You are the Secretary of Labor in Blandings. The Empress of Blandings is thinking about
introducing a national unemployment insurance program in Blandings.
(a) The Empress is concerned that introducing this benefit may result in moral hazard. One of
her economic advisors, argues that as long as the unemployment insurance benefit is financed
through taxes, there will be no moral hazard problem since individuals know that they must
pay for longer unemployment durations through higher taxes. Do you agree? Explain. (3-4
sentences)
(b) Suppose that all citizens in Blandings are identical. For each individual, the coefficient of
relative risk aversion is 2 and the elasticity of unemployment duration with respect to
benefits is 0.8. Using the social insurance framework developed in class, what will the
consumption drop be between the good and bad states of the world be at the optimal level of
social insurance benefits?
(c) You have figured out the optimal level of benefits in part (b). Now your economists tell you
that the elasticity of unemployment duration is lower than they had previously estimated.
You need to determine the new level of optimal benefits. Will this level of benefits be higher
or lower than before? Explain.
- Moral Hazard in Health Care
Suppose that John’s demand for physician office visits per year is Q = 10 – (1/20)P. The social
marginal cost of producing an office visit is $100.
(a) What is the socially optimal number of visits per year?
(b) Suppose that John has health insurance and only has to pay a $20 copayment for each office
visit. How many office visits will he make per year?
(c) Explain intuitively why the situation in (b) results in deadweight loss relative to the situation
in (a). - Causal Effects of Insurance
(a) Suppose that we try to estimate the causal effect of having insurance on health outcomes
using purely observational data, comparing those with and without insurance and controlling
for observable characteristics. Please give one reason why such an exercise could
overestimate the effects of insurance on health and one reason why it could underestimate the
effects.
(b) Explain briefly why the Medicaid program lottery in Oregon addresses the concern about
selection on unobservable characteristics and allows the authors to estimate the causal effect
of insurance on outcomes.
(c) Suppose that we had credible causal evidence that insurance does not improve health
outcomes. Does this imply that there are no welfare gains from the provision of insurance?
Explain, drawing on evidence from the Oregon experiment.
(d) Suppose we found credible causal evidence that insurance causes individuals to spend more
on health care services overall and that expanding insurance coverage does not “pay for
itself” through increased preventive care and/or reduced emergency room usage. Does this
mean that expanding health insurance is a bad idea? Explain. - Oregon Health Insurance Experiment – Readings Response
This part of the assignment MUST be completed individually.
Please read the readings for Tuesday’s class (March 10) on the course website. These papers are
all short, and you should read them in the following order:
1) “The Effects of Expanding Medicaid – Learning from the Oregon Experiment.” (NEJM
2011)
2) “The Oregon Experiment – Effects of Medicaid on Clinical Outcomes.” (NEJM 2013)
3) “Medicaid Increases Emergency Department Use: Evidence from Oregon’s Health
Insurance Experiment (Science 2014)
Please respond to the following questions:
(a) What did you find most interesting or surprising about the set of findings from the
Oregon experiment? Explain. (100-150 words)
(b) Did these findings cause you to change any of your beliefs about what types of health
care policies should be pursued when thinking about uninsured and/or low income
populations? If so, explain. If not, are there reasons you were unconvinced by the
findings? (100-150 words)
Sample Solution