[Background Material and Data on Programs within the Jurisdiction of the Committee on Ways and Means (Green Book)]
[Appendices]
[Appendix C. National and International Health Care Expenditures and Health Insurance Coverage]
[From the U.S. Government Printing Office, www.gpo.gov]




 
APPENDIX C - NATIONAL AND INTERNATIONAL HEALTH CARE EXPENDITURES AND 
HEALTH INSURANCE COVERAGE

CONTENTS

National Health Expenditures
Expenditures for Hospital Care
Trends in Hospital Utilization
	Admissions
	Average Length of Stay
	Hospital Occupancy
	Hospital Employment
Expenditures for Physicians' Services
Supply of Hospital Beds
Supply of Physicians
Health Insurance Status in 2002
	Health Insurance Coverage and Selected Population 
	   Characteristics
	Characteristics of the Uninsured Population Under Age 65
	Trends in Health Insurance Coverage
International Health Spending
References	
	
NATIONAL HEALTH EXPENDITURES

	In 1965, the year prior to the beginning of the Medicare and 
Medicaid Programs, national health expenditures were $41.0 billion. 
After adjusting for inflation, this spending figure represented 
$188.7 billion, or $944 per capita in constant 2001 dollars. Health 
care expenditures increased substantially over the next 36 years. In 
2001, the most recent year for which data are available, the Nation's 
health care bill was $1,424.5 billion, or $5,034 per capita (Tables 
C-1, C-2, and C-3). 

	The average annual rate of increase in real (i.e., adjusted 
for inflation) per capita expenditures was 5.8 percent from 1985 to 
1990. The rate decelerated to 3.5 percent for the period from 1990 to 
1995, and decelerated further to  3.1 percent from 1995 to 2000. It 
has been argued that this relatively low rate of growth in health 
expenditures during the 1990s was due to the influence of managed 
care. However, in recent years, expenditures have grown at rates 
higher than the average rates experienced during the 1990s. Real per 
capita health expenditures grew 5.1 percent in 2001. Real per capita 
health spending is projected to grow at an average annual rate of 
4.7 percent between 2001 and 2005, then decelerate to an average rate 
of 3.6 percent between 2005 and 2010 (Table C-3).


TABLE C-1--NATIONAL HEALTH EXPENDITURES, AGGREGATE AMOUNTS FOR 
SELECTED CALENDAR YEARS 1965-2010





[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


TABLE C-2--NATIONAL HEALTH EXPENDITURES IN CONSTANT 2001 DOLLARS,  
SELECTED CALENDAR YEARS 1965-2010

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



TABLE C-3--NATIONAL HEALTH EXPENDITURES: PER CAPITA AMOUNTS IN 
CONSANT 2001 DOLLARS AND AVERAGE ANNUAL PERCENTAGE INCREASES, 
SELECETED CALENDAR YEARS 1965-2010   


[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


	The majority of health spending is for personal health care 
services that treat or prevent illness and disease in individuals. In 
2001, 86.8 percent of all health spending ($1,236.4 billion) was for 
personal health care. The remaining 13.2 percent ($188.1 billion) was 
spent on health program administration; administrative costs and 
profits earned by private health insurers; public health activities; 
noncommercial health research; and construction of health facilities 
(Table C-1).
	A combination of factors can cause personal health care 
spending to increase: rising prices, population growth, increases in 
the quantity of medical services each person receives, advances in 
medical knowledge and technology, and other factors. Chart C-1 
depicts the share of personal health care spending growth attributable 
to increases in medical care prices, increases in population, and 
increases in real per capita health expenditures (what some experts 
describe as the "intensity" of care). For much of the time period 
shown in Chart C-1, prices played a larger role in personal health 
spending increases than population or non-price factors (e.g., 
improved medical technology or higher utilization). During the late 
1990s, medical care prices, constrained by managed care, grew at 
lower rates than in any other year shown. Price growth is projected 
to increase during the next decade, though not to the high levels 
experienced during the 1970s and 1980s.

CHART C-1--FACTORS INFLUENCING GROWTH IN NOMINAL PERSONAL EHATLH 
CARE EXPENDITURES, 1965-2012

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


	Hospital care and physician services are the two largest 
categories of personal health care spending.  In 2001, they accounted 
for 31.7 percent ($451.2 billion) and 22.0 percent ($313.6 billion), 
respectively, of total national health expenditures (Table C-1). 
Another major service area, prescription drugs, grew from 5.8 percent 
of all national health expenditures in 1990 to 9.9 percent in 2001. 
In real per capita terms, spending on prescription drugs increased  
11.8 percent over 2000 levels, making this category of personal 
health care the fastest growing.  It is projected to remain the 
fastest growing category of personal health care in the near future 
(Table C-3).

