Joyce et al. (2002) examined how a series of changes in cost sharing for prescription drugs
affects total expenditures for pharmaceuticals—specifically, moving from a one- to a two- or a
two- to a three-tier plan, doubling copayments in each tier, or requiring generic substitution.
This study estimated copayment elasticities that ranged from –0.22 (from doubling copayments
in a single-tier plan) to –0.40 (from doubling copayments in a three-tier plan). Adding a second
copayment tier ($20 in addition to $10) reduced average spending by 19 percent, and adding a
third level ($30) reduced average spending by an additional 4 percent.
substitution in a two-tier plan reduced drug spending by 8 percent.
A few studies have estimated the elasticity of demand for drugs in specific therapeutic
classes. Using claims data from 30 employers with 52 health plans between 1997 and 2000,
Goldman et al. (2004) examined the effect of doubling the copayment for eight classes of
therapeutic drugs. They found reduced utilization in all eight classes, ranging from 45 percent
(for nonsteroidal anti-inflammatory drugs) to 25 percent (for anti-diabetic drugs). Lansman et al.
(2005) studied the effect of introducing three-tier copayments and reached similar conclusions.
They estimated a generally low copayment elasticity (-0.16 to –0.10) for drugs to treat on-going
asymptomatic conditions, and a moderate copayment elasticity (-0.60 to –0.24) for drugs to treat
Certain subpopulations (such as people with chronic illness, the low-income, and the
elderly) are likely to have different price elasticity of demand for prescription drugs. This
section summarizes the limited evidence for these subpopulations.
Persons with Chronic Illness. In general, people with chronic illness are expected to be
sponsored (respectively) by 30 employers, Goldman et al. (2004) found that the use of disease-