Health Economics Review
Established in 1965 (under the Social Security Act)
Provides care to elderly (65+) and some disabled.
4 parts- A,B,C,D
Financed through payroll taxes, beneficiary premiums, and general federal revenue
1965: President Johnson signed Medicare bill
-Ended Stage renal disease eligible in 1973
1980-mid 1990’s: Payment reform
-Hospitals: prospective fixed payment by diagnosis in 1983
-Physicians: payment tied to the estimated cost of inputs in 1992 1997: Balanced Budget Act
-Cut provider payments by limiting growth rate in Medicare fees -Part C introduced: competitive within Medicare system
2003: Bush signed Medicare Modernization Act
-Prescription drug benefits “Part D”
Medicare Enrollment trend- Expected to double from 2000-2030 Medicare Spending- 7.7 Billion (1970)- 532.7 Billion (2012)
Sources of Financing
1. Hospital impatient largely financed by payroll tax
2. 238.6 Billion in spending
3. Could dissolve in decade
1. Outpatient and physician. Financed mostly by general revenue.
Part C- Since 1970 Medicare beneficiaries could choose between Medicare and privatized. A private plan was paid by government. Originally only a HMO.
- Originally put to reduce costs through better coordination
- Plans required A and B as well
- Could provide additional or small benefit in to Part B
Balanced Budget Act (1997)- “Medicare+Choice”
- Alternative to traditional Medicare
- Expanded options for private plans (PPO’S) (PSO’S)
Medicare Modernization Act (2003)-
-reception of drug coverage
-Changes in payment methods
-Prvate plans create small networks to reduce costs.
We also discuss several other topics like what is amphipatic?
Don't forget about the age old question of What is the role of the “support” in a communications brief?
Part C provides extra benefits. Also designed in a way where people pay higher premiums depending on sickness level to discourage cherry picking healthy patients. Not a perfect system however. From 1997-2003 Medicare PC Plans Don't forget about the age old question of What is Phospholipid?
- Raised premiums, Reduced prescription drug benefits, increased beneficiary cost sharing
Medicare Advantage Payment Plan-
-95% of fee-for service cost given to HMO in a given county.
- 1997 Congress changed payment structure and in the next few years the number of participating plans dropped in half. We also discuss several other topics like What is consciousness?
- To encourage participation MMA introduced bidding process. (Will cover more in ACA)
Part D- Prescription drugs
Seniors without drug coverage more likely to forgo medicine.
- Chandra, Gruber, Mcknight-Studied Relationship between decrease generosity of coverage and hospital savings. 1. They found 20% of the savings are offset by higher hospital spending (displacement) 2. For sick elderly it actually costs more. 3. 43% offset with chronic care. Summary- far from free, but the better alternative.
Impact of Part D-
-Large increase in prescription drug use in elderly.
-Increase Demand means more negotiation power with drug makers
Supplemental Insurance-Elderly spent almost 22% of income on health services in 2002 because of the gaps in Medicare Coverage. 89% of beneficiaries has some supplemental insurance by 2006.
Donut Hole- Affordable Care Act will eventually rid the donut hole by 2020. Which will greatly increase prescription drug spending.
-Established in 1965 along with Medicare. Provides health coverage (not services) to low income populations.
Financed by the state and federal governments. States establish their own programs within broad federal guidelines. Usually more comprehensive and generous that private. Low out of pocket costs or premiums.
Medicaid mandatory vs optional coverage and income requirements
Mandatory- 1. Physician services, 2. Laboratory and x-ray services, 3. Inpatient and outpatient hospital services, 4. Early screenings and diagnostics (under 21) 5. Rural and federally qualified health services 6. Nurse midwife services 7. Nursing facility services for those over 21 Don't forget about the age old question of How to Calculate components of the income statement?
Optional- 1. Prescription drugs, 2. Clinic services, 3. Dental, 4. Physical therapy, 5. Prosthetic, 6. Primary care, intermediate care for mental retardation, 7. Home health care services, 8. Personal care services, hospice services
Medicaid costs and the importance of long-term care- Federal Government accounts for 57% of Medicaid spending. Places particular financial pressure on states which that have to balance their budget
Financing of Medicaid- See above
Theory of private insurance crowd-out and empirical evidence- “Crowd Out” refers to the fact that the individual With Vl will drop current insurance to obtain free insurance. They have a higher preference toward other goods than toward insurance. When free public insurance is introduced, the better the benefits the more push there will be toward that then private insurance. If benefit outweighs cost than crowd out is effective if not it can be destructive. Gruber-Theorized that states with increased expansion is benefits for Medicaid should see reduction in private costs. They found 50% of increased in Medicaid coverage was due to reduction in private insurance coverage. Don't forget about the age old question of What is the meaning of Born-Haber Cycle?
Gruber article: The uninsured in the US (prior to the ACA)-
-Most industrialized nations guaranteed universal healthcare. However in the US 47 million people are still uninsured. The government is already heavily involved in the healthcare market accounting for over half of all health care spending.
-200+ Billion on subsidies. Could lead one to argue that this may not be the optimal policy. This article argues for and against. Siting major inefficiency and equity issues.
What are the main sources of insurance coverage in the US? Broken down into 3 groups. Private, public, and the uninsured.
Private- 80% of market, 90% coming from employers.
Why are people uninsured?
-High Administrative costs. Volatility in family size. Adverse selection. Rising insurance costs
- Have to help person in operating room regardless of if they have insurance or not.