	Direct payments for personal health care come from five 
general sources: consumer payments out-of-pocket, payments by 
private insurance companies, Federal funds, State and local funds, 
and "other" private funds. Out-of-pocket payments include payments 
by those without health insurance.  Out-of-pocket payments also 
include payments by the insured for deductibles, coinsurance, and 
costs not covered by insurance (excluding premiums).  "Other" private 
funds consist mostly of philanthropic contributions to the health 
care system.

	Ultimately, all health care spending is paid for by 
individuals through direct payments, cost-sharing, insurance 
premiums, taxes, and charitable contributions.  However, most of 
these payments are redistributed; what a person pays does not 
necessarily reflect how much health care that person receives. One 
who pays relatively high taxes might not have any of their health 
care financed by the government.  Similarly, there are some people 
who pay health insurance premiums, yet use less care than the sum of 
the premiums paid.  Only when individuals pay directly for the cost 
of treatment (either because they are uninsured or because they have 
not met their deductible) do personal expenditures directly reflect 
the amount of care received.

	Table C-4 shows the percent of personal health care spending 
attributable to each source.  In 1965, 52.3 percent of all personal 
health care was financed out-of-pocket, whereas private insurance 
paid for 25.1 percent and the Federal government paid for 8.1 percent.  
In 2001, only 16.6 percent of personal health care was paid out-of-
pocket while private insurance paid for 35.4 percent and the Federal 
government paid for 32.9 percent.  Much of the increase in the 
Federal government's share of health spending occurred during the 
1960s, when the Medicare and Medicaid programs were introduced.

EXPENDITURES FOR HOSPITAL CARE 
	Based on data from the Centers for Medicare and Medicaid 
Services (CMS), hospitals accounted for 31.7 percent of total 
national health expenditures in 2001, down from 41.3 percent in 
1980.  A different data source, the National Center for Health 
Statistics, provides expenses incurred by hospitals based on 
ownership structure (Table C-5). Total expenses for hospitals of 
all types reached $395.4 billion in 2000, up 6.0 percent from the 
previous year. This increase is higher than in recent years; 
expenses for hospitals of all types increased 3.9 percent in 1998 
and 4.9 percent in 1999.  However, the growth in 2000 is still lower 
than the average growth that occurred during the late 1970s


TABLE C-4--PERSONAL HEALTH CARE EXPENDITURES, AGGREGATE AMOUNTS 
AND PERCENTAGE DISTRIBUTION, SELECTED CALENDAR YEARS 1965-2010

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



TABLE C-5-- HOSPITAL EXPENSES BY TYPE OF OWNERSHIP, SELECTED 
CALENDAR YEARS 1975-2000

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


TABLE C-6 -- NATIONAL EXPENDITURES ON HOSPITAL CARE BY SOURCE OF 
FUNDS, SELECTED CALENDAR YEARS 1965-2010


[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


and the 1980s.  From 1975 to 1980, hospital expenses grew at an 
average annual rate of 13.5 percent.  From 1980 to 1985, hospital 
expenses grew at an average annual rate of 10.8 percent.  From 1985 
to 1990, hospital expenses grew at an average annual rate of 
8.9 percent.

	Based on data from CMS, expenditures for hospital care are 
financed primarily by third parties, as shown in Table C-6. In 2001, 
private health insurers paid 33.7 percent of total hospital 
expenditures. Medicare financed 29.9 percent of hospital spending 
in 2001, and Medicaid (including both the Federal and State shares) 
financed 17.2 percent of hospital spending. The share financed by 
out-of-pocket payments from individuals was only 3.1 percent in 
2001, down from 19.8 percent in 1965.

TRENDS IN HOSPITAL UTILIZATION

ADMISSIONS

	Hospital utilization in the United States has undergone 
major change in the past twenty years, greatly influenced by 
technology, health care policy, and population dynamics.  During the 
1970s, hospital admissions increased consistently and had reached 
highs of over 36 million in the early 1980s. With the introduction 
of Medicare's prospective payment system (PPS) in 1983, which 
encouraged more cost efficient treatment methods, admissions declined 
dramatically for several years, as shown in Table C-7.  After 1987, 
total admissions continued to decrease, though more slowly, and 
reached a low of 30.7 million in 1994.  Since 1994, hospital 
admissions again have increased each year, due in part to the 
growing health care needs of adults 65 and older.  The largest 
increases, since 1999, also may be attributable to the weakening 
impact of managed care.  Close to 34 million people were admitted 
to hospitals in 2001, a level roughly comparable to that of 1985.  

	Another significant trend that has developed since the 
implementation of Medicare's PPS has been the increasing provision 
of health care services in more cost-effective outpatient settings 
favored by insurers.  This has been made possible by scientific and 
technological advances which allowed surgical procedures to be 
completed safely in a shorter amount of time. Total outpatient 
visits more than doubled from 212 million in 1984 to 538 million 
in 2001, steadily growing by an average of over 5 percent each 
year (Table C-7).


