See my website for details, including a correction procedure to account for selection into college transcript data

My research portfolio includes several working papers spanning K-12, higher ed, poverty alleviation & family economics

Hire me!

hema-shah-econ.com/research

(10/10)

My work bridges literature on teacher value added & predictors of college enrollment/persistence

2 main contributions:

1️⃣ First link between K-12 teacher value added & college performance
2️⃣ First estimates of teacher value added on college admissions test scores

(9/10)

I show that teachers high value added teachers boost college performance

⬆️ college GPA
⬇️ remedial course-taking
⬇️ college dropout
⬆️ college completion

Suggesting that high value added teachers help students develop college-relevant skills

(8/10)

Enrollment impacts are too large to be explained by the signal value of a higher ACT score

Suggesting that increases in enrollment are partly driven by increases in skills ...

Question 3️⃣: Are high value added teachers developing skills relevant to college performance?

(7/10)

I show that impacts on selective college enrollment are largest among high achievers

A 1σ increase in teacher value added increases flagship university enrollment by over 3 percentage points among students in the top ventile of lagged achievement!

(6/10)

I model college enrollment using a nested logit model, allowing for the effects of value added to vary across students with different lagged achievement levels

(5/10)

2️⃣: When high school teachers boost students’ ACT scores, what happens to college enrollment?

I show that teachers who boost ACT scores ("high value added teachers") lead to

⬆️ 4-year college enrollment
(& selective college enrollment)
⬇️ 2-year college enrollment

(4/10)

1️⃣: How do high school teachers impact students’ ACT scores? ("ACT Score Value Added")

I leverage universal ACT testing in NC & show that teachers matter for ACT scores

Assignment to a teacher with value added 1σ above the mean increases ACT scores by ~0.5 points/36

(3/10)

I use novel admin data from North Carolina linking K-12 public school records with enrollment & transcript records from public 4-year universities

to answer 3 main questions

(2/10)

w/ @lisagennetian.bsky.social, Katherine Magnuson, Hirokazu Yoshikawa, Laura Stilwell, Kim Noble, and Greg Duncan

Feedback welcome!

What does this mean?
🔸Latino families’ responses suggest signals regarding intended use of income influence spending decisions
🔸Differential impacts on monetary & time investments suggest unconditional cash can enable families to respond to diverse circumstances & constraints

Effects on child development are so far inconclusive, with 1 exception. An objective assessment of child executive functioning at age 4 shows:

🔸⬆️among Latino children (who experienced large infusions of monetary investments)
🔸null effect among Black children

We compliment treatment effect estimates with MPC estimates, using the control group. We find:
🔸Among Latino families: higher marginal propensity to consume child-specific goods from government income than from maternal income
🔸Among Black families: the opposite!

We find:
🔸Similar effects on net household income across Latino & Black families
🔸Different effects on monetary & time investments in children:
🔸Among Latino families: ⬆️child-focused expenditures
🔸Among Black families: ⬇️maternal work hours, ⬆️time spent with children

🔸Responses to cash depend on preferences, norms, & structural factors which vary by race/ethnicity
🔸Key challenge: distinguishing between differential take-up & differential responses to cash
🔸That's where the BFY RCT comes in: no differences in take-up by race/ethnicity!

Thank you for sharing!!

I think the fed usually follows federal hiring freezes but not sure if they are required to!

www.jonathandroth.com/assets/files...