Dynamic nature of uninsurance- Very dynamic. Rates in the past have been 40- 50% higher then they are now. Many factors such as government internvention, unemployment rates, ect.
Why should society care about the uninsured?
Externalities: Communicable disease
Labor Market Inefficiency- reduces ability to look for jobs, reduce mobility Insured are more likely to live longer, marginal health benefit from basic insurance greatly increases.
Behavioral responses to changes in public insurance coverage Take up and crowd out- with expansion in public care leads to crowd out Firm Reactions- as price increase there is a decrease % that firm will cover total cost of the premium
Gruber comparison of policy options and their relative costs and efficiency [L. 16]
The physician utility maximization problem
- Physicians maximize utilitywhy not profit?
-Physicians are both workers who supply labor and often are small business owners -As workers, physicians get utility from income & leisure time
-However, physicians also have the ability to induce their patients to demand care, but dislike doing so
-Utility = U(Y,L,I)
- Y = net income; UY > 0 (where UY is ΔU/ΔY)
- L = Leisure; time spent not working; UL > 0
- I = degree of ‘artificial demand creation’; UI < 0
Physician labor supply curves and substitution versus income effects
Income vs. Leisure
-can result in a backward bending supply curve of labor
-initially, ‘substitution effect’ dominates & higher wages result in more hours worked
-at some point, the ‘income effect’ dominates and results in fewer hours worked as wage increases
Supplier induced demand model
Relies on the familiar concept of information asymmetry b/w provider & patient
-Theory Physicians can alter quantity demanded, all else equal, by varying the ‘accuracy of advice’
-Trade off the utility gains from additional income earned form extra services with the disutility or conscience of inducing demand
-A decrease in the profit rate will, in theory, lead to more demand inducement
-Empirical evidence supporting:
VERSION 1: Physician supply increases quantity of services
- Fuchs: estimates the elasticity of surgical operations per capita to surgeons per capita is 0.28
-That is, a 10% increase in the supply of surgeons increases the number of surgeries by 2.8%, after controlling for the fact that surgeons prefer to locate in areas with higher demand for surgery
One theoretical explanation for this is the target income hypothesis:
-Physicians expect to make a certain income (after large investment in training) and if their income is too low they induce demand to increase income
VERSION 2: Increased supply increases price
Monopolistic competition and provider quality differentiation can also explain this:
- If quality matters to consumers but is difficult to observe, then all else equal if there are more physicians then consumers have less information about quality on average (think about information coming from word of mouth recommendations in a small town vs large city. Everyone knows the doctor(s) in small town, but don’t know the full set of choices in a large city.)
- If consumers have less information about quality, they have more inelastic demand, so optimal prices rise
- Empirical evidence opposing model
Dranove & Wehner (1994) Supply-induced pregnancies
- Use same statistical approach from other studies, but test whether increase in supply of obstetricians can “induce demand”
- Find evidence that obstetricians do induced demand
- This is obviously not possible, so suggests that there is something wrong with the statistical model that is probably also wrong in the other studies as well
Patient reactions to supplier induced demand and empirical evidence
Patient accepts treatment recommendation based on physicians reputation for inducing demand, but patient selects a treatment cutoff to optimize taking physicians response into account
- Suppose patients have a certain condition have severity S, and each doctor chooses a threshold Sd above which they treat and below which they don’t
- Patients could also have a severity threshold Sp above which they are willing to accept treatment
Empirical Test of Dranove (1988)
Predicts that the market should punish physicians who induce demand “too much”
Dranove and Ramanarayanan (2008) test for this effect for caesarian sections
1) Model treatment choice conditional on observables
2) Model physicians treatments relative to predicted rate given patient observables
3) Estimate demand for physicians based on “excess rate” Result:
After controlling for differences in demand, physicians that are overly aggressive in inducing demand suffer an 8% drop in market share
Small area variation in Medicare spending
A large amount of variation in procedure use and spending occurs across small-market areas
Dartmouth Atlas of HealthCare tracks these types of variations for Medicare
Categorizing differences in spending as underuse, misuse, and overuse
∙ UNDERUSE of effective care
Services that are of proven value and have no significant trade-offs (e.g. eye exams for diabetics)
Receipt of recommended care in Medicare is inversely correlated with spending
– Too many doctors with none clearly in charge and responsible for managing organization of care
– Financial incentives in accountability are lacking
∙ MISUSE of preference-sensitive care
Misuse of care that involves trade-offs affecting quality and/or length of life where consumers tend to have strong (but potentially uninformed) preferences (e.g. hip/knee replacements)
- Practice style hypothesis: Local medical practice styles or “medical culture” has a strong influence on treatment choice rather than recommended guideline
- Imperfect information about “best” medical decisions
- Not supply driven
∙ OVERUSE of supply-sensitive care
Care whose frequency of use is not determined by well-articulated medical theory (e.g. hospitalizations, physician visits, etc.)
Variation in this care explains most of the variation in Medicare per capita spending
High spending regions not associated with better outcomes What can explain small area variations?
