Health Econ Midterm

Reasons why adverse selection would not occur?

*1. Customers misperceive their own risk:* -Ex. US and Swedish novice drivers demand less insurance because they believe they are safer than median driver *2. Customers do not act on their private information:* -Real customers have more important things to think about than small bargains on insurance -Customers may fail to realize their advantage (ex. car owners can predict how many miles they will drive in the coming year with good accuracy, yet they do not seem to take that into account when deciding how much insurance to purchase *3. Insurers can accurately observe customer risks:* -Long-term care insurers, with access to historical data and a team of analysts, are better at predicting whether middle-aged customers will eventually need nursing home care than the customers themselves *4. Advantageous selection:* -Sometimes it is healthier, less risky people who are more likely to buy insurance due to being more risk averse, having higher income, and a better understanding of insurance benefits

Different hypotheses for what causes health disparities

*1. Efficient producer:* better educated individuals are more efficient producers of heath than less well-educated individuals (ex. 1 year of schooling causes a 1.7 year increase in life expectancy) *2. Thrifty phenotype:* genetic reasons for being inefficient at producing, the deprivation of resources in utero and early childhood leads to activations of "thrifty" genes that are useful for sparse environmental conditions (ex. Dutch famine study, those in utero during famine had higher rates of diabetes and obesity --> thrifty gene told cells to hoard fat) *3. Direct income:* disparities exist because rich people have more resources to devote to health, rich individuals have an expanded PPF because of extra financial resources, causes optimal health to be higher (even applies to lottery winners) *4. Allostatic load:* prolonged or repeated stress is unhealthy and can cause an increased rate of aging (ex. Whitehall study on British civil servants, those at the lowest grade doing the most work had higher mortality rates than those at the highest grade) *5. Income inequality:* health disparities are caused by an unequal distribution of income *6. Access to care:* those with high income can afford more generous health insurance compared to those of low income (ex. Oregon medicaid exp.) *7. Productive time:* SES differences are caused by disparities and health (ex. poor health during infancy leading to higher mortality rates, lower educational achievement, and lower adult earnings) --Grossman model suggest this is more relevant for people with poor health *8. Time preference (The Fuchs hypothesis):* bad health does not cause low SES, and low SES does not cause bad health; there's a third factor (tie preference) that causes both --> time discounting (patience). Both SES and health are determined by willingness to delay gratification, so people who discount the future heavily are less willing to forgo present utility for higher future utility (ex. Marshmallow test) -Note: these are not competing theories

Modes of relationship between physicians and hospitals

*1. Physician's workbench (Harris model)* -Hospital provide a place for physicians to do their work, but do notdirectly employ them -What most hospitals in the US are *2. Physicians are direct employees of a hospital or hospital system* (ex. UK NHS) *3. Physician ownership of hospitals traditionally justified as a way to avoid the commercialization of medicine* -Typical of Japan Second and third types may avoid the conflict of interest between physician and administrators -But because doctors must also be mindful of costs, there is potential that doctors will fail to serve as ideal agents for patients (doctors may choose to save hospital money, rather than exploring every possible avenue of treatment)

Pooling equilibrium

-A contract that attracts both robust and frail individuals while also satisfying equilibrium conditions -Must be on the population zero-profit line, if to the right the firm loses money, if to the left then other firms can enter the market -Can always find a contract that deviates from the ______ _______, therefore no ______ ______ can exist, as the new firms entering the market will create adverse selection, in that they only attract robust individuals

Seperating equilibrium

-A set of contracts where one attracts frail individuals and the other attracts robust individuals, while satisfying equilibrium conditions -Only works when the robust can be distinguished from the frail -Need a contract that will not tempt frail individuals to leave their ideal contract, but will still attract robust ones (incentivize individuals to choose the coverage that is right for them, i.e. not offering full coverage to the robust will prevent the frail from being attracted to the policy) -Will occur at the intersection between the frail utility line and robust zero-profit line -Allows each type of person to maximize their utility -This can break though for example, when there are too many robust customers (share of sick individuals is very low) --> zero profit line for average population will be very close to robust line (only happens when the average line is above robust) --> creates a new region on the graph that would attract both frail and robust people

Asymmetric information

-A situation in which agents in a potential economic transaction do not have the same information about the quality of the good being transacted --In the insurance case, we are the sellers and the insurance company is the buyer, since theoretically we have more information about our health than they do -Ex. new assumption: sellers can determine car quality, but buyers cannot (all cars look identically good to the buyers) --Now any cars that sell sell for the same price, the best cars all not be offered, and possibility that the cars will not end up with the people who value them most (buyers) -Why is there only one price?: no buyer will want to buy the expensive car, because both cars will seem the same --> all sellers will have to lower their prices to match the lowest price on the market --> only the lower-quality cars stay on the market -Why are some cars not offered?: we know the market has one price, P, so consider the seller who owns the nicest car on the market it is probably worth way more than P --> no reason to remain in the market and can't get around this by advertising the quality since buyers can't see it -Causes the market to unravel and the Pareto-improving transactions do not occur (market failure)

Symmetric information equilibrium

-Achieved when firms can tell frail and robust individuals apart and can legally exclude certain risk types from certain contracts -if so they will offer the ideal contract which lies on an individual's zero-profit line

Actuarially fair insurance

-Actuarially fair means that insurance is a fair bet, i.e. when the premium equals the expected payout (asked to pay your expected expense, which is gotten by the probability of getting sick) --> r = p*q -Insurer makes zero profit/loss from actuarially fair insurance in expectation -Healthy state: IH' = IH-r --> IH' = IH-pq --> IH' = IH-p(IH-IS) --> IH' = pIS+(1-p)IH --> IH' = E[I]p -Sick state: IS' = IS-r+q --> IS' = IS-pq+q --> IS' = IS-p(IH-IS)+(IH-IS) --> IS' = pIS+(1-p)IH --> IS' = E[I]p -Customers with actuarially fair, full insurance achieve their expected income with certainty

The moral hazard pattern

-An individual faces some risk of a bad event X, and his actions can increase or decrease its likelihood -He has an insurance contract that will help pay some or all of the costs of X, if it occurs --> thus his price of X is now lower -In response to the price distortion, he changes his behavior in a way that increases the chance of X or increases the costs of recovering from X -The insurance company cannot observe this behavior change so there is an information asymmetry (otherwise the contract would have been written to discourage the riskier behavior) -The individuals riskier behavior creates a social loss, because the costly event X occurs more than it would have without insurance

Insurance and risk aversion

-As we have seen, simply by reducing uncertainty, insurance can make this risk-averse individual better off -Relative to the state of no insurance, with insurance she loses income in the healthy state (IH > IH′) and gains income in the sick state -(IS < IS′): In other words, the risk-averse individual willingly sacrifices some good times in the healthy state to ease the bad times in the sick state

Labor-leisure tradeoff

-At low levels of health, small improvements create large improvements in productive time and large drops in sick time -Once sick time and health improvement time are set, the individual allocates remaining time between work and play -Optimal allocation of time will be the point on the graph where the indifference curve is tangent to the time constraint line -When health improves, more productive time is available for use --> pushes time constraint outward (from U0 to U1) and can reach higher utilities

How healthcare is organized

-Based on two questions... 1. Is the demand curve for health care downward sloping? (Are people sensitive to the price of healthcare?) --If yes, people who face different prices or have different willingness to pay get different amounts of care 2. Do they end up with different health outcomes as a result? --If no, there is a very limited role for health econ -These questions are rarely an issue in countries where citizens receive subsidized health insurance or are eligible for free care from the gov, but in countries like the US, people facing different prices receive different amounts of health care

Tradeoffs from barriers to entry

-Because of barriers to entry, consumers have to pay above the competitive price (good because they prevent a surplus of information, but when too high they are bad because prices get too high) --> so physicians earn monopoly rents, which are stated to approach 25% of total physician compensation --Monopoly rents: wages above the competitive price due to artificial constraint of the market -Barriers to entry ensure that physicians are qualified -Costly for consumers to distinguish good physicians from bad on their own -If search costs to consumers in an unregulated market exceed monopoly rents in a regulated market, then barriers to entry may improve welfare

Ex post moral hazard

-Behavior changes that occur after an insured event happens and makes recovering from that event more expensive -Ex. using expensive drugs instead of generics, knee replacement surgery instead of painkillers -Ex. Stanford employees: after a 1967 change that required a new 25% copay, visits to the doctor declined by 24%

Ex ante moral hazard

-Behavior changes that occur before an insured event happens and makes that event more likely -Ex. leaving the stove on, skipping the flu vaccine -Ex. Ghana: insured households less likely to use mosquito nets, key for preventing malaria -Ex. Seguro Popular: low-income Mexicans assigned to receive free insurance were less likely to get a flu shot and cancer screenings

Bulk markups

-Bulk discounts: a lower per-unit price for a large purchase of a commodity -Bulk markups: a higher per-unit price for large purchases of a commodity --Insurance companies use bulk markups to protect themselves from risk customers who want a lot of insurance --This is exactly what the Rothschild-Stiglitz model predicts

