Healthcare ADVANCED

Question 1

How can healthcare be modelled (NEOCLASSICAL)?

Answer 1

• Objective: MaxU(X,H)
• Constraint: Budget: PxX + PhH = I
• Substitute B.C. into objective and take derivsatives to find the MRSx,h
• Relate this to the price ratio and solve for H; f(Px, Ph, I)
• Usually; U (X,H) = lnX + βlnH

Question 2

How can you find PED and IED?

Answer 2

• PED: δH/δPh * Ph/H
• IED: δH/δI * I/H

Question 3

What does the Grossman model illustrate (GRAPHICALLY)?

Answer 3

• QUADRANT 1 (TL): Health Vs Healthcare- health production funcion exhibiting DMR
• QUADRANT 2 (TR): Health Vs Other Inputs- P.P.F curve, which meets an I.C at the most opitmal level
• QUADRANT 3 (BL): Other Goods Vs Healthcare- Budget Constraint
• QUADRANT 4 (BR): Other Goods Vs Other Inputs- Household production function- doesn’t exhibit DMR
• δP=> Shift in PPF curve- only one side thus slope changes
• δI=> Shift in PPF curve-slope remains the same

Question 4

What are the objective and constraints of the Grossman Model?

Answer 4

• OBJECTIVE: Max U(O,H)
• SUBJECT TO:
• Law of motion of health stock: HSt = HSt-1 - δt + It
• Health Production Function: It = I(Mt, THt, Et)
• Household Production Function: Ot = O(Xt, TOt, Et)
• Time Constraint: Tt = THt + TOt + TWt
• Budget Constraint: PmM + PxX = TwW -> Written in terms of TVs

Question 5

What does the Grossman Model tell us?

Answer 5

• A rise in depreciation due to ageing population leads to a lower health stock
• A rise in wages lead to higher health stock on an aggregate level- not individual
• A rise in education leads to higher health stock

Question 6

How can an econometrics model be used to examine the healthcare sector?

Answer 6

• Build a regression in reduced form (depedant on left, regressors on right)
• Parkin and Yule (1988) Demand for Dentures:
• Dn = 157505 - 10960Pd + 2141Po - 4432I + 8867A
• Where Pd = Price, Po = Price of Others, I = Income, A = Availability

Question 7

What is difference in differences? What does this rely upon?

Answer 7

• DiD compares changes in outcomes over time between individuals affected by exogenous event and those who aren’t
• B - A: Difference in treatment group
• D - C: Difference in non-treatment group
• TREATMENT EFFECT: [D-C] - [B-A]
• Example of enrollment in schools- England Vs Scotland
• Validity of DiD relies upon ‘parallel trends assumptions’- requiring no time-varying differences
• This can be tested via diagnostic strategies; control group used to work out outcomes (Placebo effect)