The Sustainability of Health Spending Growth

1
The Sustainability of Health Spending Growth

• Glenn Follette
• Louise Sheiner
• Federal Reserve Board

2
Historically, spending growth on health has
exceeded that of consumption
3
Health Spending Projections

• Since 1970, per capita health spending has grown
  an average of 2-1/2 percentage points faster than
  per capita GDP.
• Obviously, health spending cannot continue to
  grow faster than consumption (or GDP) forever.
• Until 2001, Medicare Trustees assumed per capita
  health care spending growth would slow to the
  rate of per capita GDP growth.
• Since 2001, Trustees have assumed that per capita
  health care spending growth would slow to a rate
  1 percentage point faster than per capita GDP.
  This has been called the assumption of 1
  percentage point excess cost growth.

4
Sustainability of Excess Cost Growth from a
Macroeconomic Perspective

• Any rate of excess cost growth will eventually
  crowd out all of GDP and will be unsustainable.
• One rationale for the Trustees assumptions for
  the next 75 years is that, even with such cost
  growth, real per capita non-health consumption
  continues to grow, although slower than per
  capita GDP. No absolute crowd-out of non-health
  consumption.
• Note that what matters for real crowd-out is the
  share of health spending in consumption
• In particular, crowd-out occurs when the share of
  health spending in consumption is equal to the
  ratio of the growth rate of income divided by the
  growth rate of health spending.

5
1 percent excess cost growth and 1.5 percent per
capita GDP growth, crowd-out starts to occur when
health spending reaches 60 of consumption
6
With 1.5 percent excess cost growth, crowd-out
starts when health spending reaches 50 of
consumption.
7
With 2.0 percent excess cost growth, crowd-out
begins at 43 of consumption.
8
Crowding out criteria suggests 1 percent excess
growth is upper bound over the century

• Under baseline macroeconomic projections, health
  spending reaches 40 of consumption by 2080 just
  getting close to real crowd-out.
• With 1.5 percent excess cost growth, health
  spending would reach 67 of consumption
  crowd-out has begun.
• With 2 percent excess cost growth, health
  spending would reach 93 of consumption
  non-health consumption is declining rapidly.

9
Sustainability of Excess Cost Growth from a More
Microeconomic Perspective

• On average, real consumption continues to grow
  even with 1 percent excess cost growth. But, is
  this true across groups?
• Low-income might already spend a larger share of
  income on health spending, which would lead to
  earlier crowd out.
• Elderly or others with high medical expenses
  might also already have larger share of income on
  health spending.
• Projection of 1 percent excess cost growth
  assumes that only force driving health care cost
  growth in future will be exogenous technological
  growth.
• Implicitly assumes that insurance markets and
  public sector involvement in health care remain
  stable. Is this reasonable? Or will increases
  in health spending itself engender changes in
  public and private insurance?

10
Health Spending by Income Quintile
11
Health Spending as a Share of Income
12
Private health spending as a share of income has
been relatively flat
13
(No Transcript)
14
Simulations

• We took 2002 data of the distribution of income
  and public and private health spending for
  elderly and non-elderly
• We projected them forward to 2080 using 1 percent
  excess growth, maintaining 2002 public share of
  health spending adjusted for Medicare drug
  benefit

15
Simulation Results

• Under 1 percent excess cost growth assumption,
  private shares of spending increase over time,
  but most quintiles do not reach the point where
  private health spending is 60 of income and
  starts to crowd out other consumption.
• This is because the public share of health
  spending is very high, particularly for elderly.
  
• Note that, even under these conditions, many
  individuals will face real crowd-out, because we
  dont have perfect insurance (i.e., private
  health spending will still rise if you get sick).

16
Health Care Spending as a Share of Income
17
Caveats

• Note that this exercise does not take into
  account the taxes necessary to finance the large
  rise in public health spending built into the
  baseline.
• If these taxes were to be levied predominantly on
  top 2 or 3 quintiles, then health spending might
  still not crowd out other spending.
• But if these taxes were more evenly distributed,
  then lower-income groups would likely face real
  crowd out (since increased taxes would crowd out
  non-health spending).

18
History of Public Response

• History indicates that there will be more
  pressure on public spending than this. Public
  spending tends to increase as health spending
  burden increases.
• One reasonable view of recent history is that
  public spending increases to keep private
  spending constant as a share of income.
• Another approach is to examine the relationship
  between the public share of spending and the
  total burden on health spending.

19
Private share of income stays relatively constant
while public role grows
20
Public share is larger when health spending is a
higher share of income
21
Simulation Public Share Rises Above Baseline

• To get an idea of the potential costs were this
  trend to continue, we simulate two alternatives
• Alternative A We raise public spending to keep
  private spending for elderly and non-elderly
  constant at 2002 level.
• Alternative B We estimate public spending via a
  regression analysis of past trends.
• Chart shows the additional government spending
  necessary for these alternatives.
• Alternative A By 2030 and thereafter, public
  health spending would be 50 percent higher than
  baseline.
• Alternative B By 2030, public health spending
  would be 16 percent higher than baseline by
  2080 it would be 30 percent above baseline.

22
(No Transcript)
23
Structure of Insurance

• One striking fact is that the share of
  out-of-pocket spending as a share of income has
  been very flat over time.
• This has been accomplished by a significant
  decline in the share of health spending that is
  financed through out-of-pocket payments.
• A lower out-of-pocket share of health spending
  might be a reasonable response to increasing
  health costs A constant share might make health
  care unaffordable and/or provide too little
  insurance.
• Implicit baseline assumptions out-of-pocket
  payments constant share of health spending.
• Alternative assumption out-of-pocket shares
  continue to decline over time. Under this
  assumption, out-of-pocket spending as a share of
  health spending is halved by 2050, and about 40
  percent by 2080.
• Assuming price elasticity of health spending is
  .2, alternative assumption would raise spending
  by 3 in 2030, 9 in 2050, and 12 in 2080.
• This increase in health spending would raise
  public spending by the same percentages.

24
Out-of-pocket expenses have been stable as a
share of income, except among poor elderly
(Medicare drug benefit will help reverse this)
25
Out-of-pocket share of health spending has
declined
26
Conclusions

• From a macroeconomic perspective, 1 percent
  excess cost growth is feasible although
  difficult.
• From a microeconomic perspective, 1 percent seems
  more problematic.
• Because public share of spending is so large,
  most income groups will not face crowd out
  non-health consumption.
• However, federal health spending, already
  difficult to sustain in baseline, will likely
  face additional demands.
• Also, higher health costs will likely raise
  demands for more complete insurance, giving rise
  to further health spending pressures (or making
  it even harder to achieve 1 percent).