Present-day barriers in US physician labor market

-Caps on medical school class size -Doctors need a license to practice on their own --International med graduates have a long and arduous process to practice in the US, so its very hard to do this --Nurses and Physician Assistants are limited in scope of practice (nurses cannot prescribe medication and physician assistants cannot conduct major surgeries), Insurers have policies that make it harder for nurses to complete medical care and other insurers routinely pay nurses less than licensed physicians for providing the same service, and many states require nurses to be supervised or associated with a physician in order go get payment --Alternative medicine: Chiropractors, acupuncturists, etc. need licensure too

Weaknesses of audit studies

-Chandra and Staiger (2010) attempt to determine whether the discriminatory prescriptions noted by Shulman et al. (1999) and other are efficient or inefficient, so they analyze Medicare patients admitted to US emergency rooms suffering AMI and compared rates at which black and white patients undergo procedures involving reperfusion (catheterization or open heart surgery) -Results: found that black patients are slightly less likely to receive reperfusion than white patients even after controlling other relevant risk factors (differential reperfusion rates imply discrimination like the Shulman et al. study (the results were consistent)) -After confirming they were consistent they wanted to determine what type of discrimiation it was --If discriminatiion was taste-based, then the average benefit for black patients receiving reperfusion would be higher than white patients -Researchers find that reperfusion benefits black patients less than it benefits white patients: see larger improvements for white patients -So, physicians are either over-treating black patients or under-treating white patients -Differential rates of treatment do not necessarily imply harmful discrimination

Limited competition

-Characterizes a media market with many producers and sellers but only a few products within a particular category -Not just due to barriers to entry, but also because of insurance, non-transparent prices, moral hazard for insured patients, government set prices, and the emergency nature of health care means that patients are unable to search for the best and cheapest hospital

Internal rate of return (IRR)

-Consider two possible career choices P and C with incomes paths Ip and Ic -Internal rate of return r* is the discount rate which equalizes the NPV of both careers (or the difference between NPV (p) − NPV (c) = 0 -Formula: summation from 0 to T of [(Ip(t)-Is(t))\(1+r*)^t = 0 -Someone with a discount rate of r* values career P and career C exactly equally -if the internal rate of return increases more people will be willing to become doctors -If the internal rate of return is higher than an individuals' discount rate, he should make the investment --> future returns are worth the wait -If the internal rate of return is lower than an individual's discount rate, he should not make the investment --> future returns are not worth the wait -Internal rates of return for a professional career are substantially higher than the market interest rate -IRR in medicine is typically between 11% and 14% (also significantly higher than market interest rate) --> the market not adjusting (supply isn't increasing) due to the barriers to entry -The fact that the IRR has stayed high is curious, suggests that being a physician is highly lucrative as well as that there are barriers to entry (hence why more physicians have not entered the market) --Supply of physicians may be constrained by the number of slots in medical schools and restrictions in the immigration of physicians

Evidence for downward sloping demand curve for health care

-Consumers are price sensitive: people with different budget constraints, life expectancies, qualities of life evaluate the tradeoff between medical care and other goods differently -As Patient cost-sharing increases, the number of episodes of outpatient care decreases sharply --Rand HIE shows that people assigned to the 95% group: had 34% fewer episodes of chronic outpatient care than those in the free plan and 37% fewer episodes of acute outpatient care --Oregon Medicaid experiment (Lottery winners vs. lottery losers who were unlikely to be on medicaid): lottery winners were 24 percentage points more likely to have an outpatient visit over a 6 month period and 35% more visits on average

Inpatient and emergency room care

-Data suggest people are more likely to go to the emergency room when they know they are covered and will not have to pay any additional prices (they will go even when things are not that urgent) --Even though one would think that if it was truly an emergency people would not be deterred from visiting the emergency room regardless of coverage -Based on age - looking at the discontinuity that occurs at age 65: once people become eligible for healthcare at age 65 there is an increase in both unplanned ED admissions and other admissions, but the increase is greater for other admissions --This is consistent with the other results we have seen, when they condition is more severe people will be less sensitive to price --Also, this data and the results can only be used to generalize across this age group

Induced innovation

-Def: discoveries that result when innovators change their research agenda in response to profit opportunities -Ex. changing demographics --> As the US population aged between 1970-2000, drug companies turned their attention to drugs for the elderly (glaucoma medication, etc) -Academic and public institutions also engage in _______ _______, even though they do not usually profit directly from their discoveries(ex. large-scale production of penicillin during WWII by US Department of Agriculture, ex. In recent years, academic researchers have focused more on obesity, ex. US Army ceasing malaria research after Vietnam War ended and troops came home from malarial regions)

Private vs. public markets for insurance

-Private markets can navigate the tradeoff between moral hazard and uninsurance to provide the optimal contract if they are perfectly competitive -Because public insurance is necessarily non-voluntary, government-provided insurance lacks self-corrective features that make national insurance policy both more complex and more critical

Adverse selection death spiral

-Def: successive rounds of adverse selection that destroy an insurance market -There is nothing to stop this cycle -The heart of the problem is adverse selection: only the worst customers stay in the market when the insurer sets the premium -No way for the insurer to turn a profit in this very simple model -Both frail and robust individuals are pooled together, premium is average cost of people in the pool, frail types are indirectly subsidized by robust, robust types exit the pool, leaving only unhealthy individuals, then the cycle repeats (readjusted premium, healthier types leave)

Pay for performance: a dangerous health policy (Research insight # 4)

-Def: the catchall term for policies that purport to pay doctors and hospitals based on quality and cost measures -Has been taking a bashing -Last November, University of Pittsburgh and Harvard researchers published a major study in Annuals of Internal Medicine showing that a Medicare pay-for-performance program did not improve quality or reduce cost --> actually penalized doctors for caring for the poorest and sickest patients because their quality scores suffered -Empirical evidence suggests that pay for performance policies do not improve the health of patients, harm sicker and poorer patients, encourage doctors and hospitals to avoid sicker patients who drag down quality scores due to factors outside physicians control, cause some doctors to stop using lifesaving treatments if they don't result in bonuses, create interruptions in needed medical care, reduce job satisfaction and undermine altruism and professionalism among doctors, and cause doctors to game/lie about quality measures (ex. Medicare program that punished hospitals for hospital-acquired infections actually induced some hospitals to characterize infections acquired after admission as present upon admission or to simply not report the infection rather than reduce actual infection rates) -Judging performance based on physician skill is not a fair measure as many other things can impact performance (ex. patient genetics), and only causes harm to patients either through doctors treating a disproportionate number of sick/poor people --> doctors get penalties --> leaves less resources to treat others or sick/poor patients drag down doctor scores so they avoid treating them --> leads to serious preventable illness and increased medical costs -Overall, doesn't work: practices that care for lower income or sicker patients received greater penalties, essentially creating a reverse Robin Hood effect that may have exacerbated existing disparities in care --May improve the sales of products like dishwashers and computer products, but it is irrelevant to the complexities and professionalism of good doctoring and other human services like education

Why buy insurance?

-Demand for insurance is driven by the fear of the unknown, there's a possibility of bad outcomes, so get this to hedge against risk -Purchasing insurance means forfeiting income in good times to get money in bad times (if bad times end up being avoided, then the money is lost) --Ex. the individual who buys health insurance but never visits the hospital might have been better off spending that income else

Full-insurance line

-Describes when we are fully insured (income will be the same no matter if you are healthy or sick) -45 degree line -Represents IH equaling IS, so any contracts that are on this line are full contracts

Health disparities

-Differences in health outcomes among groups -Caused by education, race, employment grade, income, birth weight, social groups -Different causal pathways (relationship b/t health, SES, and other variables) underlie the existence of ________ -Ex. males who are college graduates are more likely to survive at eighteen than HS dropouts (education) -Ex. those who were British royalty had a higher life expectancy at birth than commoners starting in the 1740s (income or employment grade), whereas from the 1500s until then they were similar, but by the 1900s there was a 20 year difference -Ex. high-income individuals routinely self-report better health status on this scale (1-5) than low-income individuals and are less likely to have diseases (health can also be a powerful predator of future wealth) -Health disparities between social groups have been found even in non-human societies: ex. baboons in Africa --Baboons at the top of their social hierarchies are in better health than subordinated baboons, dominant baboons have higher levels of high-density lipoprotein (HDL levels) --> good cholesterol, Sapolsky has argued that this disparity arises because dominant baboons suffer lower stress -Ex. whites report better health than blacks

The Grossman model and health disparities

-Different SES groups may have different MECs (Marginal efficiency of health capital curve) --MEC indicates the return on each additional unit of health capital --Due to the 8 different hypotheses for health disparities --Differences in the parameters (health productivity, resources constraints, health depreciation rare, total productive time, and rate of time discounting) can result in the different MEC curves, due to different optimal health decisions

Who is harmed by induced innovation

-Diseases that are rare (orphan diseases) or that mostly occur in developing countries (tropical diseases) receive less attention from researchers, because there is less profit to be made -Governments have tried to harness the power of induced innovation to fight these diseases -Orphan Drug Act in the US: facilitate development of orphan drugs for rare diseases such as Huntington's disease, myoclonus, ALS, Tourette syndrome and muscular dystrophy which affect small numbers of individuals residing in the United States -Advanced purchase of yet-undiscovered vaccines for HIV, malaria, TB

Liability insurance

-Doctors can respond to liability risk through the purchase of __________ -If ________ were full and could completely insure physicians from the costs of malpractice claims there would be little incentive for doctors to practice defensive medicine -But, full _________ might induce doctors to be less diligent in protecting against errors, because the risk of malpractice now is shouldered entirely by their insurer --If so, ________ may lead to more medical errors, more malpractice claims, and more malpractice payment --Which is why insurers are reluctant to offer full malpractice insurance packages

How strong should patents be?

