V401 Case 2: An Economic Evaluation of “Safer Choices”

Overview

School programs to reduce unprotected sexual intercourse have been implemented across the United States, to reduce human immunodeficiency virus (HIV) infection, infection by other sexually transmitted diseases (STDs), and unintended pregnancy among US adolescents. Program evaluations have shown that such programs are effective in reducing unprotected sexual intercourse, substantially increasing condom use and other forms of contraceptive use among sexually active young people. Because resources to fund school-based HIV, other STDs, and pregnancy prevention programs are limited, however, program effectiveness is not sufficient to justify program implementation. Issues of practical concern to policy makers are financing (whether prevention programs are affordable) and whether the benefits of such programs exceed their costs.

Few economic studies of school programs have been conducted. Most studies of HIV prevention programs in particular have focused on those targeting intravenous drug users, adult urban women, and adult or adolescent gay and bisexual men. Moreover, few studies have assessed the reduction of STDs other than HIV, or have looked at the benefits and costs of preventingunintended teenage pregnancy. This case tries to get a better sense of the benefits and costs of a program to reduce the amount of unprotected sex, by monetizing the multiple benefits such programs can offer.

This topic is particularly policy relevant since sexually active people between the ages of 15 and 24 accounted for 50% of all new cases of sexually transmitted diseases in the United States. Researchers found about 18.9 million new STD cases occurred in 2000, for example, and 9.1 million of the cases, or roughly 50%, occurred among people between the ages of 15 and 24. This age group also accounts for the majority of unintended pregnancies, and a significant fraction of new HIV infections.

The Data and NPV Computations

Tables 1 and 2 give the basic economic data needed to compute probability distributions for the net present value of the “Safer Choices” program. The question to be answered is whether it is probable that the costs of the program are less than the expected benefits, and thus, whether such a program should be instituted widely by local school districts, assuming the financial resources are available to implement such a program. We will return to financing issues in the next section.
Costs include the usual inputs of time and materials needed to institute this kind of program. The benefits fall into the following categories:
• reduced medical costs for treating STDs
• reduced medical costs for dealing with unintended pregnancies
• increased work productivity from reducing the incidence of HIV infection
• Reduced medical costs associated with reducing the incidence of HIV infection. See Table 2 for the particular details.
Not included on this list is the quality-of-life improvement received by someone who otherwise would have contracted STDs, had an unintended pregnancy, or contracted HIV. Since it is hard to monetize the benefit “quality of life improvement,” we leave it out of the quantitative assessment. However, such a benefit might affect your decision about recommending the program, even though it cannot be explicitly considered in the quantitative analysis.
Similarly, the “value of live” is not monetized in this study, for two reasons. First, the assumption that a person who now contracts HIV will live close to a normal life span, given dramatic improvements in the efficacy of antiretroviral drugs. In this particular case, we assume that life spans are 10 years shorter for the person who contracts HIV. Secondly, monetizing the value of life is a controversial issue in cost benefit analysis. But note that by not monetizing the value of life, the approach taken here, a downward bias is imparted to our estimate of the economic value of the program – that is, our estimate of the program’s benefits will be conservative – because extending lifespans 10 years definitely has value. This issue can also be considered qualitatively in your assessment of the program.
Overall, the question is what is the probability that the net economic value of Safer Choices is positive, given the data in Tables 1 and 2? Given the information you have, can you recommend with reasonable confidence that Safer Choices provides positive net-benefits?

