III.Estimating the Impact of the Single-Payer Model in Maryland
In this analysis, we estimated the financial impact of the single-payer model on major payers for health
care in Maryland including state and local governments, employers, households and federal
government. In particular, we estimated the distributional impact of this proposal on various subgroups
of payers such as small employers and families in various age and income groups. These estimates
were developed using the Lewin Group Health Benefits Simulation Model (HBSM) which is
specifically designed to provide these detailed distributional impacts analyses for state-level health
reform initiatives.

In this section, we describe the data and methods used in HBSM to develop estimates of the impact of
the single-payer initiative in Maryland. We begin by describing the overall methodology used in the
model. We then explain how the model was adapted to provide Maryland specific estimates of the
impact of this bill on health spending by various payers in future years. Our discussion is presented in
the following sections:

Overview of HBSM
Health Spending in Maryland
Projections to Future Years
A.The Health Benefits Simulation Model
HOBS is a “microsimulation” model of health spending. The core for the model is a representative
sample of Maryland households. For each household in the sample these data provide information on
health insurance coverage, health spending, income employment and basic demographic
characteristics. The model uses these data to show how expenditures for households will change as
they become covered under a new health insurance system such as the Maryland single-payer
initiative. This micro level approach of simulating changes in spending for individual households permits
us to estimate both the aggregate impact of major health reform initiatives as well as the impact on
households of various socioeconomic groups.

For example, the model estimates the increase in utilization which will occur as coverage is extended to
previously uninsured persons. The model also determines which of the services for each individual are
covered under the plan, the reimbursement amount for these services under the plan’s cost sharing
rules, and savings to the sources of payment for this care under current law (family out-of-pocket,
employers, county hospitals, charity care, etc.). Because the model is based upon a representative
sample of the population, it produces aggregate estimates of the impact of policy proposals on total
number of persons affected, aggregate health spending, and program costs. However, because the
model develops these estimates based upon analyses performed on an individual-by-individual basis,
the model also provides estimates of the impact of these policies on various socioeconomic groups.
Using these data, HBSM produces estimates of program impacts by source of payment including:

Employer Impacts
Number of workers and dependents affected
Cost to employers
Impact on firms that do not now insure
Number of firms affected
Uncompensated care cost shift savings
Tax savings (corporate deductions for health benefits, if applicable)

Provider Impacts
Utilization by type of service/provider
Sources of payment for care
Expenditures for services by type of service/provider
Hospital uncompensated care

Household Impacts
Number of insured by income, age, sex, etc.
Family premium payments
Family out-of-pocket spending

Government Impacts
Expenditures under Medicaid expansions
Offsets to general assistance
Offsets to public hospitals
Corporate income tax losses
Tax revenues under various financing mechanism

The basic data source used in this analysis is the Maryland subsample of the March 1999 Current
Population Survey (CPS) conducted by the Bureau of the Census. These data provide detailed
information on Maryland residents by age, income, employment status and other demographic
characteristics Figure 2 shows our estimate of the distribution of Marylanders by primary source of
insurance in 2001.

Because the CPS does not include health spending data, we merged the Maryland subsample of the
CPS with the 1987 National Medical Expenditures Survey (NMES) data which includes health care
utilization and expenditures data for households across various income, age and employment status
groups. The population and income data in the database were adjusted to 2001 based upon the best
available projections for that year. Health expenditures data were then controlled to replicate aggregate
health expenditures estimates for 1998 by type of service and source of payment derived from state-
wide health expenditures estimates developed by the Maryland Health Services Cost Review
Commission.
Figure 2
Distribution of Persons in Maryland by Primary Source of Insurance Coverage in
2001 (Average Monthly Coverage Estimates: in thousands)
Source: Lewin Group Estimates using the Maryland version of the Health Benefits Simulation Model (HBSM).

B.Projections Through 2001
The household database was “aged” to be representative of the Maryland state population in 2001.
This was accomplished by adjusting the population totals in these data to reflect trends in population
growth by age and sex. The earnings and other income data reported in the household database
were also adjusted to reflect income growth projections. Finally, health expenditures were adjusted
to reflect projections of health spending by type of service and source of payment.

The population totals were adjusted to reflect Bureau of the Census projections of population levels
by age and sex in Maryland through 2001. We also adjusted the Medicaid coverage data to reflect
federally mandated expansions in coverage for children through 2001 using the Medicaid Simulation
Module of HBSM.

We adjusted the incomes reported by individuals in the database to future years. Earnings were
adjusted based upon historical data on real growth in earnings per worker. Non-earnings income
was projected based upon the historical rate of growth in non-earnings income per person. These
growth estimates were adjusted to be consistent with national income projections provided by the
Congressional Budget Office (CBO).

Health expenditures were increased based upon projections of the growth in per-capita health
spending by type of service provided by the Health Care Financing Administration (HCFA). Using
this methodology, we estimate that health spending in Maryland will reach about $20.8 billion in
2001 (these estimates exclude expenditures for public health, research, and construction). Figure 3
presents our estimates of health spending for Maryland residents in 2001 by type of service and
source of payment.
Figure 3
Health Expenditures for Maryland Residents by Type of Service and Source of
Payment in 2001 (in millions)
Source: Lewin Group estimates based upon data provided by the Maryland Health Services cost review
commission projected to 2001 using the Maryland version of the Health Benefits Simulation Model (HBSM).
Comments?
Previous Section
Next Section
Table of Contents