drs

							Lecture Notes – Physician Services Market
Why do we have physician
firms? (and what do they do?)
 dual role of physician firms
    the physician as an input into the production of
     physician services
        Demand for physician services is a derived demand.
        Affected by Economic factors
          Prices, Income, insurance coverage, prices of competing goods
        Affected by Non-economic factors
          Age, Illness events, education, marriage, etc.

    the physician as owner and entrepreneur
      Supply of Physician Services is derived from other end of
       market (input markets) such as:
        Physician Labor, nurses labor, technical assistants, etc.
        Capital and equipment
 the physician services firm vs. the market
   vertical and horizontal integration (what's that?)
   What is the alternative? Compare internal (integration) vs.
    external (market) systems.
   Health care has moved away from the market toward internal
    system with government regulation.
 possible goals of the physician firm
     profit
     leisure (of who?)
     patient's health
     What are the relationships?
 important issues to be addressed
   demand side or allocative efficiency
   supply side or technological efficiency
   public policy implications
Demand Side Performance
 Overriding issue is allocative efficiency but difficult to
  measure either benefits or costs => How to assess allocative
  efficiency?
 variations in physicians' fees between areas/physicians
      how should fees vary in a perfectly competitive market?
      if patients have incomplete information and must search?
      if quality differences exist between physicians?
      empirical predictions
          Consider 2 different goods – standardized glaucoma test vs. surgery
           for, say, prostate cancer – Which would have the largest variation in
           fees even when the market is competitive (but with asymmetric
           information)?
          Glaucoma test – why? Doesn’t pay to search for the best deal for
           small budget item.
 is the market for physicians' services competitive?
 empirical results
                          Variation in Physician Fees

                                                         Percent
             Procedure               High         Low   Difference
         Initial office visits        53          31     70.97%
         Normal delivery             1450        775     87.10%
              stress test             300        100    200.00%
         prostate surgery            2000        1200    66.67%
           hernia repair             1030        450    128.89%

     Note: all data is from the 1980s


 Conclusions = ? Get large variation even for large
  budget items =>
 Evidence that the market is not competitive.
 increases in physicians' fees over time
   what should happen to fees in both short and long-run
    if the market is competitive?
       Short-run competitive firms can make profits => prices and
        costs may vary independently.
       Long-run competitive firms can not make profits => prices
        can only vary as costs vary.
   do fee increases over time reflect increased costs?
       Examine handout to answer question – especially look at
        differences over time in how fees change compared to costs as
        measured by overall CPI.
   Conclusions
       Evidence of lack of competition in market
       More evidence of competition in recent past
Supply Side Performance
 the production of physician services
   how do you measure output
       physician visits? - advantages and disadvantages
       annual gross billings? - advantages and disadvantages
   what are the inputs in the production process?
       must be at least 1 physician (why?)
       other inputs? Nurses, technicians, etc.
       relationship between the inputs? Substitutes vs.
        complements?
 technological efficiency
   empirical studies of the use of aides by physicians
       Increases in price increase the optimal # of aides
       Increases in wages decrease the optimal # of aides
       Confirmation of economic theory on optimal use of inputs
       However, aides also found to be underutilized => inefficient
        (see T 15.1 p. 239). What other inputs are underutilized?
          MP/w should be equal if inputs are correctly utilized

          If MP/w larger for a given input => that input is underutilized
           and the reverse.
       Why are too few aides and other inputs used?
          Attempt by Dr. to give best possible care (Dr.’s care) to patient?

          How to use a monopolized input (Dr’s) to maximize profit?
 group practice studies
   Group practices produce more at lower costs => efficient.
   Recent increases in the use of group practice suggests efficiency as
    well
   Economies of scale exist and explain increasing size of physician
    firms.
   Incentive Problems in groups
      Shared profit leads to incentives to shirk. Why?
         As group size increase then incentives to shirk increases.
      If paid flat wage the incentive then still have incentive to shirk.
       Why?
      How does the firm monitor worker effort?

 referrals among physicians (why is this a concern?)
   fee splitting
   multi-specialty firms
 Conclusion = increases in competition are forcing drs to take
  advantage of aides, economies of scale, etc. => the market is
  non-competitive but the level of competition is increasing.
Models of Physician Pricing
 monopoly pricing model (collusive joint profit
  maximization model)
   the importance of entry barriers
   cheating on the cartel
   what happens, according to the model, when insurance
    coverage increases?
       Demand increases and becomes less elastic.
       Both cause price to increase. This prediction is supported by
        empirical evidence.
   why limit advertising by physicians?
 Supplier induced demand
   Two conditions lead to SID
       Assymetric Information
       Dr. as agent
       Based on empirical observation of correlation between things
        like:
          Hospital beds and hospital utilization

          Drs and utilization of Drs services

       Two possible interpreations using D/S analysis
P                                    P

                       S                                    S
                                S1



                           D1                                   D1
                   D                                    D
                                Q                                    Q

