Lecture 4 - Is the Influenza Pandemic over? (Almond,2006)

Question 1

What is the main goal of the paper?

Answer 1

To estimate the impact that being exposed to influenza pre-natally compared to post-natally has on education outcomes.

It goes beyond previous literature which often were studies on animals

Question 2

What datset is used?

Answer 2

Micro data from the 1960,70, 80 census is used as it shows birth location and date.

Influenza pandemic broke out in the late 1918s and ended at the start of 1919.

Question 3

What is the fetal origins hypothesis?

Answer 3

Babies born before the pandemic or at least 9 months after the end should be unaffected.

This is because pregnant mothers were largely affected so often the disease would be passed on to the fetus.

Question 4

What was the main regression estimated by Almond?

Answer 4

Yci= B0+B1 I(YOB=1919)+B2(YOB) +B3(YOB)^2+eic

Y- outcome variable for individual I, in cohort c. Can be school attendance, likelihood of graduating , percentage with a disability.
I is a dummy variable for if the year of birth was 1919 (to show if the YOB was the year of the pandemic)

When I=1 B1 tells us how different people born in 1919 are to people in the sample as a whole
We expect it to be negative for education outcomes, positive for disability variable.

Question 5

What were the main results of the paper?

Answer 5

Overtime, the percentage of babies that were disabled fell, however in 1919 there was a clear departure from the trend and these babies had a greater probability of being disabled. After the influenza pandemic (sep 1919) the trend returned to how it was before

Question 6

What is the empirical strategy

Answer 6

Randomized control Trial-possible to measure as people don’t self select to have influenza

In terms of average age, education income, those born in 1919 and the surrounding cohorts were the same when looking at the 1960 1970 and 1980 census. This is why we can assume that the difference in disability and years of schooling is due to the influenza pandemic.

Question 7

What was the alternate regression that was used?

Answer 7

One that looked at geography, place of birth of the baby as a measure impacting influenza and so outcomes

Yi=B0+B1ψs Y0B-1+ γ YOB+λs+ei
ψs Shows the effect that differing influenza infection rates across mothers in state s (and with YOB-1) this is during pregnancy, had on education outcomes (such as school attendance, likelihood of graduating).

We look at YOB-1 because we want to capture the effect of the mum having influenza during the pregnancy
λs Accounts for fixed effects based on which state the baby was born in.