-Downside of stronger patents: customers have to pay monopoly prices for a longer period, less incentive for further innovation by same company, and act as legal barriers to subsequent innovation by another company -But if patents are too weak, no incentive to develop new drugs!

Indifference curves in IH-IS space

-Downward sloping: willing to give up income in one state if compensated for more income in the other state -Convex: more downward sloping at low levels of IH but flatter at high levels, IH and IS are imperfect substitutes, result of risk aversion

Drug safety

-Ensured by FDA regulation -Before 1930s, no safety or testing regulations for drugs in the US magic elixirs (ex. Thalidomide in Europe prescribed to pregnant women caused birth defects --> pulled from shelves and became more heavily regulated) -Kefauver-Harris Amendment in the US (1962): companies must prove new drugs are safe and efficacious through clinical trials, stricter regulations led to a lower number of new chemical entities on the market -Doctors have prescription power -Bans on direct-to-consumer (DTC) advertising: bans in place in most developed countries except US, prevent moral hazard and reduce strain on doctor-patient relationships, but customers may not find out about new drugs that will benefit them

Cost shifting

-Even for care that is uncompensated hospitals may engage in ________ by using richer recipients to subsidize poorer recipients of care (also called cross-subsidization) -Def: redistribution of payment sources, typically occurs when a discount on provider services is obtained by one payer and the providers increase costs to another payer to make up the difference -Hard to measure -Hospitals use profits from lucrative services to subsidize uncompensated care -In United Kingdom cost-shifting occurs well before the hospital visit, in the form of tax revenue, which in progressive systems the tax burden is more so on the wealthy

Disparities in countries with universal health insurance

-Evidence shows health disparities persist even in countries with national health insurance -Ex. infants born to poor families in Canada are nearly twice as likely to have poor health than children born to rich families -There are also substantial differences in health outcomes between racial groups

Differentiated product oligopoly

-Ex. hospital industry -Strict barriers to entry (buildings, technology, staff, administration, etc.) -Entry regulated by the government to ensure the level of quality (level of reg. varies b/t countries) -Few firms in market -Services provided by each firm are not perfect substitutes (differentiated products): hospitals provide different quality of care depending which hospital's staff and doctors are more experienced, distance and travel time can also differentiate the service of each hospital (especially in an emergency)

Bhattacharya (2005) - explaining differences in returns

-Examines data on young physicians and uses an IRR calculation to find that differences in work-hours, training requirements, career length and ability (measures by test scores) can only explain half of the salary premiums for medical specialists -Differences in ability at the end of medical school do not explain any of the differences in returns between high-income and low-income specialties -Remaining differences must be due to barriers to entry to becoming a specialist

Rand Health insurance experiment (1974-1982)

-Example of a randomized experiment: assign treatment randomly to different groups, include control group, this ensures groups are statistically similar -Results of nonrandomized studies until this were mixed as to whether demand was downward sloping, but this confirmed that it was -Randomly assigned families to several different insurance plans for several years --> plans varied on how much they covered, had different copay rates (different price for same service) of 0, 25%, 50%, and 90% copay rate -Results: saw people were more likely to go to the doctor for acute problems, rather than chronic and the more coverage people had the more they went to the doctor

Physician induced demand (PID)

-Extra demand for medical goods and services induced by the advice of a physician who takes into account goals other than the patient's objectives (e.g., the physician's own financial gain) -Due to information asymmetry between doctor and patient, financial incentive for doctors to prescribe more services than needed, when reimbursement rates for various procedures change, doctors prescription practices also change -Ex. Neurologists and orthopedists owning their MRI machines ordered more MRI tests than did physicians who had to refer patients to outside firms for MRI scans -Similar findings among physicians owning specialty treatment clinics: physicians who recently became owners of back and spine clinics were more likely to recommend surgery than they were before owning their own clinics and Yip (1998) finds that thoracic (spine) surgeons compensated for a round of cuts to Medicare fees in 1990 by performing more surgeries -Tradeoffs: high reimbursement may lead to overprescription of certain procedures vs. low reimbursement might cause physicians to switch patients to other more lucrative services (late 1980s surgical fees were lowered by 6.5% to account for anticipated PID, but saw no volume change, may imply PID is small or doctors were inducing demand for more lucrative services)

The upside of moral hazard

-Extra preventative care: supported by RAND HIE and Oregon Medicaid Experiment, this is a beneficial effect if and only if people are consuming less preventative care than they should -The income effect: Insurance makes people richer by making expensive surgeries or treatments affordable when they may have been unaffordable without insurance

Life cycle of a drug

-Find chemical compound that might treat a disease -Then, test it on animals to show it is not toxic -Then, test on humans in three phases --Phase 1: low dose to healthy individuals ( 2 years) --Phase 2: dose to unhealthy individuals ( 2 years) --Phase 3: test effectiveness in preventing disease or medical conditions( 3-4 years) -Get approved for sale by FDA or similar body -Once the drug is approved for sale, the drug company has a temporary legal monopoly protected by a patent (17 years in the US) --This is the company's chance to recoup the millions of dollars spent on testing --After that time is up, other companies can produce the same drug cheaply and profits decrease sharply

Elasticity of demand

-Formula: % change in quantity demanded / % change in price --> [(Q1-Q2)/Q1]/[(P2-P1)/P1] -Can be used to compare downward sloping demand for different types of healthcare or for the same type in different studies --Ex. Suppose an individual starts with an insurance plan with a 25% copayment rate and switches to a plan with a 95% copayment rate... -Demand is relatively inelastic if it is (-1 < E < 0), ex. health services

How would you test if demand is downward sloping

-Gather data on the quantity of healthcare services (ex. Number of visits) and the price of healthcare (ex. Copay, insurance premium) -Seems generally that those with insurance are healthier than those without, but hard to conclude the direct relationship of health insurance on health --Money/income could be a potential confound - people with more money can not only afford better insurance, but they can also afford better food, a gym membership, etc. -Control for confounds by either using matched pairs or a longitudinal study --Still not perfect because you can only control for things you know and many other things could have an impact, for example risk aversion

Grossman Model

-Provides a framework that helps us think about health not just as something that happens to people, but as something that people partially determine with their choices -In the ____ _____ health plays three roles: (1) consumption good, (2) input into production, (3) a form of stock/capital (an investment) -Explains two well known empirical health phenomena: better health among the educated and declining health among the aging

Alternative explanation for positive correlation between competition and costs

-Goes against the medical arms race view -Researched by Dranove and Satterwhaite (1992) -More hospitals within a region can arise as a response to greater demand there -Higher demand alone can produce a positive correlation -Hospital in smaller markets face less demand and therefore less return to installing MRI machines, angioplasty facilities, and other technology investment, so these hospitals are unlikely to adopt the new technologies as quickly -A positive competition-cost correlation could be the result of differing regional demand and not a medical arms race -Test their hypothesis by applying an empirical strategy that controls for traits of different regional market and find little support for the medical arms race hypothesis (even before 1983)

How to solve adverse selection

-Gov forces everyone to buy contract at the intersection of the full-insurance line and the population zero-profit line -Could be free market solutions, under one crucial assumption: health differences between robust and frail only appear over time as customers age (robust when young, frail when older) -Lifetime insurance contract --Two customers pool together at age 18 and make a lifelong, contractually-binding commitment --> by age 50, one is robust and one is frail (no one knows in advance who will be there at one point) --This solves adverse selection but creates antagonistic relationships and may be legally unenforceable -Guaranteed renewable contract --Premiums are frontloaded so that robust and frail both want to remain in the contract voluntarily --But still a problem: customers can't switch insurers so there is no competitive pressure -Cochrane lifetime contract: Insurers also provide premium insurance for the future --Ex. If someone is diagnosed with cancer, they get reimbursed for immediate care expenses as well as a windfall payment to afford future health insurance premiums that will be very high --Allows the market to be more mobile and competitive because thanks to the windfall payment sick customers will not be tied to their current insurers only because they cannot afford to go elsewhere