 TABLE C-7-ANNUAL NUMBER AND CHANGE IN HOSPITAL ADMISSIONS, 
 TOTAL OUTPATIENT VISITS, AND EMERGENCY DEPARTMENT VISITS, 
 1980-2001


 [GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



Hospital emergency department visits also continued to rise steadily 
from 1984 to 1993, though at a slower rate of 2 to 3 percent (Table 
C-7). The 1986 Emergency Medical Treatment and Labor Act (EMTALA), 
which required emergency departments to perform a screening examination 
and treat or stabilize patients for transfer to another facility, may 
have influenced growth during this time frame. This was followed by 
several years (1994, 1996, 1997) of decline in total visits, perhaps 
due to restrictions by insurers.  The Balanced Budget Act of 1997, 
however, required the Medicaid and Medicare programs to reimburse 
hospitals for emergency care that a reasonable person would consider 
necessary and many States passed laws with similar "prudent 
layperson" standards. Since then, hospital emergency department 
visits have increased by 14 percent from 93 million in 1997 to 106 
million 2001.  This number is expected to grow as hospital emergency 
departments are increasingly used as a "safety net" by persons 
without health insurance.

AVERAGE LENGTH OF STAY
	
	Advances in medical technology and drug therapy contributed 
to a steady decrease in the average length of hospital stay from over 
8 days in the late 1960s to 7.1 days in 1982.  The implementation of 
the Medicare PPS system in 1983 caused the average stay to fall 
further to 6.5 days in 1985.  The change was even greater for 
patients over 65, who saw a decline in length of stay from 10.1 in 
1982 to 8.7 in 1985.  In the latter part of the 1980s as outpatient 
visits increased, patients admitted to hospitals tended to be those 
with more severe illnesses which required longer hospital stays, and 
the average length of stay became more stable and even increased for 
those 65 or over.  Beginning in the early 1990s, however, declines 
occurred which were even steeper than in the first years of PPS. The 
average length of stay in 2001 was at 4.9 for all ages, compared to 
6.4 in 1990, a decrease of 23 percent. For persons over the age of 
65, the average length of stay declined 33 percent from 8.7 days in 
1990 to  5.8 days in 2001.  Declines in the past ten years have been 
attributed to greater insurance coverage of post-acute care 
alternatives to hospitalization, an increase in managed care and 
other cost-containment programs, as well as continuing technological 
advances.

TABLE C-8--AVERAGE LENGTH OF STAY AND ANNUAL CHANGE BY AGE GROUP, 
1980-2001 

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



HOSPITAL OCCUPANCY
	Table C-9 shows that community hospital occupancy rates were 
over 70 percent in the early 1980s. The number of hospital beds was 
increasing and exceeded 1 million by 1981.  Following the 
introduction of PPS, however, occupancy rates decreased dramatically.  
From 1983 to 1986, the aggregate occupancy rate fell from 73.5 
percent to 64.3 percent. There was a slight increase in occupancy 
rates in the late 1980s, but even though the number of hospital beds 
had been steadily declining, a reduction in the average length of 
stay lowered the occupancy rate below 63 percent by 1993. Since 
1997, there have been slight increases each year in hospital 
occupancy rates and in 2001 the nationwide rate was at 64.5 percent.  
During this same period of time, the number of community hospitals 
steadily declined from 5,830 in 1980 to 4,908 in 2001. With fewer 
hospitals and beds, the increasing occupancy rate is causing a 
struggle for some hospitals to cope with growth in patient admissions.



TABLE C-9--NUMBER AND PERCENT CHANGE IN HOSPITALS, TOTAL BEDS, AND 
OCCUPANCY RATE, 1980 - 2001

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


HOSPITAL EMPLOYMENT

	Hospitals experienced a significant downturn in total 
employment levels at the time PPS was introduced, with the number of 
full-time equivalent hospital employees (FTEs) declining each year 
from 1983 to 1986 (Table C-10).  

Though variable, the number of FTEs increased by 2 to 3 percent 
through the late 1980s and early 1990s.  After jumps of 4.2 percent 
in 1994 and 5.8 percent in 1995, the growth rate slowed.  After 
reaching a low of 0.2 percent in 1999, it rose to 1.9 percent for 
both 2000 and 2001.  In 2001, almost 4.0 million FTEs were employed 
in community hospitals.   This slower rate of growth since 1996 is 
attributed to a response to managed care and steps taken by hospitals 
to assure greater efficiencies in the provision of care. There are 
concerns, however, that this decline in growth has occurred as the 
volume of admissions and outpatient visits have risen and the aging 
of our population puts increasing pressure on the health care 
delivery system. There are also questions about the quality of care 
that can continue to be provided by a registered nurse workforce 
which showed little sign of growth.  RNs comprise 24 percent of the 
workforce in community hospitals, or 958,026 RNs in 2001. (Community 
hospitals employed 92 percent of the total 1,045,501 registered 
nurses in all hospitals throughout the United States.)

