxation 1 Looney and K. taxes; US US grocery chain store to make products (cosmetics vs other
Kroft (2009) Alcohol and sales tax more salient (tags with products in the same aisle),
”Salience and two types of tax-inclusive price), then track store (one large in North
Taxation: Theory taxation the sales of products with California vs two others with
and Evidence“, scanner data similar demographics), time
American It is about salience (and not that period (3 weeks in
Economic people do not know the tax rate) Feb/March 2006 vs calendar
Review, 99(4): because: they ask shoppers year 2005 and first 6 weeks
1145-1177 about applicable tax rate and of 2006)
most get it right Develop a model to estimate
Alcohol: excise tax (included in welfare to include incorporate
price) vs sales tax (added at the salience
register)
Findings contradict the
assumption of the neoclassical
model used to guide tax policy
→ it is less likely to lower
alcohol consumption with a
sales tax compared to the
change in consumption caused
by an excise tax
Consumers underreact to taxes
that are not salient
, Lecture 4: N. Eissa (1995), Earners’ After the TRA1986 instead of 14 DiD: Compare very
Taxation 2 ”Taxation and response to brackets the personal income high-earning women with
labor supply of income tax tax had only 2 moderately earning women;
married women: changes Very top of income distribution: before and after TRA1986
the Tax Reform Tax Reform from 50% to 28% → huge Justification: women with
Act of 1986 as a Act of 1986 change moderately high earnings
natural US Changes in deductions and saw much lower reductions
experiment”, capital gains tax reduced in their marginal tax rate
NBER Working income effects Control group necessary,
Paper 5023 Focus in married women: because labor supply of
second earners usually respond women changes over time
stronger to changes in earnings The Current Population
opportunities Survey
Results: Women with very high
earnings (relatively) increase
their labor supply; The rise in
after-tax wages yields greater
labor supply; Half of effect
comes from the extensive
margin; Little effect on men’s
wages
Lecture 4: Chetty et al. US Idea: provide tax refunds to low Control group: people without
Taxation 2 (2013) Anti poverty income households with children (not eligible)
program - children Key assumption needed:
Earned Share of self-reported income at Individuals in low-bunching
Income Tax EITC - maximizing kink: proxy neighborhoods believe that
Credit (EITC) for knowledge about the the EITC has no impact in
marginal rate structure of EITC their marginal tax rates:
Estimate the effect of taxes on Spatial heterogeneity in
labor supply bunching correlated with
Variation comes from diffusion knowledge about the first
of knowledge across time kink; having children
changes EITC eligibility, but
does not change labor supply
(in low-bunching areas)
Lecture 4: J. Gelbach US; Uses quarter of birth (QOB) to Instrumental variables
Taxation 2 (2002), “Public kindergarten identify kindergarten attendance approach
schooling for (kindergarten eligibility in the First stage: regress
young children US works through birthday kindergarten enrollment on
and maternal cutoffs) QOB
labor supply”, Second stage: regress labor
American supply on predicted
Economic enrollment
Review, 92(1):
307-322
Lecture 5: Burman & US; capital Results: virtually all of the Use panel data to estimate
Lecture 2: E. Duflo (2000) Indonesia: Enrollment rates increased Date of birth and region of
Education “Schooling and constructing Per district: number of schools birth jointly determine
Labor Market schools to be constructed was exposure to the program
Consequences of (1973-74; proportional to the number of Identification assumption:
School 1978-79) primary school age children not there is no time-varying and
Construction in enrolled in school region specific effect that
Indonesia: Increase in years of education correlate with the program
Evidence from an Increase in wage rates (could be violated if changes
Unusual Policy → higher economic returns to in education returns correlate
Experiment”, education with initial level, or if other
American programs were rolled out at
Economic the same time
Review, 91(4), Remedy: control for
795-813 enrollment in 1971, control
for access to water and
sanitation programs
Cost- benefit analysis:
compare building and
maintenance cost of school
with additional wealth
generated
Lecture 3: R. Chetty, A. Salience of Manipulate price tags at a major Triple difference approach:
Taxation 1 Looney and K. taxes; US US grocery chain store to make products (cosmetics vs other
Kroft (2009) Alcohol and sales tax more salient (tags with products in the same aisle),
”Salience and two types of tax-inclusive price), then track store (one large in North
Taxation: Theory taxation the sales of products with California vs two others with
and Evidence“, scanner data similar demographics), time
American It is about salience (and not that period (3 weeks in
Economic people do not know the tax rate) Feb/March 2006 vs calendar
Review, 99(4): because: they ask shoppers year 2005 and first 6 weeks
1145-1177 about applicable tax rate and of 2006)
most get it right Develop a model to estimate
Alcohol: excise tax (included in welfare to include incorporate
price) vs sales tax (added at the salience
register)
Findings contradict the
assumption of the neoclassical
model used to guide tax policy
→ it is less likely to lower
alcohol consumption with a
sales tax compared to the
change in consumption caused
by an excise tax
Consumers underreact to taxes
that are not salient
, Lecture 4: N. Eissa (1995), Earners’ After the TRA1986 instead of 14 DiD: Compare very
Taxation 2 ”Taxation and response to brackets the personal income high-earning women with
labor supply of income tax tax had only 2 moderately earning women;
married women: changes Very top of income distribution: before and after TRA1986
the Tax Reform Tax Reform from 50% to 28% → huge Justification: women with
Act of 1986 as a Act of 1986 change moderately high earnings
natural US Changes in deductions and saw much lower reductions
experiment”, capital gains tax reduced in their marginal tax rate
NBER Working income effects Control group necessary,
Paper 5023 Focus in married women: because labor supply of
second earners usually respond women changes over time
stronger to changes in earnings The Current Population
opportunities Survey
Results: Women with very high
earnings (relatively) increase
their labor supply; The rise in
after-tax wages yields greater
labor supply; Half of effect
comes from the extensive
margin; Little effect on men’s
wages
Lecture 4: Chetty et al. US Idea: provide tax refunds to low Control group: people without
Taxation 2 (2013) Anti poverty income households with children (not eligible)
program - children Key assumption needed:
Earned Share of self-reported income at Individuals in low-bunching
Income Tax EITC - maximizing kink: proxy neighborhoods believe that
Credit (EITC) for knowledge about the the EITC has no impact in
marginal rate structure of EITC their marginal tax rates:
Estimate the effect of taxes on Spatial heterogeneity in
labor supply bunching correlated with
Variation comes from diffusion knowledge about the first
of knowledge across time kink; having children
changes EITC eligibility, but
does not change labor supply
(in low-bunching areas)
Lecture 4: J. Gelbach US; Uses quarter of birth (QOB) to Instrumental variables
Taxation 2 (2002), “Public kindergarten identify kindergarten attendance approach
schooling for (kindergarten eligibility in the First stage: regress
young children US works through birthday kindergarten enrollment on
and maternal cutoffs) QOB
labor supply”, Second stage: regress labor
American supply on predicted
Economic enrollment
Review, 92(1):
307-322
Lecture 5: Burman & US; capital Results: virtually all of the Use panel data to estimate