    S increases                          D increases

    Drs income falls                     QS increases


    Responds by inducing D               P and Q rise


    P and Q rise
 Target income model
   The first graph is this model
   Notice that the second graph is simplier and evidence suggests it is
    more likely to explain the empirical observation.
 Price rigidities
   Difficult to change price quickly (why?)
   As input prices rise => prices rise slower
   Leads to inducement but only enough to decrease the surplus
   Disapears over time
 Disutility of discretion model
   Problem with SID = why are drs satisfied with a given target – why
    not higher?
   Assume U = U(Y, W, D)
      Y=Drs income, + impact; W = Drs hours of work, - impact
      D=Disutility caused by inducing demand, - impact
      Inducement causes tradeoff between extra income and more work
       and disutility from inducement.
      Therefore, a limit to inducement because of tradeoff
 Profit maximization
   SID has another potential tradeoff
   Increasing D increase profit but also increases costs (diminishing
    returns).
   Tradeoff between the two leads to some limit to SID

 Problems  identification
   Causality – See figures above
     Graph 1 says increase in S causes increase in D

     Graph 2 says increase in D causes increase in Qs

     Causality asks which comes first

 Empirical Evidence on SID
     Some evidence in favor of SID from price rigidity studies
     Is there evidence of SID caused by imperfect information and
      agency?
     Compare initial visits to Drs (patient initiated) vs followup visits (Dr.
      Initiated) => find evidence of SID from differential impact of D.
     Some evidence exists of SID although not large impact
 Small Area Variation
     Figure 10.5 p. 216
     Shows variation in utilization between geographic areas.
      Why?
       Differences in level of competition

       Differences in information/practice styles of Drs.

       Figure 10.6 shows these differences which lead to different
        practice styles.
     How to test the model? How do we measure practice style?
       Studies show that education/feedback/surveillance of Drs
        changes practice style.
       Comparison of relatively homogenous areas still finds SAV.

       Multiple regression studies controlling for other factors still
        find SAV (although smaller).
 Public Policy
    conclusions about the level of competition in the
     physician services market
         Market is not competitive.
         The level of competition is increasing.
    possible public policy proposals to increase
     competition
         Regulation
             The source of much of the inefficiency in the market is
              regulation (e.g., licensing laws lead to non-optimal use of
              labor in health care markets) => reduce regulation?
             Problems

         Medicare payment reform
         Others?
     The Physician as Labor
        why become a doctor? (what matters when making the
         decision?)
                     the cost of becoming a doctor
                     the benefits of becoming a doctor
                                                          Earnings MD
                     graphical

             $                                            Earnings BA/BS




             $0
                                                          time


Graduate                          Graduate
with BA/BS                        with MD
                                             Earnings MD


                                C
$                                            Earnings BA/BS




        B


        A                                    time




 Area A = Direct costs of MD
 Area B = Indirect costs of MD
 Area C = Benefit of MD
 Invest if C > A + B – not quite, also add discount rates
 Empirical Results – see handout on rates of return for MD
    education
     Why specialize?
         the cost and benefits of specialization

         empirical rates of return

 Why have licensing of health care professionals
    3 theories
         Each focuses on a different group benefiting from licensing.

         Public Interest Theory

               Licensing benefits the public - consumers

         Capture Theory

               Licensing benefits the professionals being licensed.

         Political Economy Theory

               Licensing benefits the regulators => sell licensing to highest bidder,
                sometimes public and sometimes professionals.
    Empirical Evidence tends to support political economy theory
         See T 15.4 p. 344 – licensing varies by state => test the impact of licensing on
          professional fees => conclude sometimes fees increases but sometimes not.
         Quality – if licensing increases quality => supports public interest.
                  Again sometimes find quality increased, sometimes decreased => support for
                   political economy theory.
 labor supply of physicians
     why do physicians supply labor?
         labor as a consumption good
             Work because it gives us utility => implications?

         labor as an investment good
             Work because it gives us money => implications

         the relationship between labor and leisure
             Leisure is what you do when you’re not working or
              sleeping/eating/etc.
             Tradeoff between labor and leisure

         what is the wage rate?
             Wage rate equals the price paid for labor

             Wage rate also equals the opportunity cost of leisure
 what does an individual's supply of labor look like?
      Two effects of a wage increase
         Focus on the labor/leisure tradeoff => when leisure ↑ or ↓
          then labor ↓ or ↑.
         substitution effect – w ↑ => consumption of leisure ↓
          because leisure is now more costly (↑ opportunity cost of
          leisure) => substitute less time intensive types of leisure for
          more time intensive types => ↑ quantity supplied of labor.
         income effect – w ↑ => income rises (↑ price of labor) =>
          leisure is a normal good => consumption of leisure ↑ =>
          ↓ quantity supplied of labor.
         Both effects occur simultaneously => which is largest?

             Empirical question – examine behavior as wages
               increase
             Empirical studies of physician labor supply shows the
               following:
    SL
             Income Effect outweighs
W            Substitution Effect




                                     Theoretical Maximum =
                                     16x7




         Substitution Effect outweighs
         Income Effect




                           112           Hrs/Week
    Do market labor supply curves also bend backwards?
          Empirical questions and answer = no. Why not?
             New entry as w increases.

             Fixed components of labor supply

                 One example is malpractice insurance, which induces
                  drs not to reduce labor supply as wage increases.
 Physician Location Decisions
    Hotelling Model
    Say have a concession stand on a beach:
    Where do you locate? Assume customers evenly distributed on beach.
          Why? Because it reduces travel time and maximizes demand and profit.
    Where does second, third, etc. firm locate?
 Same principle for new entry.
 Also same principle is true for Drs.
       Locate where they can have the maximum number of customers.
 Predictions from the model:
       As the number of drs increases => should see more communities
        with drs.
       Same is true of specialists.
       Empirical evidence supports these predictions.