Medical Costs:

short-term v. L o n g – Term forecasts

Actuaries Club of Philadelphia

February 16, 2010

Thomas E. Getzen, Ph.D.Temple University, Dept. of Risk

CEO, International Health Econ Assoc.

• Short-rundata, contracts, already exist

• Long-Runmacroeconomic trends, constraints

GDP regression slide

$10,000 $15,000 $20,000 $25,000 $30,000 $35,000 $40,000

$-

$1,000

$2,000

$3,000

$4,000

$5,000

$6,000

$7,000

[Figure 13.1 U.S. Health Expenditures and Income per capita, 1960-2005]

Income (2000 dollars)

Hea

lth

(20

00 d

olla

rs)

Slope = 0.228 r2= .977

but, SR almost no correlation

-4.0% -2.0% 0.0% 2.0% 4.0% 6.0% 8.0%

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

9.0%

[Figure 13.3 Annual percentage rates of change]

Income (% change from last year)

Hea

lth

%

WHY?Budgets are already set for 2010--what happens

now cannot change much.

Multi-year employment contracts.

Generally slow inertial adjustment of Health Care.

implication: Recession which has just begun to depress HC spending will continue to do so through 2010, 2011, 2012, 2013.

employment

Health

Total

Correlations slide

-2.0% -1.0% 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0%0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

9.0%

[Figure 13.4 Annual percentage rates of change (lagged moving average)]

Income (average of 3–5 years back)

Hea

lth

%

Long-Run?

• Its all about the money---

Long Run ---

There must be a limit, eventually, somewhere, we will run out of money to spend.

0%

10%

20%

30%

[Figure 18.1 Growth Curve: Health Share of GDP 1850 to 2099?]

GDP + 0% to + 1%

GDP + 1% to + 0%

GDP + 2% to + 5% ?

S-curve Growth

Trends

1) Short-run fluctuations --why the recession and inflation timing will make any "reform" program look good in 2012.

2) Actuarial Uncertainty -- in the long run, the uncertainty is vast. Clients want precise answers--and it is just not possible to predict hc trends 20 years out all that well. Why saying that a DB retiree benefits plan is "102% funded" is almost meaningless. Our only certainties are Death, Taxes -- and the federal budget constraint.

Short-Run Fluctuations

• Inflation• Recession• Underwriting Cycle• Premium/Cost shifting & offsets

Inflation

Inflation moves HC$ 1:1 over the long run.

But, short-run, only 1/3 picked up each year.

Implication: A surge in inflation in 2011 will “undercount” the medical cost increase in 2011 - 2012 - 2013.

Recession (GDP)

• LR income (GDP) effects dominate.

Underwriting Cycle

Bottomed out in 2008?

implication: as cycle turns up, will tend to raise premium costs above med cost.

Underwriting cycle slide

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

18.0%

20.0%

[Figure 5.2 The "Underwriting Cycle" of Health Insurance 1988–2008]

source: Kaiser/HRET survey of employer health benefits and U.S. National Health Accounts

Increase in insurance premi-ums

Growth in medical costs

Premium-shifting

Continued employer “Premium Cutting” by shifting cost to employee co-pays, contribution%, deductibles, etc.

Pressure from need to balance Medicare, Medicaid cuts.

implication: ???

Short-Run Forecasts

• Understanding lags, inertia, inflation and underwriting cycles revolutionized SR forecasting in the 1990s.

• Outlook for 2012 -- combination of “enforced” and “illusory” cuts will make cost growth look very modest.

How Accurate Can Actuaries Be?

Accuracy Measured

• About 1% to 2% error per year.

• “fluctuations” versus “trend errors”

Short Run Accuracy --

• Actually not bad.• Most good firms and knowledgeable

actuaries can do better than econometricians and forecast models.

• SR there are many different situations, and many factors, that all tend to wash out over the long run.

Long Run Forecasts

• CMS• CBO• SOA

Some Differences, but all use

• Same (CMS) ST % changes (first 5 years)

• Transition Period• Same long run assumption:

GDP+0% (ie, equal to growth in wages)

SOA-Getzen Model

Model Input Page

Model Output page

$Health Chart

QuickTime™ and a decompressor

are needed to see this picture.

Share Chart

QuickTime™ and a decompressor

are needed to see this picture.

# # # Thank You! ###