TABLE C-10--ANNUAL NUMBER AND PERCENT CHANGE IN HOSPITAL 
EMPLOYMENT, 1980 - 2001



[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



EXPENDITURES FOR PHYSICIANS' SERVICES

	Health care expenditures for physician services were $313.6 
billion in 2001, an increase of 8.6 percent from 2000. This amounted 
to 22.0 percent of national health expenditures. 

	Third-party payments financed 88.8 percent of physician 
services in 2001. Private health insurance was the single largest 
payer, accounting for  48.1 percent of total spending on physician 
services. The share of physician services paid by private insurance 
increased from 35.3 percent in 1980 to 47.8 percent in 1993. After 
1993, the share of physician services financed by private insurance 
remained relatively stable. From 1993 to 2001, private insurance 
financed between 47.2 percent and 48.6 percent of total physician 
expenditures. The stabilization in the share of physician services 
funded by private insurance is likely the result of the increased 
use of managed care during the 1990s.

	Public expenditures on physician services have grown 
relatively slowly. In 1980, the public sector financed 30.5 percent 
of all physician payments; in 2001, the public sector financed 33.6 
percent of all physician payments. Of the $105.4 billion spent by the 
government on physician services in 2001,  $63.9 billion was for the 
Medicare program. Out-of-pocket payments by individuals for physician 
services have decreased from 30.2 percent in 1980 to 11.2 percent 
($35.0 billion) in 2001 (Table C-11). 

	As measured by Consumer Price Index (CPI), inflation in 
physicians' fees has outpaced that of the U.S. economy as a whole 
since 1980. The inflation rate of 2.8 percent for 2002, however, is 
the lowest since 1964 and the same rate of growth as 1999 (Table 
C-12). Also, the "excess" rate of increase in physician services 
prices above overall inflation since 1993 decreased to an average  
1.2 percentage points per year, down from 2.9 percentage points for 
the years from 1981 to 1992. 

	Mean physician net income (after expenses but before taxes) 
rose from $144,700 in 1988 to $205,700 in 2000 (Table C-13). This 
increase represents an average annual growth rate of 3.0 percent. 
Average annual increases in physician net income between 1988 and 
2000 varied by specialty, geographic area, and employment status. On 
average, physicians in the specialty of radiology experienced the 
highest average annual increase in net income between 1988 and 2000 
(4.2 percent). Over that same time period, physicians practicing 
anesthesiology and obstetrics/gynecology experienced the lowest 
average annual increase in net income (1.9 percent). Physicians in 
the West North Central region experienced the highest average annual 
increase in net income between 1988 and 2000 (3.5 percent). (The West 
North Central region includes North Dakota, South Dakota, Nebraska, 
Kansas, Minnesota, Iowa, and Missouri.) Physicians in the South 
Atlantic region experienced the lowest average annual increase in 
net income between 1988 and 2000 (2.2 percent). (The South Atlantic 
region includes Delaware, the District of Columbia, Maryland, 
Virginia, West Virginia, North Carolina, South Carolina, Georgia, 
and Florida.)

	Table C-14 shows the distribution of physicians' net incomes 
in 2000 by specialty and geographic region. While the average net 
income of all physicians was $205,700, one-half of all physicians 
earned $175,000 or less. One-fourth of all physicians earned $123,000 
or less, while one-fourth earned $250,000 or more. Median incomes 
across all physician specialties remained far apart in 2000. The 
median income for physician's specializing in diagnostic radiology 
and cardiovascular diseases was $300,000, while pediatricians 
reported median incomes of $125,000.

	On average, 40.8 percent of physician practice revenue is 
from some form of managed care contract (Table C-15). Of all 
specialties, obstetrics/gynecology and pediatrics receive the 
highest percent of revenues from managed care  (56.4 percent and 
54.9 percent, respectively). Psychiatry received the lowest percent 
of revenues from managed care, 28.6 percent. The share of practice 
revenue received from managed care was lowest in East South Central 
States. (The East South Central region includes Kentucky, Tennessee, 
Mississippi, and Alabama.) The share of practice revenue received 
from managed care was highest in New England States. (New England 
States include Maine, New Hampshire, Vermont, Massachusetts, 
Connecticut, and Rhode Island.)

SUPPLY OF HOSPITAL BEDS 
	The national supply of community hospital beds per 1,000 
population steadily increased from the 1940s, reaching 4.5 beds 
per 1,000 population in 1980.  By 2000, the number of beds dropped 
to 2.9 per 1,000 population. Among the 9 Census regions, the East 
South Central experienced the largest increase, from 1.7 per 1,000 
population in 1940 to 5.1 in 1980. By 2000, this number had declined 
to 3.8, but was second in rank to West North Central which had 3.9 
beds per 1,000.  In contrast, the New England and Pacific regions 
never rose above their 1940 levels.  The Middle Atlantic, East 
North Central, and Mountain Census regions experienced increases 
in the 1970s, but are now also below their 1940 levels.  While all 
Census regions experienced a decrease in the number of hospital beds
between 1990 and 2000, the New England (-3.0), Mountain (-2.9), and 
Pacific (-2.5) regions experienced the largest drops in average 
annual percent changes.  The change in the Middle Atlantic region 
was the smallest at -1.9 percent (Table C-16).  