Distributional and Financial Analysis
We now consider how Safer Choices might be financed. One option would be to have the taxpayer finance the program. KHT1 below shows the distributional effects of Safer Choices when taxpayers finance the program. The table reflects the fact that insurance companies receive the fraction “a” of medical cost savings on the assumption that insurance companies would otherwise have paid the fraction “a” of these costs. The student beneficiary of the program receives the fraction (1-a) of the benefits – effectively, the value of the savings in copayments they would otherwise have had to have been paid had the program not benefited the student in one of the possible ways it might. On the assumptions that (a) the cost of the program is completely covered by taxpayers and (b) the employers gain in productivity just equal the wages they would pay out, we can see that Safer Choices benefits two groups: (a) the students who use it and benefit from it: they gain a lifetime stream of co-payment savings, plus additional wages from additional work and (b) insurance companies, who get a lifetime stream of savings in reimbursements for medical expenses. To get a sense of the magnitude of the benefits to these two groups, numerically specify KHT1 based on the mean npv estimates for the stakeholders in question. Choose the means from either the college educated or non college educated npv distributions.
A second KHT should be specified for an alternative financing arrangement whereby insurance companies pay schools to adopt “Safer Choices” (see KHT2). Again choose means from either college educated or non-college educated distributions to specify the KHT
Note that if the financial returns to health insurance companies from avoided future reimbursements are higher than the subsidy they would have to pay to school districts to institute safer choices, it would be in their interest to do so. It is for this reason that many employer-provided insurance program, such as the TIAA-CREF programs which cover IU faculty, reimburse preventive care at 100% in the hopes that people will live healthier lifestyles — reducing health insurance payouts later down the road. In the same way, it might make sense for insurance companies to pay schools to encourage students to lead healthier lifestyles.
Note: you can assume that “a”= .8, and therefore that “1-a”= .2 In short, students affected by the program (or their parents) would normally make a 20% copay, with health insurance companies covering 80%.
Case Outputs
The basic output is a no longer than a 3 page double-space memorandum (this page length EXCLUDES ANY TABLES YOU MIGHT INCLUDE) that describes the analysis, presents the results, and then makes a recommendation whether or not Safer Choices is a program local school districts should promote, with also a recommendation about the financing mechanism. As with the pervious case, your memo should be broken down explicitly into sections with bolded headers as follows:

Introduction
Analysis Method and Assumptions
Results
Policy Recommendation

Your assessment should be based on four figures/tables.
Figure 1 shows the NPV distribution of the program for non-college educated students, with the summary statistics @Risk provides (mean, standard deviation, minium, maximum, etc.) Also, do the overlay that shows the kind of distribution generated.
Figure 2 shows the NPV distribution of the program forcollege-educated students, with the summary statistics @Risk provides (mean, standard deviation, minimum, maximum, etc.) Also, do the overlay that shows the kind of distribution generated.
Table 3 is KHT 1 based on the mean estimated from the distributions in either Table 1 or Table 2.
Table 4 is KHT2 based on the means estimated from the distribution you chose for KHT1.That is, do either non-college educated or college-educated for both the KHTs.

Ground Rules
As in The previous case, you are encouraged to form working groups and collaborate on the analysis. But you must write up the memo yourself.

Let me know if you have any questions.

Table 1: Program costs per 1000 students in initial period (period zero)
Cost Category Cost
Teacher training 59000
Teaching 36153
Peer facilitators 53166
Site coordinator training 48913
Site coordination 12121
Curriculum packages 8000
Implementation manuals 1063
Activity kits 1250
Photocopies for students 58
Photocopies for teachers 64
Videos 600
Total 220388
Table 2: Program Benefits

Cost Category Value and Time Period Cases Avoided per 1000 students
Avoided medical cost of STD treatment (a weighted average of bacterial STDs (Chlamydia, Gonorrhea, PID) and viral STDs such as Herpes
Lognormal distribution
Mean $75 per case avoided,
Sigma=$5

realized every year for 51 years, starting in program year (period 0)
PERT
Low=0
Medium=12
High =15
Avoided cost of one unwanted pregnancy (a weighted combination of cost of live birth, prenatal care, or pregnancy termination)
Lognormal distribution
Mean=$5,200 per avoided pregnancy,
Sigma=$900

realized in the program year, and the following year (Periods 0 and 1)
PERT
Low=0,
Medium=2
High=6
Value of work productivity gain from avoiding HIV — non college educated
Normal distribution:
Mean=$14.00 per hour
Sigma=$2.00 per hour

2000 hours per year over 40 years starting in year after program commences (period 1)
Lognormal
Mean=2
Sigma=.4
Value of work productivity gain from avoiding HIV — college bound
Normal distribution
Mean= $18.00/hour
Sigma=$3/hour

2000 hours per year over 40 years starting in the 4th year after program commences, for 36 years (starting in period 4)

Lognormal
Mean=2
Sigma=.4
Cost estimate of value of medical cost savings from avoiding HIV (savings in antiretroviral drugs and other treatments)
Normal distribution
Mean=$15,000 per case avoided per year
Sigma=$1000

