Advanced economics

• Factor accumulation is not the key 
• Technology diffuses quickly, so theoretically is unlikely to explain national variation 
• Profits and prices in the health sector are important – the US has significant monopoly power 

• Find r^2 of 0.92 when explaining health spending using GDP 
• Finds healthcare is a luxury good (e of 1.3)
• Contradicts micro level of elasticity, which are lower than 1 
• Uses cross sectional macro OECD data from 1970s 

• Measures the size of factors attributing to growth in health spending – 0.25-0.5 

  • Growth of insurance 
  • Aging 
  • Income growth (elasticity of 0.4) 
• Assumes that the unexplained residual is the result of growth in tech 
• This means there is not such a large welfare loss from rising spending – making use of new opportunities 

• Increasing medical technology accounts for 0.25-0.5 of growth 
• Income effect – macro level elasticity 1, pure income effect of 0.6-0.9. Accounts for  0.25-0.5 
• 5-20% of total growth due to increases in prices 
• 10% due to expansion in insurance   
• 'income plays a critical role, primarily through the actions of governments and employers on behalf of pools of consumers 
• Dynamic moral hazard – the decision to pay and the decision to use do not occur at the same time – innefficiently broad application of new technologies 
  • Incentive for over use can be reduced using DRGs, rather than fee for service 
• 'the rate of technological innovation is influenced by the size of the market, which in turn is influenced by income and insurance' - aggregation issues
• Difference between micro and macro is insurance
  • Households pay for care net of insurance
  • Countries have mechanisms to adapt spending to match incomes  

• Productivity growth is intrinsically low for health spending  
• It's a service industry 

• Suggests healthcare is a staple, not a luxury good, but not statistically significant from 1 
• Estimate sensitive to functional form  - studies use linear
• Using exchange rates rather than PPP has issues because they are not market specific - growth in prices varies
  • spending elasticity not income elasticity
• Relies on using micro level models to understand macro data – but there's no reason to believe that the two act in the same way  
• micro OVB - income distribution
• macro OCB - government preferences/structure
• Using PPP leads to a staple
• intepretation
  • relationship is not causal
  • R^2 is not causal

• US has higher levels of spending, but growth not out of line 

• Healthcare productivity depends on: 

  • Heterogeneity of treatment effects across patients 
  • The shape of the health production function 
  • The cost structure of procedures such as MRIs with high fixed costs and low marginal costs 

• Three types of medical technology 

  • Highly cost effective interventions with little chance of overuse – ART 
  • Treatments highly effective for some but not for all  
  • Grey area treatments with uncertain clinical value such as ICU days 
• Adopting 3 and too much of 2 leads to high spending growth (as is the case in the US) 
• Some technology induced growth is spending is welfare enhancing, some is not 

• Three key changes 
  • Massive expansion in health technology in US 
  • Expansion of insurance coverage (22% - 73% 1960-84, and greater breadth) 
  • Personal health expenditures have significantly increased – as %  
• The expansion of health care insurance has paid for the development of cost increasing technologies and these have expanded demand for insurance 

• Claims 
  

  • R and D is affected by method of payment for available technologies (I.e. affected by expected utilisation, which is affected by insurance) 
    • DRGs encourage providers to be cost conscious 
  • Demand for insurance depends on available technology 
  • Growth in spending is a product of the interaction between insurance and tech 

• Clinicians decide quantity of services to provide, to trade off patient benefit and hospital surplus 

• Forms of contract 

  • Full cost reimbursement: clinicians maximise patient benefit 
  • Prospective payment: clinicians are sensitive to cost, if they weight cost and treatment benefit equally (no private social benefit), social optimum is obtained, otherwise, under provision 
  • Cost sharing: correct fraction can induce optimum trade off between cost and benefit 

• Trade off between cost and one dimensional quality  
• Demand does not reflect quality 
• Treatment benefit from quantity and quality 

• Provider motivation 

  • Fully benevolent provider – block and cost and volume can achieve optimal 
  • Partially benevolent providers – some degree of cost sharing is optimal  
  • Full self interest- first best not possible -  the second best can be achieved by cost per case 

• Demand reflects quality 
• An optimal trade off between cost and quality can be achieved via prospective payment when all patients are identical  
• When a provider cannot refuse patients who require high treatment costs or discriminate patients by qualities 
• optimally designed prospective payments can implement the efficient quality and cost reduction efforts, but cost reimbursement cannot induce any cost incentive 
• When the provider can refuse expensive patients, implementation of the first best requires a piecewise linear reimbursement rule that can be interpreted as a mixture of pure prospective payment and pure cost reimbursement.  
• Adjusting prospective payment may not be effective, it alters cost reduction incentive 
• Quality discrimination can address the issues of dumping under prospective payment 
• Screening for dumping and cherry picking requires effort 

• Measures market power of providers according to markup in the NHS internal market 
• It was found that prices were not solely related to cost 
• Some evidence that prices were lower in more competitive areas 

• If quality has one dimension, a hybrid of block payment and cost per case can be optimal 
• Inefficient to treat all those that demand healthcare, consumers are not sensitive to price and supply can be constrained 
• In this setting, price for quantity schedules will lead to quality being too low, even when costs are known  
• When capacity is constrained, it may be beneficial to base payments on the waiting list, as this is a further indicator of quality 