• Supply and Demand Factors
– Explain between 40 and 75 percent of variation according to several analyses
• Imperfect information, physician uncertainty, and differences in practice styles likely account for most of the remaining 25-60% of variation
Gawande article on small area variations in medical spending Productivity spillovers and the marginal benefit of medical care
Productivity Spillovers (Chandra & Staiger 2007) With productivity spillovers there may different equilibria across locations
– Productivity spillovers occur when the share of patients treated in a certain way (e.g. more or less intensively) help other patients treated in that way
– Leads to some locations being more intensive and some less because marginal patients are more appropriate for a treatment based on the other patients treated in that area
Example: performing more surgeries makes surgeon better at doing themincreasing net benefit to getting surgery
Empirical evidence on the effects of financial incentives on physician care
Financial Incentives (Clemens & Gottleib 2014) : Physicians in many areas had exogenous shock to reimbursement rate from Medicare
- Exogenous change in prices
- 2% increase in payments leads to a 5% increase in care provision - elective procedures respond more strongly
- increases the pace of technology diffusion
- no impact on health
Physician report cards and their effects on quality versus risk selection
Consumer choice related to quality even without report cards
– Report cards help spread information more efficiently than word-of mouth, and may contain different kinds of information than consumers would typically use to form opinions
– Consumers have difficulty understanding report cards
– Physician agents under-use report cards
– Consumers respond to report card information
o Generally healthier individuals more likely to use them
Risk Adjustment & Selection
When patients vary in “difficulty” and providers know that their outcomes are being monitored, there is an incentive to select only the easiest patients to treat
Quality measures require risk adjustment:
– Estimate the probability of a bad outcome based on observable patient attributes
– Predict the likelihood that a provider would have been successful given the patients characteristics
– Adjust outcomes for the relative difficulty of patients treated
Empirical evidence on the effects of report cards on outcomes, access, matching, and total welfare
Dranove et al (2003)
– Increase in total CABG volume driven by work on less severe patients – Improvements in matching
– Increased spending overall
– Worse outcomes for more severe patients with little gain on healthier
o More skilled surgeons prefer to avoid very complicated cases, so severe patients are worse off after report cards are released
– Conclude that report cards decreased social welfare overall
Physician pay for performance: performance incentives and practical challenges to implementation
P4P intention is to alter the incentives to providers to increase emphasis on quality of care
Financial incentives for physicians vary between two extremes – FFS: Encourages the use of more care
– Capitation: Encourages the use of less care
Neither specifically encourages the use of quality care
Are the performance measures reliable?
Are the financial incentives adequate?
– Enough to cover costs of collecting data, etc...
Will providers participate and will patients respond to quality recognition?
Will physicians pick healthy patients or ‘teach to the test’ at the expense of other outcomes?
– Risk adjustment must be sufficient
– Patient compliance
– Patients would have an incentive to understate improvements in health (reduces reimbursement price)
– Providers have incentives to overstate the difficulty of the case, assuming payment accounted for difficulty
o If payments aren’t risk adjusted this could cause selection problems, where physicians prefer treating healthy patients
Empirical evidence on pay for performance effects
Premier Hospital Quality Incentive Demonstration:
Quality scores improved in the first two years of the demonstration: – From 87.5 percent to 94.4 percent for patients with AMI (heart attack) – From 84.8 percent to 93.8 percent for patients with coronary arter bypass graft. – From 64.5 percent to 82.4 percent for patients with heart failure. – From 69.3 percent to 85.8 percent for patients with pneumonia – From 84.6 percent to 93.4 percent for patients with hip and knee replacement. Variance in quality scores also diminished
What financial incentives do HMOs use to save money?
Main incentive structures are:
- Physician level global budget for the cost of treating all the patients of one physician
- Group level incentive if the costs for of an entire group of physicians (group size 3 to 30) were below budget in a year
∙ Gaynor et al findings
- Find that for a physician group of size 10, the group incentive leads to reduction in medical spending of about 5%, which saves the HMO $3.2 million per year
- Physician’s primary source of income in the HMO still comes from fees for services—they keep 80% of revenue they generate
As a result, group incentives should have a larger effect on reducing referrals to other doctors for care, less effect on reducing care
- Find that about 91% of the reduction in expenses caused by the group incentive is achieved by reducing expenses from referrals, not from own care
- Corroborates evidence that physicians make decisions in part based on own financial interests
Mechanisms for coverage expansion under the ACA
Expansion of Medicaid eligibility for adults up to 138% of federal poverty line
Income-based insurance premium credits and subsidies for insurance purchased on the exchanges, up to 400% of federal poverty line
Tax credits for small businesses that provide insurance for employees
Medicaid eligibility changes
Prior to ACA, Medicaid program had no mandatory coverage of adults without children
– Some states voluntarily chose to offer limited coverage, but most (44/51) didn’t
Under the ACA as originally written, the income limit for mandatory Medicaid coverage for everyone under 65 was expanded to 138% of FPL
Financing of Medicaid expansion
To finance expansion of the program, federal government pays for 90% of costs of newly eligible enrollees, states pay remaining 10%
– Initially federal government pays 100%, but that rate gradually declines to 90% by 2020, then stays at 90%
National Federation of Independent Business v. Sebilius Supreme Court case, and consequences for Medicaid expansion
∙ Impact on variation in eligibility across states
– in states that opt out of Medicaid expansion there will now be a gap between Medicaid and exchange subsidies
∙ Coverage gap caused by ruling
– Many families earning below 100% of the poverty line but above the Medicaid threshold in their state will remain uninsured, and will fall into a coverage gap
Individual insurance premium subsidies
Federal credits to purchase insurance through exchanges
After credits, price of insurance faced by consumer is capped at: – 2% of income if income between 100-133% of FPL
– 3-4% of income if income between 133-150% of FPL
– 4-6.3% of income if income between 150-200% of FPL
– 6.3-8.05% of income if income between 200-250% of FPL
– 8.05-9.5% of income if income between 250-300% of FPL
– 9.5% of income if income between 300-400% of FPL
Credit amount is based on the price of the second cheapest “silver” plan in the area
– Amount of credit equals price of 2nd cheapest silver plan minus price cap based on income rules above
Subsidies for copayments, coinsurance, and deductibles for plans purchased through exchanges
All out-of-pocket costs are reduced for “silver” plans to make the actuarial value of plans higher for low-income groups
Effects of cost-sharing subsidies on moral hazard at both individual and market level
The economic costs of affordable access to health insurance through cost sharing subsidies equal:
– DWL from taxation necessary to pay for the subsidy
– Plus the DWL from moral hazard created by reduction in cost sharing
Could also be a reduction in DWL from moral hazard in unsubsidized market due to price increase for healthcare caused by subsidy
Small business tax credits
∙ For small businesses with 10 or fewer workers and average annual earnings of $25,000 or less, business can purchase insurance for workers through the exchange and get a tax credit of 50% of the employer’s contribution to the premium
∙ Tax credit phases out from 50% to 0% as size of firm increases from 10 to 25 workers, and as average annual income increases from $25,000 to $50,00