TABLE C-11 -- EXPENDITURES FOR PHYSICIAN SERVICES BY SOURCE OF 
FUNDS,  SELECTED CALENDAR YEARS 1965-2010 


[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



TABLE C-12 -- ANNUAL PERCENTAGE CHANGE IN SELECTED COMPONENTS OF 
THE CONSUMER PRICE INDEX FOR  ALL URBAN CONSUMERS, 1960-2002


[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



TABLE C-13 - PHYSICIANS' AVERAGE NET INCOME1 AFTER EXPENSES BUT 
BEFORE TAXES,  SELECTED CALENDAR YEARS 1988-2000  

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



TABLE C-14--DISTRIBUTION OF PHYSICIAN NET INCOME AFTER EXPENSES 
BUT BEFORE TAXES,  BY SPECIALTY AND CENSUS DIVISION  

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]




TABLE C-15--AVERAGE PERCENT OF PRACTICE REVENUE FROM 
MANAGED CARE, 2001



[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



TABLE C-16--COMMUNITY HOSPITAL BEDS PER 1,000 POPULATION AND 
AVERAGE ANNUAL PERCENT CHANGE BY REGION AND STATE, SELECTED 
YEARS 1940-2000


[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



SUPPLY OF PHYSICIANS

	Physician shortages in the 1950s and 1960s led to Federal 
and State initiatives to increase the supply of physicians. The 
resulting growth in the number of physicians per capita led to 
forecasts that an oversupply was imminent and that the nation had 
too few generalists and too many specialists. In 1994, the Council 
on Graduate Medical Education (COGME) reported that by the year 
2000, "an expected staffing ratio of one physician for every 370 
Americans,  was more than adequate and the nation would do well to 
greatly scale back its residency programs and shift training away 
from specialists and toward primary care." (Weiner, 2002)

	Tables C-17 and C-19 indicate that while the total number 
of physicians grew by 40 percent from 1970 (334,028) to 1980 
(467,679), over the next ten years the number grew at a slower rate 
(31.7 percent) to 615,241 in 1990. From 1990 to 2000, the total 
number of physicians grew 32.3 percent to 813,770. In 2000, the total 
physician-population ratio was 294 per 100,000 persons or one 
physician for every 340 Americans.

	The predicted physician surplus did not materialize, due in 
part to the lack of sustained impact of managed care.  According to 
the American Medical Association, the percent of physicians in 
primary care and primary care specialties (general and family 
practice, internal medicine, obstetrics and gynecology, and 
pediatrics) has remained relatively stable at approximately 34-40 
percent since 1970.  (See Tables C-17 and C-19 for number of 
physicians by specialty.)  Some hospitals, however, are now reporting 
it difficult to recruit specialists such as radiologists, orthopedic 
surgeons, anesthesiologists, and cardiologists (Thrall, 2003).

	There are also concerns about the geographic distribution of 
physicians and a continuing problem with access to care in more rural 
areas.  Table C-18 shows variations in the supply of non-Federal 
physicians relative to population by State.  In 2000, the District of 
Columbia had the highest ratio (718 physicians per 100,000 population) 
while Idaho had the lowest ratio (178 physicians per 100,000 
population).

	The adequacy of physician supply is once again the subject 
of debate. Richard Cooper and colleagues at the Medical College of 
Wisconsin forecast that the demand for physician services will 
continue to increase as the U.S. economy grows.  They warn that as 
the U.S. population expands and ages, if the per capita number of 
physicians remains flat, today's small national shortage of 
physicians will become a deficit of 200,000 by 2020 (Cooper, 2002).  
The Council on Graduate Medical Education is reported to be 
reconsidering its earlier position in the light of a study by 
Ed Salzberg of the University of 

TABLE C-17--PHYSICIAN SUPPLY BY MAJOR CATEGORIES, SELECTED 
YEARS 1970-2000



[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


Albany Center for Health Workforce Studies that forecasts a shortage 
of about 150,000 physicians by 2020 (Romano, 2003).

	The long training "pipeline" makes it difficult to predict 
physician supply. According to the Association of American Medical 
Colleges (AAMC), since reaching an all-time high of 47,000 for the 
1996 entering class, the number of medical school applicants 
declined to 33,501 for 2002.  They projected an increase of 4 to 6 
percent for 2003, based on the number of individuals who took the 
Medical College Admission Test in 2002 and initial applications. 
While the number of residents in training declined from 98,143 in 
1997 to 96,410 in 2001, there was an increase to 98,258 for 2002.  
The number of medical school graduates remained more stable, though 
declining slightly from 16,143 in 1998 to 15,640 in 2002 
(Table C-20).