• If quality is multi dimensional and not correlated in the same way that the payer preference is correlated, even this will be inefficient 
  

  • There are more dimensions then instruments 

• Some providers might have altruistic tendencies, and if altruism is high enough, there may be a negative relationship between competition and quality  

• Use change in pay in VA hospitals as an instrument for analysing competitiveness of nurse pay 
• Labour supply to individual hospitals is inelastic in short run – short run monopsony, high wage setting power 
• Localised competition  
• Marginal product high above wages 

• Previous empirical findings 

  • Persistent shortages 
  • Increases in minimum wage can increase employment  

• Find that labour supply is not responsive to wages in the short run 
• Vacancy rate is high  

• Special features 

  • Female dominated – career breaks 
  • Second earners 
  • No close job substitutes for skills 
  • Non wage factors affect attachment to job – treatment benefit  
• Wage increases are likely to be inneffective in increasing labour supply  

• Increasing intrinsic motivations by altering conditions is likely to be more effective  

  • Promotion and training 
  • workload 
• Spouse wage matters 

• Labour supply is inelastic  
• Build a model that takes into account employment options and work related utility, not purely as a result of wage 
• Compares willingness to work shifts Vs daytime and suggests the differential wage is unnecessarily high  
• Find income elasticity with respect to different types of nursing (e.g. shift Vs day) 
• Job satisfaction is more important 
• norway 

• 'firm labour supply is upward sloping, independent of market structure' - I.e. concentration 
• Wage level and concentration are not correlated in the US 
• Wage decline with increases in hospitals concentrationn- suggests classic monopsony in the short run, but not the long run  

• Wage is not found to be correlated with mobility – new monopsony 
  

  • RNs display greater inter employer mobility than the general population 
• Therefore, market for RNs is not a new or classic monopsony 

• Nursing is a vocation, increasing the wage attracts lower quality candidates 

• Vocation 

  • Intrinsic motivation 
  • Willing to exceed contract  

• Healthcare is an individual necessity and a national luxury 
• Individuals purchase insurance, within the pool, their level of spending is related to need 
• Most variation at the individual level is due to need  
• Units of observation must correspond with the scale that decisions are made  

• the hospital is multi product 
• It can shift focus from some areas to other areas 
• Uses PROMs instead of mortality as its relevant to more patients- based on gains 
• Find that the introduction of patient choice of hospital may have had a negative effect on elective surgical quality 
• Choice leading to quality competition implies patients can judge quality 
• Did not find a correlation between AMI and PROMS 
• IV 
  • Sicker patients may choose higher quality hospitals because quality is more important to them 
  • A high quality hospital may attract patients from farther afield, thus giving it a larger market radius, and making it appear more competitive 

• AMI mortality is the best and most common measure of quality 

  • Benefits 

    • Not subject to gaming 

    • Clear link between quality of treatment and outcomes 

    • Emergency procedure, taken to nearest hospital with no choice 

    • Correlates well with other measures of quality  

  • Costs 

    • Not relevant to everything hospitals do 

• Mortality fell during competition and fell more in more competitive areas 

• Costs did not rise 

• Some providers might have altruistic tendencies, and if altruism is high enough, there may be a negative relationship between competition and quality  
• Competition will reduce the focus on quality 
• Under competition – treat fewer patients – less treatment benefit and less revenue – less incentive to invest in quality 

• Providers facing more rivals had higher mortality, before the introduction of choice 
• After the introduction of choice, this pattern was present, but smaller 

• Do hospitals that face more competition have better management practices? 
• Use marginal constituencies as an instrument 
  • There was an association 
• Theory 
  • Competition based on a fixed price will improve quantity 
  • Competition where providers set their own prices and undetermined effects 
• Effects depend on the features of markets 
• Market structure may be endogenous to quality 
  • New entrants may to choose to locate in areas where quality is poor 
• A large portion of healthcare uses are elderly and frail 
• Evidence from the UK 
  • In the UK, information has been made available to help people choose 
  • Evidence that people choose GPs on quality, as well as location 
  • When competition was introduced, demand for high quality hospitals increased  
  • Mortality fell and quality rose, they fell more in areas where there was more choice 
• Mergers and consolidation don't improve quality and increase spending 
• Competition based on fixed prices improves quality 
• There is a lack of understanding of how decision makers respond to competition 

US timeline

• 1960s and 70s 

  • Cost reimbursement by providers 

  • Competition on quality 

• 1980s 

  • Prospective payment 

• Post 1980s 

  • Managed care 

  • Capitation payments per individual 

UK timeline 

• 1990s – internal market 

• 2000s – patient choice, any willing provider, price negotiation replaced with prospective payment 

• Switch from cost reimbursement in US 
• Measure market concentration using the distances patients live from hospitals 
• Measure quality using AMI mortality 
• Mortality rates were higher in more concentrated markets 

• Low risk patients in less competitive markets receive more intensive treatment than in more competitive markets, but have similar health outcomes  

• High risk patients in less competitive markets receive less intensive treatment than in more competitive markets and have significantly worse health outcomes