Single-period utility model

-Health enters the model as a consumption good: Ut=U(Ht,Zt) --Ht is the level of health --Zt is a composite good representing everything else -Choices between H and Z may pose interesting tradeoffs -Time constraints within a single period: can either have time sick (doesn't increase utility) or time working (produces income), where time sick is entirely determined by health and is not voluntary

Why educated people have better health explained by the Grossman model

-Health is positively correlated with socioeconomic status (SES) -The Grossman model explanation for positive SES gradient in health says that the well-educated individuals are more efficient producers of health --So, for any given hour dedicated to health recovery or dollars spent on health-related goods)(M) a college graduate reaps more health improvement than a high-school dropouts they have a better understanding of doctor's instruction and more sophistication about purchasing medicine -The more educated person gains more health stock for each unit of health investment (more efficient at each level of health investment)

Empirical evidence that when reimbursement rates for various procedures change, doctors prescription practices also change

-Hickson et al. (1987) randomly assigned a group of pediatric residents to be either paid by a fixed salary or paid according to the number of services rendered -Fee-for-service physicians scheduled more visits than salaried physicians and scheduled more visits than recommended by the American Academy of Pediatrics --Allows them to increase the amount of money they will make

Medical Arms Race Hypothesis

-Hospitals compete on quality by adopting the best medical technologies available to appeal to physicians and their patients, this results in a race for each hospital to have the best medical technologies available and may cause overconsumption of medical technology -Most studies find a positive correlation between competition in an area and the cost per admission as proxy evidence for a ________ (ex. a hospital competing with many other hospitals in the area would invest more in medical resources and also generate higher expenditures per patient) -Robinson and Luft (1985, 1988, and Luft et al. 1987) find that: --US hospitals in more competitive markets have higher expenditures --Hospitals tend to have higher ratios of employees to patients --More likely to invest in facilities hence the ability to perform angioplasty and coronary bypass facilities and more mammography, heart surgery, catheterization services, offer them even if the demand for is not that high it just shows they provide the "best" care

Do physicians work too many hours?

-Hours: over 60 per week, on call residents could work up to 30 consecutive hours -In 2003, implementation to limit number of hours/week for US doctors: no more than 80 hours a week --> saw no change in patient mortality and many residents still work over 80 hours a week, but report only 80 hours -Trade-off between longer and shorter work hours: --Long: fatigue may impair physicians cognitive abilities and in turn may affect patient health --Short: requires more hand-offs by physicians and thus greater chance for error --Do not know which effect dominates -Fatigue and cognitive abilities: saw surgeons with no sleep needed 14% more time and committed 20% more errors, but if they work longer hours error can be reduced by eliminating the need to change doctors

Herfindahl-Hirschman Index (HHI)

-How economists determine whether a market is an oligopoly or not -An index of market concentration found by summing the square of percentage shares of firms in the market -If HHI closer to 1 means few firms in the market (highly concentrated) -If HHI closer to 0 means a large number of firms in the market -Us Department of Justice and Federal Trade Commission share the responsibility of regulating competition in markets labels concentration: --Concentrated: HHI > 0.18 --Highly concentrated: HHI > 0.25 --HHI in hospital markets in the average American metropolitan area was 0.33 in 2006 (so concentrated market with low competition)

Partial insurance

-IS′ < IH′--> IS-r+q < IH-r --> IS+q < IH --> q<IH-IS -Doesn't achieve state-independence: the closer q is to IH − IS and the farther from 0, the fuller the contract -Fullness of the contract determines the size of the payout (q)

Returns to specialization

-Pursuing a highly paid specialty field is almost always worthwhile, even considering the longer low-paid residency -Nicholson (2002): Internal rates of return for physicians entering into radiology, orthopedic surgery, general surgery, obstetrics, gynecology, and anesthesiology in 1998 exceeded 25%, so even if residency positions did not pay, it would be still worth the wait -High salary differential in competitive equilibrium explained by specialists worjing more, training longer, retiring eralier, and being more highly skilled

Shift to outpatient care

-In 1974, the number of hospitals peaked and has been declining ever since and they no longer play a dominant role in providing medical care -Why?: technology advances have reduced recovery times (ex. laparoscopic surgery: surgeons rely on a small camera rather than a gaping incision to see into the patient's body, reduces recovery time) and because insurers increasingly design hospital payment to incentivize shorter hospital stays (changed from fee for service (FFS) to diagnosis-related groups (DRG) (Medicare in 1984)) --FFS: paid based on numbers of days in hospital and for each service performed (incentivized doctors to keep patients in the hospital way longer by providing as much care as they can) --DRG: payment based on patient's diagnosis at the time of hospital admission (since the diagnosis is given at admission along with cost, so doctors no longer have an incentive to keep patients in the hospital as any extra procedures will cause the hospital to lose money) -Despite the shift we saw occur in the US, in Japan, hospitals are still paid on a FFS basis (though recent reforms have introduced a partial and voluntary system of prospective payment) --Average length of stay declined in past decades (result of advances in medical technology), but it is still substantially higher than the average length of stay in the US as hospitals have little incentive to avoid lengthy hospital stays

Price competition

-In most markets, suppliers attempt to attract customers by offering lower prices than their competitors --keeps markets efficient, forcing firms to offer their clients a good value or face the risk of bankruptcy --But when limited, this check on inefficiently high prices may disappear -Insurance can interfere with price competition and may eliminate it completely --Insured patients have diminished incentive to search for the lowest-cost provider (moral hazard) --> why would they if they aren't paying for it --Search costs are high and price information can be hard to come by (ex. patient may be rushed to nearest ER without concern for price. the full price of treatment is in most cases uncertain (depends on diagnosis, complications etc.)) --In many health care systems, medical prices for group of treatments are defined by the government (Medicaid and Medicare do it for their enrollees; UK's NHS determines the payment for hospital care) ---Thus, hospital cannot lower prices to attract more patients, Instead insurance companies have the most bargaining power

Hosiptal bills

-In nationalized systems such as the UK and Sweden, the government decides how to fund hospitals -In systems with private parties payers typically negotiate with hospitals to decide the bill -According to public price lists or charge-masters, the cost of a chest x-ray in 2004 ranged between $120 and $1,519 across seven California hospitals (tremendous variability!!), so price depends on the bargaining power of hospital and insurer, but in actuality, buyers (both insurers and patients) rarely pay the charge-master price -Instead, hospitals and insurers (both private and public) periodically negotiate rates -Rates vary with relative bargaining power of hospital and insurer -The same hospital may receive different rates from different insurers

What may have shifted extent of the medical arms race over time

-Increased popularity of managed care organizations (MCOs) during the 1980s in the US might have tampered down the medical arms race -MCOs vertically integrate the payer and provider of health care, they are sensitive to both hospital costs and patient bills -This sensitivity may limit any medical arms race if hospitals are more reluctant to invest in expensive technologies -Zwanziger and Melnick (1988): positive correlation between competition and costs found in 1983 disappeared by 1985 (so less evidence than originally thought) -Conor et al. (1997) find that the correlation for hospitals nationwide greatly subsided by 1994 -But it seems that rise of outpatient clinics that also compete with hospitals led to a medical arms race again

Moral hazard with health insurance

-Insured people take risks with their health that similar uninsured people would not take, and demand more expensive treatment from their doctors when they get sick -This is the downside of health insurance because it raises society's level of health care expenditures

Currie and Schwandt, PNAS 2013 (research insight #3)

-Large amount of literature on the relationship between month of birth, birth weight, and gestation, but recent work cast doubt on causal interpretation --Mothers with low SES give birth in months that are associated with poorer birth outcomes -Within-mother analysis of seasonal patterns in health at birth analyzed the seasonality at birth using administrative records on 1,435,213 children and by following the same mother over time, they eliminate differences in fixed maternal characteristics as an explanation for seasonal differences in health -Figure A, B, C, D - suggests that being born in a cold month can be really bad, especially if you are exposed to germs -Birth weight also has similarities based on when you are born -They find sharp declines in gestation length among babies conceived in May: pre-maturity goes up by 10% -Likely explanation: higher influenza prevalence in Jan and Feb 2009 (H1N1 pandemic-began and peaked earlier and involved many more cases) -Policy Implications: --Seasonal variations in nutrition matter for birth outcomes --Flu shots might be effective in fighting seasonal deterioration in length of gestation --Further research is needed to evaluate the effectiveness of specific interventions

Hospital competition

-Like suppliers in other industries, hospitals compete with one another for customers (patients), but there are some differences with respect to other markets (not in the US): --In many countries, access to health care is a basic right --Hospitals are forbidden from denying anyone care --Building and staffing a new hospital requires high initial investment cost and often some form of government approval --High barriers to entry can impede competitors from entering the market --> no free entry (ex. high costs, attracting doctors, bring admin, etc.) --Commonness of health insurance in the market distorts the typical economics of supply and demand (demand because there is less of a cost burden on patients when they are insured, so they may demand more care than if they had to pay for it all on their own)

Accreditation Council for Graduate Medical Education (AGCME 2003)