∙ Tax-exempt small businesses can get credits too, but credits are 30% smaller
A. True/False Explain.
Indicate whether each of the following statements is true or false and then explain why you think this. Include in your explanation any pertinent institutional details and economic reasoning (including appropriate graphs and equations). Points will be awarded for correct, concise, clear answers with minimal irrelevant detail. Explanation is required to receive points.
1) As a result of the so-called “managed care backlash,” HMO and PPO plans are much less common today than they were ten years ago.
False. HMOs are somewhat less common, but PPOs are much more common than they were in the late 1990.
2) A typical episode during which a person goes without health insurance in the United States tends to last longer than 1 year.
False. The median episode of uninsurance lasts much less than one year. The Gruber article discusses evidence of this, in that the rate of uninsurance at any point during a year is about 40-50% higher than the rate of uninsurance at any single point in time.
3) By 2015, there are expected to be almost no uninsured individuals in the U.S. because of the recently passed Affordable Care Act.
False. The remaining uninsured (~30+ million people) will consist of illegal immigrants and people who do not receive insurance through their employer, do not qualify for Medicaid, and choose not to purchase insurance through the exchange, in addition to people who do not qualify for Medicaid in the 24 states that chose not to accept the Medicaid expansion of income eligibility.
4) Under a capitation arrangement between physicians and insurance companies, there is an upper limit, or “cap”, on the total amount of the medical care that a patient can receive.
False. Under a capitation arrangement physicians are paid a fixed amount for the care of a patient, without regard to the actual services provided. The revenue to the physician is set in advance, but the amount of medical care the patient receives is not limited in any way.
5) One important reason that incomes for surgeons are higher on average than incomes for primary care doctors is that there are greater restrictions in the supply of new surgeons.
True. The entry barriers to becoming a surgeon (in the form of limited residency slots relative to the number people who want to become surgeons) are greater than the entry barriers into primary care medicine.
6) In the article by Gruber (2008) on covering the uninsured, Gruber estimates that if a public policy were to reduce the rate of uninsurance by 5-10 million people, it would be a more efficient use of public fund to expand to public insurance programs like Medicaid than to issue tax credits that could be used to purchase individual insurance in the private market. (Although the definition of “efficient” could be subjective, for the purpose of this question assume that efficiency means maximizing the value of insurance coverage per dollar of public spending.)
True. Gruber estimates that it would cost between 93-99 cents of public funding to provide $1 worth of health insurance if the coverage expansion is implemented as an increase in the income threshold for Medicaid eligibility. If instead the expansion is implemented by giving a similar low-income group tax credits to purchase insurance in private individual insurance markets, it would cost between $2.35-$4.01 of public funds to provide the same $1 worth of insurance coverage.
7) The reason why there is a shortage of physicians and nurses in the US is that, relative to other countries like Japan and Australia, the ratio of physician incomes to average incomes in all jobs is lower in the US, which decreases the relative rate of return to attending medical school.
False. Physicians in the US earn more than 5.5 times more than an average worker. In other countries this ratio is much lower, around 2-3.
8) Under the Affordable Care Act, all individuals in the US are required to have health insurance or else they are breaking the law and could be arrested.
False. The individual mandate under the ACA is a tax penalty for noncompliance. You can’t be arrested for choosing to not purchase insurance.
9) Recent proposals to change the definition of full-time employment in the Affordable Care Act to 40 hours a week would cause more people to gain employer-provided insurance under the ACA compared to the current definition of full-time.
False. The full-time employment definition under the ACA affects the cutoff above which firms must pay a fine under the employer mandate if they fail to provide insurance for their workers. The mandate does not apply to workers who are not full time. Raising the definition of full time from 30 to 40 hours would decrease the strength of the mandate and reduce the number of people with employer-sponsored insurance.
1) Describe the differences between HMO, PPO, POS, and indemnity insurance plans in terms of the ways that each plan uses specific supply-side restrictions to control medical spending.
The four dimensions on which these plans differ are in their use of provider networks, gatekeeper physicians, selective contracting, and utilization reviews. HMOs use all of these supply-side approaches to reducing medical spending, while indemnity plans use only utilization review. PPOs have no gatekeepers, so there is no need to get referrals to see a specialist, for example. POS plans are like a hybrid between HMOs and PPOs, and they use gatekeepers, selective contracting, and utilization review. However, in POS plans consumers can opt out of the network if they choose, and face higher cost sharing. In that sense POS plans don’t have restrictive provider networks.