	The supply of international medical graduates (IMGs) also 
could have a significant impact on the U.S. physician workforce.  
IMGs comprised 17 percent of total physicians in 1970.  By 2000, 
this had risen to 24 percent of the total physician population 
(Table C-17). The number of residency positions occupied by IMGs 
fluctuated over the period 1971-95.  Due to stricter immigration laws 
and more rigorous competency requirements, IMGs dropped from 
41 percent of all residents in 1971 to about 17 percent in 1985.  
Since then the number of IMGs in training in the United States more 
than doubled, from 12,509 in 1985 to 25,783 in 2002 (26 percent of 
all residents in training). In 1998, the Education Commission for 
Foreign Medical Graduates (ECFMG) began requiring that IMGs pass a 
basic clinical skills examination.  According to the ECFMG, "the 
number of IMGs seeking and receiving certification has decreased, but 
the quality of the applicants appears to have improved and the number 
of IMGs certified annually continues to adequately fill GME positions 
not taken by U.S. medical graduates." (Whelan, 2002) While the 
portion of IMGs in the U.S. physician workforce continues to 
represent approximately one-quarter of the physicians practicing in 
the U.S., since 1999, there has been a decrease in the number of 
foreign nationals and a trend toward a higher percent of U.S. citizens 
in the pool applying for certification.

	There is considerable uncertainty as to whether there is 
enough evidence to suggest a coming physician shortage.  Those who 
are skeptical that there is a doctor shortage suggest it is limited 
to problems with recruitment of specialists such as radiologists, 
orthopedic surgeons, anesthesiologists, and cardiologists, or to 
hospitals in certain areas because of specific market conditions.  
They also point to the difficulties in predicting how the economy 
will shape health care, e.g, if consumer-directed plans become more 
common, demand for services could fall.  New kinds of health care 
delivery systems also could emerge with changes in licensing allowing 
clinicians to do some of the work now restricted to doctors 
(Thrall, 2003).


TABLE C-18--NON-FEDERAL PHYSICIAN/POPULATION RATIOS1 AND RANK BY 
STATE, SELECTED YEARS 1970-2000



[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]

TABLE C-19 -- PHYSICIANS TOTAL AND BY SPECIALTY, PERCENT 
DISTRIBUTION, AND PERCENT GROWTH, FOR 1970, 1980, 1990, 
AND 2000




[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



TABLE C-20 -- MEDICAL SCHOOL GRADUATES, FIRST-YEAR RESIDENTS, AND 
TOTAL RESIDENTS, SELECTED YEARS 1965-2002

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



TABLE C-21 -- INTERNATIONAL MEDICAL GRADUATE RESIDENTS BY 
CITIZENSHIP, SELECTED YEARS 1971-2002


[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


HEALTH INSURANCE STATUS IN 2002
	Most people have some form of health insurance. In 2002, an 
estimated 84.8 percent of the total noninstitutionalized population 
had public or private coverage during at least part of the year.  
However, an estimated 43.6 million Americans, or 15.2 percent of the 
population, were without coverage in 2002. Almost all the uninsured 
were under age 65; consequently, 17.2 percent of the nonelderly 
population was uninsured. This section examines characteristics of 
both the insured and the uninsured populations in 2002 (Peterson, 
2003), and reviews trends in health insurance coverage over the 
1987-2002 period.

	Estimates of health insurance coverage in 2002 are based on 
analysis of the March 2003 Current Population Survey (CPS), a 
household survey by the Department of Commerce's Census Bureau.  
Each year's March CPS asks whether individuals had coverage from 
selected sources of health insurance  at any time during the 
preceding calendar year. Thus, the March 2003 CPS reflects 
respondents' recollection of coverage during all of 2002.  

HEALTH INSURANCE COVERAGE AND SELECTED POPULATION CHARACTERISTICS


Age
	Table C-22 provides a breakdown of health insurance coverage 
by type of insurance and age.  In 2002, compared to other age 
groups, those under age 5 had the highest rates of coverage in 
Medicaid, the State Children's Health Insurance Program (SCHIP), or 
some other program for low-income individuals (30 percent). Young 
adults ages 19 to 24 were the most likely to have gone without health 
insurance for the entire year.  While most in this age group (55 
percent) were covered under an employment-based plan, 31 percent had 
no health insurance.  Young adults are often too old to be covered 
as dependents on their parents' policies and, as entry-level 
workers, do not have strong ties to the work force.  In addition, 
some may feel that they are in good health and choose to remain 
uninsured, spending their money on other items.  After age 25, the 
percentage of people without health insurance decreases.  Of those 
age 65 and over, 95 percent were covered by Medicare, and less than 
1 percent were uninsured for the entire year.  The remainder of this
section focuses on the population under age 65.