-Limits on the number of hours that residents in the US are permitted to work -Doctors in training were not to work in excess of 24 consecutive hours and no more than 80 hours in a single week -Acted a natural experiment to estimate the impact on patient outcomes of limiting resident work-hours, but only teaching hospitals with resident programs were affected by the new limits --> studies found little to no difference in mortality outcomes in the 2 years after the reform --Possible explanation: not completely effective policy (84% of survey respondents violated regulation at least once)

Barriers to entry

-May explain the high IRR -For example, in 19th century, becoming a doctor was simple, anyone could do it since there was no regulation about training, bur nowadays the market for physicians are tightly regulated -Regulated by restrictions to enter medical school and practice medicine -These barriers to entry allow physicians to earn monopoly rents (extra wages above the competitive level that accrue to producers in a market where supply is artificially constrained) -American Medical Association (1847) created pre-reqs for medical school, made med school 4 years, required doctors to have a license to practice -1910 Flexner Report helped shut down low-quality med schools -Result: less med schools and less med students

Full insurance

-Means no income uncertainty (IH′ = IS′) --Regardless of healthy or sick, final income is the same -Risk-averse individuals prefer full insurance to partial insurance (given the same price) -Payout: given IH′ = IS′ --> IH+0-r = IS+q-r --> IH = IS+q --> q = IH-IS (the payout from a ___________ contract is the difference between incomes without insurance)

Discount factor

-Measure of how much less an individual values future income over present income -Lies between 0 and 1 -Formula: δ = 1/(1+r) --Where r is the discount rate, analogous to the market interest rate that would make a person with discount factor δ indifferent between saving for tomorrow and spending today -Ex: δ = 0.90 corresponds with r = 0.11 -Those who are very patient have high discount factors and low discount rates r

True vs. measured data

-Measured demand will be biased -When prices, are low measured demand is overestimated, meaning it over selected those who prefer low prices (opposite can be said for high prices, it is underestimated)

Relationship between hospitals and physicians

-Modern hospitals combine the great expertise of medical professionals with expensive technologies -Running a large hospital is a coordination challenge that requires extensive manpower for tasks such as managing the operating room, conducting blood tests, X-rays, and dispensing drugs -People tend to think of "good hospitals" as having good doctors, but a good hospital is not necessarily the one with the best doctors, but the one whose doctors and support staff work together best -Harris argues that hospitals should be thought of as two separate economic entities: the physician staff and the administrative staff --Doctors often not direct employees of the hospitals, which can undermine their incentive to control hospital costs (often they're an individual who basically rents the space in the hospital buildings) --Administrative staff does not interfere with doctors' treatment of patients, they control budget, but lack the medical knowledge to judge the relative value of different innovations for patient health --This bargaining process between doctors and administrators can result in wasteful spending to install cost-effective technology --> this allocation of decision power contributes to explain why hospital costs have grown so much in past decades

Oregon medicaid experiment

-No assignment to different insurance plans -Compared two groups of low-income adults, those who won the lottery to receive insurance and those who lost and received no insurance -Results: those who won the lottery and received medicaid went to the doctor more often than those who lost and received no coverage

Recent trends in US hospitals

-Number of hospitals has stayed relatively flat with a drop off starting in the 90s -Number of beds has been decreasing even more so than hospitals since 1967 -Number of outpatient visits has overall been increasing with the only major decrease occurring in 1987 -Length of stay in hospitals has been decreasing due to more efficient treatment thanks to improved technology and the shift to outpatient care -Despite all technological advancement of the last century hospitals remain dangerous places: still places where sick people are housed in close quarters and spread diseases to one another, breeding grounds for drug-resistant bacteria that threaten the health of the elderly and the immuno-compromised (hence why the trend towards outpatient care is welcome)

Measuring price sensitivity with elasticity

-One simple measure of price sensitivity for each type (inpatient/outpatient) of care is the slope of the line plotted between two points of the demand curve, but can't compare points of demand for different services, so need a measure of price sensitivity that is no affected by the units in which price or quantity are measured -This measure is the elasticity of demand

Does the price of health care affect health?

-Oregon Medicaid Experiment tracked participant survival rates --No evidence of significant differences between the survival of Medicaid lottery winners and lottery losers in the first year after the lottery -Meyer and Wherry (2012) find that Medicaid increased survival rates of black children by between 13% and 18% -RAND HIE also failed to find a mortality difference between treatment groups --Yet, free insurance did seem to have an impact on the mortality rates among the most vulnerable participants --The high-risk people on the free plan were 10% less likely to die than high-risk participants on the cost-sharing plans -Out of 23 health comparisons, only 3 resulted statistically significant (blood pressure, myopia and presbyopia)

Elasticity in terms of healthcare

-Outpatient visits will be more elastic than inpatient since inpatient is more severe -Infant healthcare is less elastic than dental care because people will do whatever they can to help their child -When the price elasticity of demand for healthcare is zero, behavior does not change when you are insured, so moral hazard would not occur (moral hazard only occurs if a price distortion induces a change in behavior)

Insurer profits

-Premium: r -Payout: q -Probability of sickness p E[Π] = Expected profits --> E [Π(p, q, r )] = (1 − p)r + p(r − q) = r − pq -In a perfectly competitive insurance market, profits will equal zero: E[Π(p,q,r,)] > 0 → r = pq -An insurance contract which yields positive profits is called unfair insurance: E[Π(p,q,r,)] > 0 → r > pq (insurer would never offer a contract with negative profits)

Price controls

-Price ceilings set or negotiated by the government (ex. Italian government publishes list of maximum permissible prices for each drug, ex. NHS in UK sets the price at which they are willing to purchase drugs (monopsony power)) -Reduce incentive for research: tradeoff between access to existing drugs and incentives to develop new ones, same as patent tradeoff -US has no broad price controls, but other countries free-ride on US market for future innovation and drug companies count on making their money off US consumers

Moral hazard and risk reduction

-Q: Given the social loss associated with moral hazard, should insurance policies that create moral hazard be prohibited? --Ex. pollution is generally regarded as an externality that needs to be taxed or otherwise regulated - Should the provision of insurance be similarly taxed or even banned? -A: No, because insurance itself provides positive welfare gains that may offset the harm from moral hazard --Asymmetric information creates a tradeoff between more insurance coverage, which generates more moral hazard, and less insurance coverage, which increases risk exposure

Feasible contract wedge

-R1 = over full insurance: get more income if you are sick than when you are healthy (implausible contract), no reason you would want this -R2 = under indifference curve going through E: individual prefers E (no insurance) to any contract offered in this region, would not want any of these (why pay for something that gets you the same as E, which costs nothing) -R3 = northeast of zero-profit line: companies will lose money on these contracts, since above zero-profit line -F = _______ _______ _______: only area where both customers and insurance companies want to meet -Curved line: indifference curve of the typical individual going through E

How Insurance providers react to moral hazard

-Raising premiums for fuller coverage -Since, in a competitive market, premiums are a function of the probability of sickness, we can re-label the y-axis in the previous graph (prob of sickness) to the premium per unit of coverage -The upward-sloping curve represents the set of contracts that an insurer may offer to a person, who, without insurance, would have probability p0 of sickness

Positive correlation between risk and coverage

-Recall the separating equilibrium: high-risk types have full insurance (1), low-risk types have incomplete insurance (2)

Zero-profit line

-Represents the set of contracts such that the premium is exactly the same as the expected payout (no profits for insurance company) -Runs through the endowment point, E -Also can be thought of as the actuarially fair line, so contracts on this line will be actuarially fair -Slope depends on the purchaser's probability of sickness, higher the prob, the flatter the line -Any contract below the line will result in profit for insurance companies, any contract above the line will result in loss of money for insurance companies --> so no company will offer contract above this line

Heterogeneous risk types

-Robust types: low probability, p, of getting sick -Frail types: high probability of getting sick -Slope of zero profit line depends on probability of sickness, so who has a steeper zero profit line? --The population zero-profit line will fall between the frail zero-profit line and the robust zero-profit line --Robust will have a steeper line (for the same increase in premium, the insurance company can afford giving them a higher payout since there is less of a probability of them actually getting sick) --Frail will have a flatter line (opposite reasoning of above, more likely to get sick so company cannot afford as high of a payout) -How do indifference curves vary for robust and frail individuals? --Robust will be steeper (p is low, so will want a larger increase in income when sick to make it worthwhile) --Frail will be relatively flat (p is high, so any small increase in income will be beneficial to them, care less about income when healthy)

Why health declines as you age explained by the Grossman model

-Says it is too costly to stay young forever since the depreciation rate (γ) increases with age, the cost of capital (r + γ) rises, therefore individuals need to invest more to maintain the same level of health over time, so at some point they can't keep up, hence health must decline over time -The increasing depreciation (γ) makes it less and less attractive to invest scarce resources in health --> leads to a "why bother" attitude