2) A hospital has two types of patients, privately insured patients and Medicare patients. Suppose Medicare cuts its reimbursement rates to hospitals, but the reimbursement rate is still higher than the average variable cost of treating Medicare patients. Describe, using any relevant graphs/diagrams, what economic theory suggests will happen in the private sector market for hospital care.
In theory, the private sector market will be unaffected by the change in Medicare reimbursement. In the private market hospitals choose prices to maximize their profits. Profit maximization is determined by setting marginal revenue equal to marginal cost. Neither the marginal revenue curve nor the marginal cost curve in the private market are affected by the Medicare reimbursement rate, so optimal prices are also not affected by Medicare reimbursement rates.
3) Describe the main way in which the design of consumer-directed health care (CDHC) plans is different from the design of other insurance plans. All else equal, how does this main difference in plan design affect the financial risk protection of consumers and how does it affect moral hazard?
The main differences between CDHC plans and other typical health insurance plans are that CDHC plans have higher deductibles, but come with a health savings account that can be used to pay for part of the deductible. If the health savings account is not spent, the consumer can keep the balance in the account to use for future medical expenses. This gives consumers an incentive to be price-sensitive. If they don’t value some type of medical care, they are less likely to buy it if the money will come out of their private health savings account than if the insurance company will pay for it. As a result, moral hazard will decrease, but the consumer can have similar levels of risk protection as a typical plan, because the health savings account can be used to pay for the larger deductible amount, if necessary.
Selection in HMOs and Fee-for-service Insurance Plans:
Suppose there is a continuum of consumers who all have medical needs determined by the severity of their illnesses, X, where X ranges from 0 (healthiest) to 10 (sickest). Each person can choose between a fee-for-service (FFS) insurance plan and an HMO. The FFS plan has a high coinsurance rate but no supply-side restrictions on the use of medical care. The HMO has a lower price and lower coinsurance rate, but also has supply-side restrictions that consumers dislike, especially when they are sicker. Everything else about the plans is the same. Suppose consumers are willing to pay an amount of money V to avoid the supply-side restrictions in the HMO option, where V depends on the severity of their illness:
The differences in prices and coinsurance rates cause there to be a difference in total out-of-pocket spending under the two plans as well, where the total cost to the consumer in the FFS minus the total cost in the HMO plan equals E:
a) What is the average severity of consumers that choose the FFS plan?
Consumers will sort into the FFS or HMO plan based on their severity. The crossing point is where V=E, which occurs when X=5. The consumers with severity between 5 and 10 will choose the FFS, and the average severity is 7.5.
b) What is the average severity of consumers that choose the HMO plan?
Similar to a), consumers with severity between 0 and 5 will choose the HMO plan, and the average severity is 2.5.
Now suppose these two insurance plans become subsidized at a 50% rate as a result of a health insurance reform policy. The subsidy causes E to change to:
c) After the subsidy, what are the average severities of consumers that choose the FFS and HMO plans?
When E changes the crossing point changes to 1.25. The average severity in the HMO is 1.25/2 and the average severity in the FFS plan is (10-1.25)/2
d) Is the policy change likely to affect the total deadweight loss from moral hazard in the population? Why or why not?
In each of the plans the generosity increases after the subsidy, decreasing the marginal out of pocket cost of medical care, and increasing moral hazard.
On the other hand, there could potentially be differences in the level of moral hazard in the FFS plan and the HMO plan. For example, if the marginal out-of-pocket costs are higher under the FFS plan and
the supply-side restrictions under the HMO aren’t as effective at reducing moral hazard as the higher cost sharing under FFS, then the total deadweight loss from moral hazard could also depend on the how many people choose the FFS plan versus the HMO plan. Since more people choose the FFS plan after the subsidy, it’s possible that total moral hazard costs increase due to selection.
e) Is the policy change likely to affect the total deadweight loss from adverse selection in the population? Why or why not?
There is less adverse selection in the market after the subsidy. Recall from the Akerlof model that an indication of adverse selection is that only the sickest people remain in an insurance market. After the subsidy, most of the population is in the same insurance pool as the sickest people, reducing adverse selection. Another way to think about this is to conceptualize the subsidy growing very large. As the subsidy gets close to 100% the FFS plan becomes nearly free, so everyone will choose that plan. When everyone chooses the same insurance plan there is no adverse selection. Increasing the subsidy tends to reduce adverse selection costs.
Econ 5860: Health Economics
Professor Kurt Lavetti
Ohio State University
Spring 2015 Problem Set 2
Due Date: Feb 25, 2015 by the beginning of classPlease submit assignments on Carmen.osu.edu or on paper in class
A. True/False Explain. Indicate whether each of the following statements is true or false and then explain why you think this. Include in your explanation any pertinent institutional details and economic reasoning (including appropriate graphs and equations). Please provide concise, clear answers with minimal irrelevant detail. Explanation is required.
1. Suppose you are interested in estimating the elasticity of demand for medical care. A bad way to do this would be to compare people who are healthy to people who are sick, and estimate how prices affect the amount of additional care that sick people choose to purchase.
∙ FALSE Common sense & theory suggest that demand shifts out when illness occurs (sick people need/demand more health care than a healthy person) , as well as the way in which demand reacts to changes in price in a market will show elasticity. The more ill a person becomes the more inelastic their demand for health care is, and vise-versa.It would be important to also consider the effects of health insurance
2. For a private insurance company, providing coverage for monthly gym memberships can lead to higher profits for the plan.