TABLE C-22--HEALTH INSURANCE COVERAGE BY TYPE OF INSURANCE AND 
AGE, 2002

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]



Other Demographic Characteristics
	Table C-23 shows the rate of health insurance coverage by 
type of insurance and selected demographic characteristics - race, 
family type, region, poverty level, and citizenship - for people 
under age 65. In 2002, whites were least likely to be uninsured 
(12 percent), while Hispanics were most likely (34 percent). The 
rate of employment-based health coverage was highest among whites 
(74 percent), and the rate of public coverage was highest among 
blacks (25 percent). 

	People residing in two-parent families were most likely to 
have employment-based health insurance (74 percent) and least likely 
to be uninsured (13 percent).  People in a family headed by a single 
mother were most likely to have public coverage (37 percent), 
compared to other family types, while people in a family headed by a 
single father were most likely to be uninsured  (28 percent). Although 
people in single-father families were more likely to have employment-
sponsored health insurance than those in single-mother families, that 
difference was eclipsed by the higher rates of public coverage in 
single-mother families.

TABLE C-23--HEALTH INSURANCE COVERAGE BY TYPE OF INSURANCE AND 
DEMOGRAPHIC CHARACTERISTICS FOR PEOPLE UNDER AGE 65, 2002



[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


	People were less likely to be uninsured if they lived in the 
Midwest (13 percent) or the Northeast (15 percent) than if they lived 
in the South (20 percent) or West (19 percent).  At least 70 percent 
of those living in the Northeast and Midwest had employment-based 
health insurance, compared to 63 percent in the South and the West.  


TABLE C-24--HEALTH INSURANCE COVERAGE BY EMPLOYMENT CHARACTERISTICS
FOR PEOPLE UNDER AGE 65, 2002

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


	Among individuals with incomes at least two times the poverty 
level, 11 percent went without health insurance compared to 34 percent 
of the poor (i.e., those with incomes below the poverty level). Only 
20 percent of the poor received health coverage through employment, 
and 44 percent had public coverage.  Eighty percent of people with 
incomes at least two times the poverty level were covered through an 
employer, and only 5 percent had public coverage.

	Non-citizens were more likely to be uninsured than people 
born with  U.S. citizenship (i.e., "native") - 45 percent versus 
15 percent, respectively.  Non-citizens accounted for 8 percent of 
the population under 65 but were  20 percent of the uninsured. About 
42 percent of non-citizens were covered through employment, compared 
to 69 percent of native citizens.

Employment characteristics
	For the second year in a row, the prevalence of job-related 
health insurance fell, to 67 percent among the nonelderly in 2002. 
Table C-24 shows the rate of health insurance coverage by the 
employment characteristics of the primary worker in the family. In 
2002, only 9 percent of workers in large firms (1,000 or more 
employees) and their dependents were uninsured, compared to 
34 percent in small firms (less than 10 employees).  People who 
reported working in small firms and their dependents accounted for 
15 percent of the under 65 population but 29 percent of the 
uninsured.  Insurance coverage varied according to industry, as well.  
The category of arts, entertainment, recreation, accommodation and 
food services had the highest proportion of uninsured workers and 
dependents (34 percent).  Only 5 percent of those associated with 
employment in public administration were uninsured, and none of 
those associated with employment in the armed forces were uninsured.

CHARACTERISTICS OF THE UNINSURED POPULATION  
UNDER AGE 65	

	People who lack health insurance differ from the population 
as a whole: they are more likely to be young adults, poor, Hispanic, 
or employees in small firms.  Chart C-2 illustrates selected 
characteristics of those under age 65 who were uninsured for 2002.  
Approximately 17 percent of the uninsured were 19 to 24 years old, 
even though this age group represents less than 9 percent of the 
under 65 population. 

 	Although Hispanics were more likely to be uninsured than any 
 other racial or ethnic group, they made up less than 30 percent of 
 the uninsured.  Whites (non-Hispanics who report being only white) 
 were the most numerous racial or ethnic group among the uninsured.

	More than half (55 percent) of the uninsured were full time, 
full year workers or their spouses and children.  Approximately 
18 percent had no attachment to the labor force.  More than three-
quarters of the uninsured were above the poverty level.  Even though 
the poor accounted for only 12 percent of the under 65 population, 
they represented 24 percent of the uninsured.  For the second year 
in a row, more than a quarter of the uninsured were not native-born 
citizens (that is, they were either not citizens or were 
naturalized citizens).

TRENDS IN HEALTH INSURANCE COVERAGE

	Historical trends in coverage by type of insurance for the 
noninstitutionalized U.S. population under age 65 are shown in Table 
C-25. Because of changes in the CPS questionnaire, direct comparisons 
of the years shown can only be made within these three time periods: 
1987-1993, 1994-1999, and 1999-2002.

	Beginning with 1994 data, the survey included additional 
questions about private coverage and the order of questions was 
altered such that questions about private coverage preceded questions 
about other forms of health insurance. Consequently, differences in 
coverage between 1993 and 1994 are a function of changes in the CPS 
questionnaire in addition to actual changes in coverage. It is not 
possible to assess the impact of each on the estimate for 1994. 
Beginning with 1999 data, the survey included a followup question to 
those who reported no source of insurance to confirm that they were 
indeed uninsured.  This question resulted in the release of new 
estimates for 1999, reducing the number of uninsured estimated 
through the survey.  Table C-25 first shows the 1999 results before 
the verification question was asked, followed by the results that 
reflect the impact of the verification question, which is now 
standard in the survey.