History of Hospitals

-Scientific progress transformed the hospital from a place to die into a vital input for improving the health of a population -Mid 20th century: growing concern that there was an insufficient number of hospitals -Rising cost of care put hospitals out of reach of the poor -1946: Hill-Burton Act increased the number of hospitals in the US (most important policy for hospitals at the time) -Congress gave money for building hospitals --Any hospital receiving money had to provide free/low cost care to the poor --Preference toward underserved areas -Result: more hospitals (+16%) and more hospital beds (+12%) --Federal funding and the rise in demand together inspired a spree of building hospitals around the country which continued more or less unabated until 1974

Costs of investing in health

-Suppose market investments pay an interest rate of r -Depreciation due to aging (γ) acts as a second type of cost of investing in health --To guarantee the same rate of return as the market investment opportunity, health must pay a return of at least r + γ --Think of r as the opportunity cost and γ as the deprecation --Health becomes more expensive when γ rises --To have the same health later, I need to buy more now --Depreciation acts like a tax on any health the individual produces -The MEC curve determines the optimal amount of health for an individual --At the optimal level of health, the marginal cost of investment (r + γ) balances the marginal benefit of health investment

Health as an investment good

-it is a form of human capital (like knowledge, or education) -As any form of capital... --It depreciates in value over time --You can invest in it, increasing its value in future periods -Characteristic of health as a capital good: at low levels of health, small investments have enormous returns to productive time (downward sloping marginal efficiency of health curve shows the diminishing marginal returns to health) --Ex. the same investment in health if you are very sick will give you a much larger return than if you are at a higher level of health to start

Rothschild-Stiglitz Model in the IH-IS space

-Take HE and SE points and bundle them to one point E (endowment point) in IH-IS space -Point E shows the income of an individual in both the healthy and the sick state (don't know which you will end up in) -Point C represents a partial insurance contract caused by a horizontal shift = premium (r) and vertical shift = payout if sick (q − r ) from insurance -From HE to HC you lose a small amount of utility, from SE to SC you will gain much more utility (when you are relatively poor, the same increase in money will provide more utility than it would to rich people in the event that you actually get sick) -in the IH-IS space (C1, C2, C3, C4) represent contracts --Individual prefers C2 to C1: receives increased income when healthy, same when sick --Prefers C1 to C3: sees increased income when sick, same income when healthy --Prefers C2 to C4: sees same income when healthy, increased income when sick --Cannot compare preference between C1 to C4 and cannot compare preferences to E, people may have different preferences (ex. someone who is more prone to illness may prefer C1 since that income is higher, which would benefit you if you are sick very often) --> need indifference curves to really show these preferences

Efficient discrimination

-Taste-based is always inefficient: some factors other than patient's well-being is affecting treatment decision -Statistical may be efficient: efficient if there's medical evidence to treat racial groups differently --Ex: optimal hypertension treatment is different for blacks than for whites: diuretics and angiotensin converting enzyme (ACE) inhibitors should be used as first-line treatment for black patients, where as beta blockers and calcium channel blockers are the preferred treatment for white patients

Moral hazard

-Tendency for insurance against loss to reduce any incentive to prevent or minimize the cost of loss (ex. the situation in which individuals use more services when insured than if paying out of pocket) -individuals behave differently after contract is signed than if no contract -Ex. carelessness and fraud that can lead to losses and are the result of decisions by humans

Pros and cons of experiments

-The Oregon experiment focused on a low-income population Rand HIE studied a nationally representative population --> so, the Oregan results could not be generalized to a population beyond low income people -Rand HIE used a direct randomization of insurance coverage -Oregon experiment only indirectly related to insurance coverage (lottery winners were only 25 percentage points more likely to be covered in the year following the lottery than the lottery losers were) -The Oregon Medicaid Experiment included an uninsured group that was in part randomly assigned -RAND HIE did not include any participants who were totally without insurance -Both experiments provide evidence of demand for health care being downward sloping

Expected value

-The _________ of a random variable X, E[X], is the sum of all the possible outcomes of X weighted by each outcomes probability -If the outcomes are x1, x2, ..., xn, and the probabilities for each outcome are p1, p2,..., pn respectively, then E[X] = p1x1 + p2x2 + pnxn --Ex. an individual does not know whether she will become sick, but she knows the probability of sickness is p between 0 and 1, prob of sickness is p, prob of staying healthy is 1 - p, IS= income if she does get sick (income lower than if you were healthy, due to less productivity or the money you pay to insurance), IH>IS= income if she remains healthy --> E[I] : E[I] = pIS + (1-p)IH

Arc elasticity

-The choice of a starting point makes a difference in the calculation of elasticity, so the alternative is to measure elasticity at the midpoint between two endpoints -Formula: E(arc) = [△Q/(Q1+Q2)]/[△P/(P1+P2)]

Expected utility

-The expected utility from a random payout X E[U(X)] is the sum of the utility from each of the possible outcomes, weighted by each outcome's probability -If the outcomes are x1,x2,...,xn, and the probabilities for each outcome are p1, p2, ..., pn respectively, then E[U(X)] = p1[U(x1)] + p2[U(x2)] +...+ pn[U(xn)]

How do health insurers limit moral hazard?

-The extent of moral hazard depends on both how sensitive demand is to price and the amount of price distortion caused by insurance -Insurers cannot alter customers' price sensitivity (which is a property of their demand functions), but they do have ways to reduce the price distortion due to insurance... *1. Coinsurance:* insurance provision in which enrollees pay a percentage of each medical bill, and the insurer covers the remaining portion *2. Copayments:* insurance provision in which enrollees pay only a fixed amount, called a copay -Coinsurance/copayments maintain positive marginal costs for the insured and limit insurance coverage so it is no longer full *3. Deductibles:* set minimal levels of expenses below which the insurer does not help reimburse medical expenses (ex. a person insured with a deductible of $1,000 pays for his first thousand dollars of health care expenditures out-of-pocket, insurance policy then helps pay for expenses beyond the thousandth dollar) -Limits/eliminates the moral hazard from insurance, however, if the deductible is low enough, moral hazard may still persist *4. Monitoring:* some insurance companies try to observe and guide the preventative measures their customers take, while others choose to supervise the medical care that customers receive (ex. employee incentive programs with payouts if you see a nutritionist, do yoga once a week, or get a fitness test)

What determines the level of price sensitivity

-The extent of price distortion is a function of the completeness of the insurance -The fuller the insurance, the greater the price distortion -The extent of price sensitivity depends mostly on the nature of the risk being insured, and how controllable it is --Consider Huntington's disease vs heart attacks

Adverse selection

-The oversupply of low-quality goods, products, or contracts that results when there is asymmetric information -So people will choose a good that may not actually be the best choice for them -Ex. insurance attracts patients who are likely to use services at higher rates than average; low risks tend to be underinsured or there may be no insurance market

Hospital competitions impact on patient outcomes

-The standard prediction that more competition will improve welfare in the face oligopolistic pricing may not hold in this market --Increased hospital competition may actually lower consumer welfare by worsening patient outcomes --Exact effect of competition depends on many variables in the market -Kessler and McClellan (2000) study the impact of hospital competition on all non-rural Medicare patients in the US treated for acute myocardial infarction (AMI) --> found that more competitive regions yielded better results -However Gowrisankaran and Town (2003) also study AMI patients on Medicare in LA --> found worse mortality rates in the more competitive parts of the country -Possible reason for the difference: nationwide data vs. local data -Even in markets with private insurance the effect of competition on patient outcomes is unclear (too complicated to determine for sure)

Risk aversion

-The tendency to prefer a sure gain of a moderate amount over a riskier outcome, even if the riskier outcome might have a higher expected payoff --What drives the demand for insurance -Can be modeled through utility from income U(I) --> utility will increase with income, U'(I) > 0 -Marginal utility for income is declining: U''(I) < 0 (means the higher the income gets the less utility we see with every dollar)

Physician agency

-The transfer of authority/decision-making power from the patient to her physician -Patients trust physicians to act as perfect agents for their health -Doctors foremost concern should be patients well-being, not their own financial status or reputation -Physicians face incentives that may cause them to deviate from perfect agency (the optimal course of action for patients) --Due to financial, legal, or personal reasons doctors may overprescribe or not treat certain patients (ex. what led to the Opioid epidemic), which can be enhanced by informational asymmetry (you as the patient do not know what the best treatment is, so doctor's can easily mislead you)

Moral hazard and asymmetric information

-There is price distortion in insurance markets because insurance companies cannot monitor everything patients do and price their actions accordingly -Ex. it would be impractical for an insurance company to count how many cheeseburgers its customers eat -Ex. insurers cannot tell if a patient really needed to see his doctor ten times in the last month

Expected utility and expected income

-There's a crucial distinction between expected utility (E[U(I)]) and utility from expected income (U(E[I])) -For risk-averse people, U(E[I]) > E[U(I)]: prefers the utility they would get from their expected income to the expected utility they would get from their actual (but uncertain) income

Type I and Type II errors

-Type I error = bad drug is approved (e.g. Vioxx) -Type II error = good drug is rejected or delayed (e.g. beta blockers) -More incentive to avoid type I errors because of media attention