∙ TRUE offering a gym membership is effective way to attract their target customers, fit people. While also avoiding attracting unfit/sick people who don’t go to the gym.
3. In Figure 1 below, an individual with high levels of education has marginal efficiency of investment given by MEI. The marginal efficiency of
investment for an individual with low levels of education is given by MEI’.
∙ FALSE 1st, higher levels of education result in higher income/wages & as income increases so do standards of living and health. 2nd, MEI =
- G=Marginal effect of health stock on
- W=Wage rate
- SP=Marginal value of health investment
» So, as a persons market value per unit of time (wage) increases, the value of having fewer sick days (G*W) rises,MEI curve shifts right & up, so the opposite is true in this case.Note that SP also rises, but by as much.
4. Supplemental ‘Medigap’ insurance policies reduce the federal costs of Medicare by covering 20% of Medicare Part B costs.
∙ FALSELower coinsurance increases quantity consumed because of moral hazard effect
» Higher quantity consumed ’s MKT prices
B. Analytical Problems
5. An individual has a health insurance plan with a deductible of $1500 and a coinsurance rate of 20%. Their demand curve is Q=20-0.1*P, and the equilibrium market price of medical care is $100 per unit. What quantity of medical care would the individual choose to
∙ Q = 200.1(0) = 20 ; 0=200.1p p=200 ; Q=200.1(100) = 10 ; Q=200.1(20) =18 ∙ 1500/100 = 15
∙ consumer would consume 10 units
6. Consider a version of the Akerlof model in which neither buyers nor sellers observe car quality (though somehow – please suspend your disbelief – both buyers and sellers enjoy higher utility from higher quality cars). For this question, please assume that both buyers and sellers recognize that neither can observe car quality.
Sellers’ utility function is given by US = M + ∑ xi and buyers’ utility is given byUB = M + ∑ 32 xi where M is the level of consumption of non-car goods and xi is the quality level of car, and there is a uniform distribution of quality of the cars held by sellers, xi~U[0,2]. In this market, is there is a price, p, at which all cars will sell? If not, prove there is no such price. If so, calculate what prices will work.
∙ Yes, the information is symmetric—equally bad on both sidesboth will make best guess that a given car is of average qualityQ=1
- Owners have a reserve value on their car, say = $5,000 X Q
- Nonowners more eager so value them more, say = $7,500 X Q
- So, at price (p)= $7,500 and avg. Quality (best guess) = 1 all owners are willing to sell their car (75005000=2500) and all nonowners are willing to buy at this price.
- Thus, the market exists and clears (supply = demand)
7. Demand for Insurance. The next two questions refer to Figure 2 below.
Suppose individual A and B both have a 1⁄2 probability of receiving Y and a 1⁄2 probability of receiving YL. Individual A is more risk averse than individual B.Label which of the graphs above describes individual A and individual B’s utility function. Then show each
individual’s willingness to pay on the graphs and explain which individual has a higher willingness to pay for actuarially fair insurance.
∙ Individual (A) graph1
- Willingness to pay at point of intersection
[U(YL), YL ]
∙ Individual (B)graph 2, has higher willingness to pay, given a fair premium
- Willingness to pay at point of intersection
[U(YL) , YL]
a. Suppose the insurer offers an actuarially unfair, full insurance contract. Describe the cutoff “price” at which each individual would become indifferent between paying for unfair insurance and choosing to become
uninsured. Is the cutoff price for A greater than the cutoff price for B?
∙ The cutoff “price” at which each individual becomes
Econ 5860: Health Economics
Professor Kurt Lavetti
Ohio State University
Problem Set 1
Due Date: Feb 11, 2015 by the beginning of class
Please submit assignments on Carmen.osu.edu
1. Applying microeconomic concepts to a randomly selected news story that concerns health/health care.
Look for a health/health care news story in a news source of your choice this week (http://www.kaiserhealthnews.org/ is a good health news source in case you don’t have one in mind)
a) Please write down the title of the story, the name of the newspaper, and the date (and attach the 1st page of the story if you can).
b) Write one paragraph briefly describing the story in your own words.
c) Explain in a paragraph what economic issues you see at play in the
story (be careful in justifying why they are economic, as opposed to, say, ethical issues).
2. Using supply and demand graphs to explain trends in the health sector.
Use a diagram to depict supply and demand for health care. Illustrate what would happen in the diagram if each of following events happens to the health care market. Please provide a full explanation for each change you show, and what it will do to price and/or quantity in the new equilibrium.
i) Everything else stays the same, but we find more efficient ways of producing the same health care
ii) There’s a communicable disease outbreak and everyone gets sick
iii) There’s an economic recession and income falls
The first important key to answering questions like this is to understand where the supply and demand functions come from.
Recall from micro theory that the supply function is derived from the marginal cost function. Supply functions have nothing to do with consumers at all, they simply represent the production technology of the firm and the input prices it faces.
