Health exam

• Studying health economics is vital
• Health care in the economy

  • Massive and expensive
  • In the states, ⅙ of 2008 GDP is spent on health care
  • Compared to expenditure in the 1960s where $1/20 of GDP was spent on healthcare
  • Trends have been fairly similar around the globe
  • Health sector has grown massively around the developed world
• Reasons for growth in health sector


  • Increase in the wealth of scientific discoveries and technological improvements
  • Size of the healthcare sector (doctors, pharmaceuticals)
• Health


  A major source of uncertainty and risk
• Health economics was not considered different from economics
• Kenneth Arrow in 1963


  Established that the economics of health is different
• Ways health is different from other goods

  • Uncertain about how much health care → need health insurance
  • Sellers of services (physicians and hospitals) are also different from typical sellers (can't advertise or do price competition)
  • Choice of treatment doesn't depend on socioeconomic status
  • Omnipresence of health insurance in the market distinguishes it from other markets
  • Insurance markets are unusual because of information asymmetries (adverse selection, moral hazard)
  • Health care markets are rife with externalities (e.g., skipping a flu shot)
• Government around the world are deeply involved in financing health care systems
• Size of taxes you pay depend on decisions of politicians and government about how to manage the health care system
• After WWII more governments got involved in health care markets as many countries have introduced new-government financed health insurance programs (medicaid- US, NHS- UK)
• Government expenditure on health is large in the US but is larger in countries like UK, Sweden, or Canada where the gov't is responsible for the majority of health care expenditures
• With an aging population, pressure on health insurance systems paying for health care
• Health care is a growing item in the balance sheets of the gov't
• Debate about whether or not expensive medical technologies are to be adopted (cost effectiveness)
• Insurer distorts prices
• Welfare economics


  Disagreements in health policy are inevitable and turn into normative issues (how the world should be): for many of us adequate health care is a human right
• Role of health economics

  To decrease the unnecessary disagreement on health policy by providing evidence (e.g, tax on fatty food make people healthier?). But one should also translate this knowledge and incorporate them into policy
• Economic theory can predict slope but it cannot predict the degree of responsiveness (e.f., Demand)
• Measurement of the economic behaviour is crucial to determine responsiveness to co-payments introduction or increase in taxes
• Health production function


  HS = f(HC, L, E, G)

  HS (health status) HC (health care) L (lifestyle) E (environment) G (genetics)
• A person's income is an important determinant of their lifestyle, environment, and possibly health care quality and quantity
• Relationship between personal income and health

  • Higher income allows buying health-improving stuff, e.g., pharmaceuticals, good housing, and food, uncovered treatments, gym memberships
  • Higher labor income increases investment demand for health (Grossman)
  • Higher income may directly cause better health through reducing stress
• If increasing someone's income leads to better health, economic policies affecting income indirectly affect health
• Income distribution and related policies in particular may be important in determining population health
• If income doesn't cause health, then changing people's income won't affect their health
• Correlation vs Causation


  • Interested in finding out if there is a causal effect of income on health
  • May also be interested in the reverse effect of health on income
  • Positive correlation between two variables does not necessarily imply causation
• A one-unit increase in income, holding IQ constant, causes health to rise by zero units
• A one-unit increase in IQ, holding income constant, causes health to rise by 2 units
• Income and health are positively correlated, but income does not cause health
• If we observe IQ, statistically holding IQ constant would make the correlation between health and income go away

  
• Most research in health economics is econometric : Research using statistical methods to try to uncover causal relationships
• Causal relationship


  Can be claimed if we can determine people's income by flipping a coin : if heads we make their income high, if tails we make their income low. If we find that income and health are correlated in this world, we know that income cause health.
• Randomized Controlled Trials (RCTs) are important for showing causation in health econometrics
• Causal impact


  An impact that is caused by the program itself. To get this effect we need to compare how people who participated in the program performed compared to a situation where they would have not been in the program (the counterfactual).
• The counterfactual is never observed in the real world. We often infer the counterfactual from what happened to other people or by looking at what happened to participants before the program.
• Before and after comparison is not ok to assess the effect of a school-enhanced physical education program on body size, as many other factors can affect a child's body size.
• Selection bias is an issue when comparing those who participated in the enhanced physical education program to those who did not.
• Randomized evaluation


  Can help uncover causal relationships by randomly assigning people to treatment and control groups, creating a valid counterfactual.
• Steps in a randomized evaluation

  1. Define program eligibility

  2. Randomly assign units to treatment and control groups

  3. Implement the treatment for individuals in the treatment group only
• Non/Quasiexperimental Evaluation Methods


  • Before and After
  • Participant- non Participant (Cross sectional comparison)
  • Multivariate Regression
  • Statistical Matching
  • Difference-in-Difference comparisons
  • Regression Discontinuity
  • Instrumental Variable