New Year's Resolutions (research insight #1)

-Typically you will see people trying to get healthy at the start of the year (seen around other time periods too (ex.following birthdays or other holidays)) -Loss aversion and the scale: people tend to be more affected by a loss than by a gain --Ex. when losing weight people will be more upset by gaining 3 pounds than how happy they would be if they had lost 3 pounds - if someone was loss averse they would stop weighing themselves in order not to see the loss, thus losing the benefit of using it as a reminder to stay healthy --To correct for this a modified scale was created which showed categories rather than numbers, in research they saw that those using this scale were doing much better than those who weren't --Also is why too much information in some cases can actually be bad -How to keep resolutions: write down a plan (just making a plan leads to greater results), put something of value on the line, Bundle your temptations (ex. only allowed to watch a show while you are working out), seek social support

The ideal insurance contract

-U(A^F) > U(A^P) > U(A) -For anyone risk-averse, actuarially fair & full insurance (A^F) contract offers the most utility, hence it is called the ideal insurance contract

Formal treatment of the used car model

-U(sellers) = sum from j to n of Xj+M -U(buyers) = sum from j to n of (3/2)Xj+M --Buyers value cars 50% more than sellers --Xj = quality of the jth car owned (can be any number from 0 to 100) --M = utility from other goods -Buyers do not know the true quality of a particular car, but they do know the utility function of the sellers and know the distribution of cars available for sale -A seller will put a car on the market if selling it will increase his utility --> If a seller sells his car of quality X for P dollars, he loses X units of utility but gains P dollars --> hence he will only put car j on the market if P>Xj -Of the cars that remain, the average quality (E[X|P]) is only 1/2 P -Figuring out when buyers buy is trickier due to uncertainty, they must think in terms of expected utility --> buyers will buy if: (3/2)E(Xj) > P -Buyers do not like cars enough to buy a car of quality 1/2 P for a price of P, so no cars sell and the market unravels (note: if value is not 3/2 this can change, ex. 5/2)

Hospital competition in a public health system

-UK most hospitals are government-run and many government reforms have tried to inject competition into the market (ex. reforms in the 1990s permitted competitive bidding by hospitals for government contracts, others in 2000s permitted patients to choose which hospitals they visited for surgery and granted more autonomy to hospital administrators) -Gaynor et al. (2010) and Cooper et al. (2011) focus on the 2000s reforms --Saw greater patient choice in hospitals and more autonomy in hospital administration, even while health care prices remained centrally decided by the government --Mortality rates from AMI were lower in markets with more hospital competition on quality -Propper et al. (2008) finds that 1990 reforms that increased competition reduced waiting times, but raised mortality rates --Lack of information on quality prevented them from discriminating on that so they were left to look at other things such as waiting times, which clearly was not the best indicator --Learning by doing may explain why greater competition may hurt patient outcomes: a monopolistic hospital may charge high prices, which lowers patient welfare, but may gain experience through access to a large number of patients which benefits the local community and more than compensates for the high prices

Why do health disparities exist?

1. Early life events 2. Income levels 3. Stress of being poor 4. Work capacity 5. Impatience 6. Adherence to medical advice -Policy importance of understanding causes of disparities before addressing them

For-profit and nonprofit hospitals

-US has both, but majority are nonprofit (75%) -Benefits of nonprofit status: tax empty, donors also receive tax deduction -Costs of nonprofit status: cannot sell stock, cannot distribute profits to owners, restricted to certain charitable activities -Theories for their existence: 1. Government failure theory: politics ineffectively help people in need, so the government responds to this concern by providing subsidies for medical care, but those most interested in charity will not think this is enough, so nonprofits act as the governments vehicle to organize the labor and capital of those who demand more charity 2. Altruistic-motive theory: some entrepreneurs prefer altruism over profits, so they are willing to accept reduced profits in exchange for more output or higher quality output 3. Asymmetric information: non-profits exist because for-profit firms are less trustworthy in the performance of actions that are hard to observe so donors trust nonprofits with more money 4. Nonprofits are for profits in disguise: profits are distributed as higher wages or non-monetary benefits --> has mixed study results

Expected utility without insurance

-Uncertainty about which outcome will happen, though she knows the probability of becoming sick is p with expected utility of E[U(I)] = pU(IS)+(1−p)U(IH) -Consider a case where the person is sick with certainty (p = 1): E[U] = U(IS) equals the utility from certain income IS (Point S) -Consider case where person has no chance of becoming sick (p = 0): E[U] = U(IH) equals utility from certain income IH (Point H) -What if p lies between 0 and 1?: expected utility falls on a line segment between S and H lines --For p=0.25, persons expected income is: E[I] = 0.25IS + (1 - 0.25)IH

Returns to medical training

-Unlike most occupations, returns to medical training are very back-loaded (will only see the benefit much later on) -Medical school and residency expensive in direct costs and opportunity costs (ex. time) -So, those who choose being physician are patient enough to value future returns -Career path through come of surfers vs. doctors: --US: doctors experience a spike after specialization vs. surfers start to drop off as they get older because their bodies can no longer handle strenuous physical activity --Europe: doctors do not see as large of an increase after specialization vs. surfers start to drop off as they get older because their bodies can no longer handle strenuous physical activity (same as US)

Net present value

-Way of calculating value of all future streams of income (from todays perspective) -Formula: summation from 0 to T of δ^t(income of t) --δ^t: the discount factor, measures how much less an individual values future income over present income, lies between 0 and 1 (small if impatient (more likely to become a surfer), large if patient (more likely to become a physician)) -Those with high discount rate have high NPV from being a physician Those with low discount rate have low NPV (and maybe even negative NPV)

Elastic demand curve

-When a given percentage price change leads to a larger percentage change in quantity demanded -As price increases, the amount demanded decreases

inelastic demand curve

-When a given percentage price change leads to a smaller percentage change in quantity demanded -People will always demand the same amount regardless of price (ex. life saving medicine)

Uncompensated care

-When a patient lacks the resources and insurance to pay for care (for some uninsured patients, even a discounted bill may be too costly to bear) --If not paid, the case for these patients goes uncompensated --Studies indicate that 6-7% of hospital expenditures are uncompensated --Medicare estimates uncompensated care for the whole population reached 20.8 billion in 1999 -Who pays for _________? --In the US, Medicaid and state governments provide explicit subsidies for uncompensated care through Medicaid disproportionate share (DSH payments are given to hospitals that provide uncompensated care), these payments can serve as an incentive for hospitals to increase their indigent care (Duggan 2000) --Certain state governments also levy taxes on insurance payments to hospitals to fund a program dedicated to medical education and uncompensated care --Medicare and some local governments make payments that are ostensibly for other purposes but effectively help defray the cost of uncompensated care

Who is harmed in a market with imperfect information?

-When firms cannot tell frail or robust individuals apart, frail customers still receive full insurance at an actuarially fair price The robust receive inferior insurance contracts that are not as full as they like - the robust are quantity constrained

Quality competition

-When hospitals no longer have to compete on price, they will raise prices and just compete on _____ -Measured by comfort of beds, availability of advanced technologies, or effort/vigilance of the staff -If a hospital can get away with charging high prices, it can attract patients with lavish, high-quality care and then charge them or their insurers a fortune for their stay (Medical arms race hypothesis) -Medical arms race + physician induced demand can → high health care expenditures will rise without corresponding improvements in health outcomes

Symmetric information

-Where consumers and producers have access to the same information about a good or service in a market (crucial concept) -Ex. used car market where sellers advertise cars, and buyers can accurately assess the condition of each car for sale, some buyers will be willing to pay more for cars in good condition; others are happy to get a deal --Outcome: each car sells for a different price, depending on its quality -Pareto-improving transaction: a transaction that leaves all parties at least no worse off (ex. all the cars end up with the people who value them the most) -One goal of a market is to make sure all Pareto-improving transactions take place -In the market we have described, there is nothing to stop all Pareto-improving transactions from taking place

Oligopoly markets

-__________ firms are not price takers --Can price above marginal cost (where as in the free market the price is given by the forces of supply and demand), not complete control, but some --Earn positive profits --Have market power: downward sloping demand (if they increase price they will lose some customers, but not all), creates high prices for consumers, less productive output, and lower social welfare --It's not a monopoly, there is still some competition, thus their power is limited --Yet, the few firms on the market may collude and agree to maintain high prices together (forbidden by gov., but can use strategic pricing instead)

Conditions for moral hazard to occur

1. The cost of a risky or wasteful action to an individual is reduced, usually as a consequence of insurance 2. Asymmetric information prevents an insurer from adequately pricing the action 3. That individual responds to the price distortion by changing his behavior - either by taking more risks or demanding more covered goods and services

Heterogeneous risk preferences

Assuming homogenous risk types and _______ _______ _________: -Negative correlation between risk and coverage -Bulk discounts will exist to encourage people to buy more insurance -No death spiral with insurance pooling: people who are more risk averse will stay in the pool when premiums rise but will still be the same risk type to the insurer