The demand function comes from consumers’ maximizing their utility functions subject to the fact that they have a scarce amount of wealth to spend. So the things that affect the quantity of goods demanded by a consumer are: the consumer’s preferences (which are represented in the form of his/her utility function), the prices of goods, and the level of wealth. Since we graph the demand function on a graph whose axes are price and quantity, we can represent the relationship between quantities demanded and the price of the good in question directly on the graph. Changes in things that are not represented on the graph may cause the demand curve to shift. For example, changes in wealth, changes in preferences, and changes in the prices of other goods (like substitutes or complementary goods) can cause the demand curve to shift.
i) If we find more efficient ways to produce health care then the only thing that changes is the technology used by firms that produce healthcare. Only the supply curve will shift. An increase in efficiency will make it cheaper to produce a given amount of quantity, so the marginal cost curve will shift down. The supply curve is the upward sloping part of the marginal cost curve, so it too will shift down. Remember that price is on the vertical axis, so a downward shift means lower price and higher quantity, as shown below.
ii) Each person who gets sick from a communicable disease outbreak will want to buy more medical care, all else equal, and will be less responsive to prices once they become sick. Since the individual demand curves will shift out and become more inelastic, the total aggregate market demand curve will also shift out and the elasticity of demand will fall. The equilibrium price and quantity of medical care consumed will both rise, as shown below.
iii) An economic recession in which income falls will affect people’s budget constraints. If income falls, the demand curve will shift downward, and the equilibrium price and quantity will both fall, as shown below.
3. Provide a health-related explanation why the real hourly earnings of workers fell between 1975 and 1995 despite a significant increase in productivity. Consider the gap between hourly earnings and productivity, which has grown larger over time since 1975. Relative to the average worker, would you expect the growth in the relative earnings productivity gap over time to be larger or smaller for high income workers, any why? (Think about the “relative earnings-productivity gap” as real hourly earnings as a fraction of productivity)
Recall this graph from lecture 4. Historically, real hourly earnings of workers have risen as productivity rose over time. However, as health insurance and other non-wage job benefits became more expensive in real terms, health insurance became a much larger fraction of total compensation on average. In fact, since health insurance costs grew faster than productivity, real wages actually decreased in order to pay for the higher health insurance costs.
For high income workers, the share of earnings devoted to health insurance is relatively small, so paying for the higher cost of health insurance doesn’t offset all of the productivity gains. There are still gains left over after paying for health insurance, allowing real wages to rise.
Consider a low income worker, in contrast. Suppose the worker starts with an income of $20,000 and gets health insurance that costs $5,000. Suppose productivity gains increase real compensation by 10%, but health insurance costs jump from $5,000 to $10,000. Total compensation will be 10% higher, or $30,000, but $10,000 of that will go to health insurance, leaving wages unchanged despite the large productivity growth. If the worker had instead started with an income of $100,000, then even after the increase in health insurance costs they would have a wage increase to $105,000.
This difference across income levels in the share of productivity growth devoted to rising health insurance costs is one reason why income inequality, when measured just based on earned income, has risen over the past several decades.
4. Consider the following change in medical technology. Based on your cost-benefit analysis, would you recommend switching to the new technology? Please show all calculations.
A new surgical procedure reduces the amount of recovery time necessary. The new procedure costs $25,000, and there’s a 50% chance that you will live 1 year and a 50% chance you will live 9 years. Because of limitations associated with the surgery, your QALY weight will drop from 1 to 0.6.
The old procedure costs $75,000 and had a 50% chance of surviving 2 years and a 50% chance of surviving 7 years. The QALY weight is 0.7.
Assume the value of a QALY is $100,000, and ignore discounting and consumption costs.
The benefit of the new procedure is:
(0.5*1*100,000+0.5*9*100,000)*0.6 = $300,000
The benefit of the old procedure is:
(0.5*2*100,000+0.5*7*100,000)*0.7 = $315,000
The new procedure has $15,000 less in benefits, but costs $50,000 less than the old procedure. You should use the new technology based on this cost benefit analysis.
5. Suppose Colorado decides to introduce a tax on the sale of marijuana. Assume the elasticity of demand is -0.4, and that supply is perfectly elastic.
i) If the tax increases the price of marijuana by 15%, what will happen to the amount that is sold?
ii) Suppose marijuana is addictive, and that consumers who choose to buy it are rational addicts. Explain what the theory of rational addition predicts will happen to the quantity sold today if the tax increase begins next year. (Note: the theory of rational addiction says nothing about stockpiling goods to avoid paying taxes, so that is not the answer I am looking for.)
iii) (Bonus) Suppose there is a legal market with perfectly elastic supply that pays the tax, and a secondary black market that evades the tax. The elasticity of supply in the black market is 1 (the number doesn’t matter, just assume that it’s neither perfectly elastic nor perfectly inelastic), and assume that after the tax it’s cheaper to buy from the black market at low market quantities. Draw an example of what the market supply curve would look like. (Hint: the market supply curve represents the lowest price at which you could buy a given quantity of a good.) Now suppose the state is short on money so they raise the tax rate. Show using a supply-demand diagram what will happen to total tax revenue.
i) The elasticity of demand equals the % change in quantity divided by the % change in price. The percent % in price is 15% and the elasticity of demand is -0.4, which implies the % change in quantity is -0.4*15% = -6%. Quantity sold will fall by 6%.
ii) The theory of rational addiction, which we covered in lecture 7, suggests that people become addicted to goods despite being perfectly rational and forward looking. This is contrary to the myopic theory of addiction, which assumes that people become addicted because they have some cognitive limits, or make decisions based on incorrect information. It is also contrary to the imperfectly rational model of addiction, which suggests that addicts have time-inconsistent preferences that cause them to value utility in the present much more than they care about costs that occur in the future. Only the theory of rational addiction assumes that addicts have perfect information, are rational, and forward-looking.