	Between 1987 and 1993, the percentage of nonelderly Americans 
without health insurance increased by about 3 percentage points, from 
about 14 percent to 17 percent. During this period, the percent with 
employment-based coverage declined steadily, the percent with Medicare 
or Medicaid increased, and the percent uninsured continued to steadily 
increase.  

	Between 1994 and 1999, the percent with employment-based 
coverage increased by 2.5 percentage points from 65.3 to 67.8 percent. 
Between 2000 and 2002, the percent with employment-based coverage 
declined for the first time since the early 1990s.  From 1999 to 
2002, the number of people with government-provided health insurance 
increased by 5.1 million.  Following declines in the number of 
uninsured from 1998 to 2000, the number of uninsured rose in 2001 
and 2002 - to 43.3 million (17.2 percent of the nonelderly 
population) in 2002.



CHART C-2--CHARACTERISTICS OF THE UNINSURED POPULATION UNDER 
AGE 65, 2002


[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


		
TABLE C-25--HEALTH INSURANCE COVERAGE FOR THE NONINSTITUTIONALIZED 
 U.S. POPULATION UNDER 65, 1987-2002



 [GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


 INTERNATIONAL HEALTH SPENDING
	This section analyzes trends in health expenditures for the 
29 Organization for Economic Cooperation and Development (OECD) 
countries from 1960 to 2000. Table C-26 illustrates total health 
expenditures as a percentage of gross domestic product (GDP). Of the 
countries reporting data in 1960, the mean percent of GDP spent on 
health care was 4.0 percent. In 2000, the mean percent of GDP spent 
on health care by OECD countries was 8.1 percent. For all years for 
which data are available, the share of GDP spent on health care by 
the United States was higher than the OECD average. Since 1982, the 
United States has spent a larger share of its GDP on health care 
than any other OECD nation.

	In terms of dollars per capita, the United States spent 
$4,631 on health care in 2000, more than double the OECD average of 
$1,967. The per capita amount spent in the United States was the 
highest of all the OECD countries for which data are available. 
Switzerland spent $3,222 per capita on health care, the second-
highest amount of all OECD nations.

	The public sector plays a relatively small role in financing 
U.S. health care costs compared to other OECD countries. In 2000, 
44.3 percent of health care in the United States was financed by the 
government. This share is smaller than any other OECD country. By 
contrast, the public sector funded 91.4 percent of health care in 
the Czech Republic in 2000. This share is larger than any other 
OECD country.



TABLE C-26 -- TOTAL HEALTH EXPENDITURES AS A PERCENTAGE OF GROSS 
DOMESTIC PRODUCT, PER CAPITA HEALTH SPENDING AND PERCENTAGE OF 
MEDICAL EXPENDITURES PUBLICLY FUNDED, OECD COUNTRIES FOR SELECTED 
CALENDAR YEARS 1960-2000

[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]


REFERENCES

American Hospital Association.  Hospital Statistics. Individual 
years, 1981-2003.

American Medical Association. (various years). JAMA [Each year of 
this journal devotes an issue to medical education. The most recent 
is 2002. 290(9).] 

American Medical Association. (1998a). Physician marketplace 
statistics 1997/98. Chicago. 

American Medical Association. (1998b). Socioeconomic characteristics 
of medical practice 1997/98. Chicago. 

American Medical Association. (1999). Physician Socioeconomic 
Statistics, 1999-2000 Edition. Chicago. 

American Medical Association. (2002). Physician Characteristics and 
Distribution in the United States, 2002-2003 Edition. Chicago. 

Association of American Medical Colleges. (2003, October 30). Decline 
of medical school applicants continues in 2002.  Press Release.

Cooper, Richard A. (2002, January/February).  Economic and demographic 
trends signal an impending physician shortage.  Health Affairs, 21 (1), 
140-154.

National Center for Health Statistics.  Health, United States, 2002.  

National Center for Health Statistics.  National Hospital Discharge 
Survey. Individual years, 1980-2001.

Romano, Michael.  (2003, June 16).  The vanishing doctor surplus.  
Modern Healthcare, 33 (25), 28.

Thrall, Terese Hudson.  (2003, February).  Mds: [missing doctors?].  
Hospitals & Health Networks, 77 (2) 32-36.

Weiner, Jonathan P.  (2002, January/February).  A shortage of 
physicians or a surplus of assumptions?.  Health Affairs, 21 (1), 
160-162.

Whelan, Gerald P.  (2002, September 4).  The changing pool of 
international medical graduates seeking certification training in 
U.S. graduate medical education programs.  JAMA, 288 (9), 1079-1084.