Adverse selection in real markets

Asymmetric information models make three predictions about these markets 1. Positive correlation between risk and coverage 2. Bulk markups 3. Adverse selection death spiral

Evidence for racial discrimination

Audit study (Shulman et al. 1999) -Gave fictional patient histories (ex. chest pain - that may be an indicator of impending acute myocardial infarction) -Used black and white actors -Patients told doctors same script (the fake symptoms), background, and hand motions -Only difference was the race of patient/actor -Results: physicians less likely to recommend standard treatment if patient was black

The training of physicians

Entry into med school is competitive and selective worldwide -In the US, average 50% of applicants are accepted into at least one school Length of medical school varies across country -US & Canada applicants must first get a bachelor's degree -European applicants go directly from high school Medical school can be super-expensive -US: 140k to 225k for four years -European medical training often heavily subsidized Ratio of expected deaths for inpatient medication errors by month -The spike in expected number of events occurs in July because it is the month when new doctors are first entering residency, this translates to increased negative outcomes in treatment Residency: period of on-the-job training following medical school -In addition to classroom work, physicians-in-training must also gain hospital experience -New residents lack experience, and when new residents arrive at a hospital, empirical evidence shows that medical errors go up (ex. July effect for US, August killing season for Uk)

How moral hazard leads to social loss

Ex. consider an individual who loves cheeseburgers but is at risk for a heart attack -Without health insurance: his cost for each cheeseburger includes both the price of the burger and the increased chance of a heart attack -With health insurance: the cost of each cheeseburger declines, since the insurer picks up the costs of heart attack care --The change in price represents the price distortion --The angle between the DC and 90 straight line represents the extent of price sensitivity (the larger the angle, the more responsive behavior is to price distortions, and the larger the social loss) -In this case, social loss takes the form of extra money, labor, time, and effort that others expend on caring for heart attacks caused by cheeseburger overconsumption

Income with insurance

Let IH′ and IS′ be income with insurance -Sick: IS' = IS+q-r -Healthy: IH' = IS+0-r -Remember that risk-averse consumers want to avoid uncertainty so for them, optimally IH′ = IS′

Information asymmetry between doctor and patient

Patients cannot assess whether an extra test or procedure ordered by a doctor is necessary hence that when reimbursement rates for various procedures change, doctors prescription practices also change --> but the patient assumes the doctor would prescribe the best medication regardless of this, which may not be the actual case

Volume-outcome relationship and learning by doing

Positive correlation between number of cases and the outcomes of those cases -Learning by doing: where by doing procedures you become better at them (would support that people should go to hospitals with higher volumes) -Selective referral patterns: doctors may prefer to direct their patients to the most highly skilled surgeons which could detract from the correlation between volumes and outcomes -Evidence for both hypotheses (learn by doing and selective referral) -Strength of volume-outcomes relationship in hospitals varies dramatically across different procedure types Surgical mortality rates for various Medicare procedures, by hospital volume -See some evidence for volume being positively correlated with lowered mortality rates (looking across the volumes you can see the mortality rates getting lower)

Evidence on mortality rate

Rand experiment: -For all participants: people who have free healthcare have a slightly lower risk of dying, but overall there is no difference between insurance plans -For high risk participants: the difference between those who receive healthcare and those who have copays is greater, with copay people having a higher mortality rate Oregon experiment: -The mortality rate is the same between those who recieved/won medicaid and those who did not -When considering other factors much larger effects were seen in categories such as self-reported health good, self-reported health fair or better, health about the same or better over last 6 months, number of days of mental health good, and did not screen positive for depression in the last 2 weeks -The results in this experiment may seem stronger than the Rand experiment because people were taken from a lower socioeconomic sect of the population or due to a placebo effect that having insurance may just make them feel better (gives them peace of mind) -This evidence does not directly indicate that there is no benefit to one type of insurance plan as mortality may not be the best measure, since it is heavily impacted by things beside coverage, such as genetics

Hospital experience vs. physician experience

Should you prefer having your surgery with an experienced physician or in an experienced hospital? -McGrath et al. (2000) find hospitals with more surgical experience have fewer complications than a single physician with high experience -Finding makes sense because teams of medical workers can collaborate on surgeries, so individual physician experience less impactful as they are all on their own with no one to fall back on or ask for help

Types of racial discrimination

Taste-based -Preferential treatment for certain groups of patients -Can be conscious or unconscious Statistical -Stereotypes on biology or behavioral tendencies -May infer something about you not based on your prior visits but due to your belonging to a particular groups ^^Can be efficient or inefficient -Some discrimination may harm patients, but others may benefit them (inefficient) -Some populations more be more at risk for some diseases, so it could be helpful to look at that specifically based purely on race (efficient)

Defensive medicine

There are other reasons why doctors might deviate from optimal medical practice (besides financial incentive)... -Defensive medicine (not wanting to get sued) or the over utilization of testing services, which protects against malpractice lawsuits, doctors may also reject or refer away risk patients (or litigious ones - more likely to sue) to reduce malpractice risk -Doctors fearful of lawsuit may overprescribe (and overcharge) for only marginally useful procedures -Mello et al. (2010) estimate that medical liability system in the US costs $55.6 billion annually, the estimate includes costs of actual payouts from malpractice claims, attorney fees, higher medical costs due to defensive medicine, but oats social costs (damage to reputation from liability suits) as it is difficult to quantify -Jena et al. (2011): between 1991 and 2005, 7.4% of all US physicians covered by a large liability insurer faced at least one malpractice claim in any given year, with average successful claim of$274,887, and although only 22% of claims resulted in payouts, but even unsuccessful malpractice claims are costly (time, reputation etc.) -Kessler and McClellan (1996): 1980 reforms putting caps on the damages patients could recover through malpractice suits, so doctors could no longer be as fearful of being sued, lowered their risk, and the lwer liability pressure led to a 5-9% decline in medical expenditures for patients with serious diseases -Helland and Showalter (2006): reforms led to 10% decrease in expected liability costs Doctors increased work-hours by 2.85%

Patents in developing countries

Tradeoffs between patent strength are thought about differently in low income countries -Monopoly prices weigh more heavily on low-income populations -Free rider effect: if the US has patent protections, companies will develop new drugs even if there are weak patent protections in India Price discrimination -In theory, drug companies could sell their drugs for different prices in different countries -In practice, black-market importation makes this impossible

Is hospital competition good for patients?

Typically, competition improves quality and lowers prices -Overall, evidence for this is growing but requires an abundance of information to be available to patients as well as government regulation to ensure hospitals are not hiding important information But... -Commonness of insurance hinders price competition -Patients are typically referred to hospitals by physicians, so hospitals compete for physicians -Medical arms race hypothesis: greater competition among hospitals for physicians can result in redundancy in and over-consumption of medical technologies, this can actually increase costs without improving quality -Lots of empirical research about the effect of hospital competition on patient outcomes: mixed findings and different policy implications

Did UberX Reduce Ambulance Volume? (Research insight #2)

Uber and Ambulances -Emergency medical transport in an ambulance can easily exceed over a thousand dollars and usually this cost comes as a great surprise to the patient, as insurance only partially covers it or refuses to pay anything if they deem it wasn't necessary --They don't think about the cost when calling one, act out of pressure/fear, but once that wears off they are sucked by price -Unnecessary ambulance use is partly due to lack of other transportation, if other transportation existed they would use it --> many have now started to seek alternate, cheaper transport to the emergency room through ride-sharing services (ex. Uber/ Lyft) -Also, while ambulances typically go to the closest hospital, ride-sharing services allow the patient to choose which hospital they go to (potentially important factor as these facilities may have differing results for the same illness, with higher-cost hospitals associated with better outcomes) -The introduction of the Affordable Care Act (ACA) has slowed ambulance response times by 19% Uber -Expanded into the ride-sharing industry, with UberX in 2012 -By May 2017, Uber reached 5 billion rides, were in 76 countries and over 450 cities -Market effects due to Ubers entry: reduced number of drunk driving casualties, reduced citations, and reduced fatal accidents (at the county level, but not in metropolitan areas) Ambulance data -Ambulance response volume: the number of instances in which an ambulance is used for emergency transport, as submitted by the member states and sorted by zip code -Ambulance rate: city ambulance volume multiplied by one thousand and divided by city population -For some areas of the country Uber has been around for many years, but in the central US it is a more recent addition (shows there is longitudinal data available to study Uber) --> will look at if the introduction of Uber there was a reduction in ambulance rates -Can see that before the entrance of Uber the ambulance rate was fairly flat, but after there is a decrease --> shows that the entrance of Uber really did effect ambulance rates

Equilibrium for insurance contracts

Will be at _______ if... -All individuals select the contract that offers the most utility -No contract in the set earns negative profits for the firm offering it -There exists no contract or set of contracts outside the set that, if offered, would attract customers and earn at least zero profit -True _________ will occur at the intersection of the zero-profit and full-insurance line, utility curve will also intersect there