Under this theory, if the tax on an addictive good is expected to increase in the future, rational addicts will anticipate that if they consume a lot of the addictive substance now, their stock of addictive capital will increase, and in the future they will have to pay more in taxes because they will be more addicted to the good. Anticipating this, a rational addict would decrease their consumption of the good today in order to reduce their addictive capital tomorrow, and thus purchase less of the addictive good tomorrow. If the future tax increase is large enough, a rational addict could also stop consuming the addictive substance entirely to avoid purchasing any of the addictive good at the higher future price.
iii) The demand curve in the black market is upward sloping, and the demand curve in the legal market has a horizontal supply curve. The market supply curve is defined by the lowest price at which one could buy any given quantity. At low quantities this is the supply curve in the black market, and at high quantities it is the supply curve in the legal
market, so the market supply curve is upward sloping at first, and then has a kink after which it is flat.
An increase in the tax rate in the legal market will shift the kink to the right. The effect of this tax increase on tax revenue depends on where the demand curve intersects the new supply curve. Assuming that both markets existed in the first case (ie the demand curve is far enough to the right that it intersects the old supply curve to the right of the kink) there are two possible scenarios. If the demand curve intersects the new supply curve to the left of the kink, then tax revenue will fall to zero, so the increase in the tax rate will push everyone into the black market and actually reduce tax revenue. If the demand curve intersects the new supply curve to the right of the kink, then the change in tax revenue equals the change in quantity in just the legal market alone multiplied by the size of the tax rate increase, as shown in the graph below.
b c d
Initial tax revenue = b+c+d
New tax revenue = a+c
The change in tax revenue = (a+c)-(b+c+d)=a-b-d
Sno tax Q
Econ 5860: Health Economics
Problem Set 3
Indicate whether each of the following statements is true or false and then explain why you think this. Include in your explanation any pertinent institutional details and economic reasoning (including appropriate graphs and equations). Explanation is required.
1. For‐profit and non‐profit producers can simultaneously operate in equilibrium in competitive markets.
∙ TRUE: nonprofit & for profit producers can be viewed as differentiated. The type of competition among nonprofits is more regulated/restricted than among forprofit hospitals. In a market where only one of the two types exist, quality
of health services could possibly (reasonably) decrease.
Where both differentials exist, health services are better,
consumers have more value, which is the case (mostly).
2. If there is an infinite supply of altruistic donors for the non‐profit firms, the answer to Question 1 remains the same.
∙ False, nonprofits are more likely enter a market b/c their incentives are a mix of profit and altruism (Lakdawalla &
Philipson 2006). If supply is purely altruistic, there exists
no mix of profit, and they may decide not to enter the
3. In Medicare Part D, as total drug expenditures by an individual increase, the share of total outof pocket costs to the individual in the form of copayments and coinsurance as a fraction of total expenditures decreases.
∙ False, according to Medicare Part D amount is paid as a % of the cost, therefore no matter which way expenditure goes (up or down) the % paid by elderly stays same.
4. The term “Crowd Out” in the context of public health insurance refers to the fact that when Medicaid or other public benefits are expanded there is a tradeoff in access because waiting times to see doctors increases.
∙ FALSE: The term “crowdout” refers to the fact that the individual w/preferences (VL) will drop their previously held, employer provided insurance in order to get the free public insurance. (VL = values insurance less)
5. Gruber (2008) estimates that expanding public health insurance would cause more crowdout than providing vouchers to individuals for purchasing nongroup health insurance.
∙ TRUE: Gruber estimates the total amount of “crowdout” from public insurance expansion be about 17%, & the total resulting from nongroup tax credits to be 6.6%, therefore expanding public health insurance would cause roughly 10.4% more crowd out than nongroup tax credits would.
6. In his article on “Covering the Uninsured in the US,” describe what Jon Gruber means when he says that uninsurance status is very dynamic.
∙ The pointintime estimate according to the CPS data misses important dynamics of the uninsured. Estimates of uninsured
over an entire calendar year are about 1/22/3 as large.estimates of those uninsured at any point in the last year are close to 40— 50% higher. This is due in part to the fact that all uninsured are not poor, 20% are in families with incomes above $50,000 per year, 70% have family head that is a fulltime, but isn’t offered health insurance or chooses not to enroll.
7. (Optional, Not Graded) List one question related to US health care policy that you would like to know the answer to, and we have not yet discussed in class.
Analytical Question: Selection in HMO PlansSuppose there is a continuum of consumers who all have medical needs determined by the severity of their illnesses, X, which are uniformly distributed between 0 (healthiest) and 10 (sickest). Each person can choose between a fee-for-service (FFS) insurance plan and an HMO. The plans have the same premium, but the FFS plan has a high coinsurance rate while the HMO has a low coinsurance rate. Unlike the FFS plan, the HMO also has supply-side restrictions that consumers dislike, especially when
they are sicker. Everything else about the plans is the same. Suppose consumers are willing to pay an amount of money V to avoid the supply-side restrictions in the HMO option, where V depends on the severity of their illness:
The differences in coinsurance rates cause there to be a difference in total out-of-pocket spending under the two plans as well, where the total cost to the consumer in the FFS minus the total cost in the
HMO plan equals E:
a) What is the average severity of consumers that choose the FFS plan?
500(X) = 1000 + 200(X)
X = 3.33
V = 500(3.33) = $1665
b) What is the average severity of consumers that choose the HMO plan?
E = 1000 + 200(X)
E = 1000 + 200 (3.33)
E = $1666