Realized each year for 40 years
Lognormal
Mean=2
Sigma=.4
KHT1-Taxpayer financed Students School Administrators Employers Tax Payers Insurance Company Net
Avoided medical costs of STDs (1-a) B1 aB1 B1
Avoided medical costs of unintended pregnancy (1-a) B2 aB2 B2
Avoided Medical costs of HIV (1-a) B3 aB3 B3
Benefits of improved quality of life/ length of life
Worker Productivity Gain from reduced HIV incidence B5 B5
Program Cost -C1 -C1
T1 -T1 0
T2 -T2 0
Net (1-a)[B1+B2+B3]
+T2 T1-C1=0 B5-T2=0 -T1 a[B1+B2+B3] B1+B2+B3+B5-C1

KHT2-Insurance Company Financed Students School Administrators Employers Insurance Company Net
Avoided medical costs of STDs (1-a) B1 aB1 B1
Avoided medical costs of unintended pregnancy (1-a) B2 aB2 B2
Avoided Medical costs of HIV (1-a) B3 aB3 B3
Benefits of improved quality of life/ length of life
Worker Productivity Gain from reduced HIV incidence B5 B5
Program Cost -C1 -C1
T1 -T1 0
T2 -T2 0
Net (1-a)[B1+B2+B3]
+T2 T1-C1=0 B5-T2=0 a[B1+B2+B3]-T1 B1+B2+B3+B5-C1

V401 Case 2: An Economic Evaluation of “Safer Choices”

Overview

School programs to reduce unprotected sexual intercourse have been implemented across the United States, to reduce human immunodeficiency virus (HIV) infection, infection by other sexually transmitted diseases (STDs), and unintended pregnancy among US adolescents. Program evaluations have shown that such programs are effective in reducing unprotected sexual intercourse, substantially increasing condom use and other forms of contraceptive use among sexually active young people. Because resources to fund school-based HIV, other STDs, and pregnancy prevention programs are limited, however, program effectiveness is not sufficient to justify program implementation. Issues of practical concern to policy makers are financing (whether prevention programs are affordable) and whether the benefits of such programs exceed their costs.

Few economic studies of school programs have been conducted. Most studies of HIV prevention programs in particular have focused on those targeting intravenous drug users, adult urban women, and adult or adolescent gay and bisexual men. Moreover, few studies have assessed the reduction of STDs other than HIV, or have looked at the benefits and costs of preventingunintended teenage pregnancy. This case tries to get a better sense of the benefits and costs of a program to reduce the amount of unprotected sex, by monetizing the multiple benefits such programs can offer.

This topic is particularly policy relevant since sexually active people between the ages of 15 and 24 accounted for 50% of all new cases of sexually transmitted diseases in the United States. Researchers found about 18.9 million new STD cases occurred in 2000, for example, and 9.1 million of the cases, or roughly 50%, occurred among people between the ages of 15 and 24. This age group also accounts for the majority of unintended pregnancies, and a significant fraction of new HIV infections.

The Data and NPV Computations

Tables 1 and 2 give the basic economic data needed to compute probability distributions for the net present value of the “Safer Choices” program. The question to be answered is whether it is probable that the costs of the program are less than the expected benefits, and thus, whether such a program should be instituted widely by local school districts, assuming the financial resources are available to implement such a program. We will return to financing issues in the next section.
Costs include the usual inputs of time and materials needed to institute this kind of program. The benefits fall into the following categories:
• reduced medical costs for treating STDs
• reduced medical costs for dealing with unintended pregnancies
• increased work productivity from reducing the incidence of HIV infection
• Reduced medical costs associated with reducing the incidence of HIV infection. See Table 2 for the particular details.
Not included on this list is the quality-of-life improvement received by someone who otherwise would have contracted STDs, had an unintended pregnancy, or contracted HIV. Since it is hard to monetize the benefit “quality of life improvement,” we leave it out of the quantitative assessment. However, such a benefit might affect your decision about recommending the program, even though it cannot be explicitly considered in the quantitative analysis.
Similarly, the “value of live” is not monetized in this study, for two reasons. First, the assumption that a person who now contracts HIV will live close to a normal life span, given dramatic improvements in the efficacy of antiretroviral drugs. In this particular case, we assume that life spans are 10 years shorter for the person who contracts HIV. Secondly, monetizing the value of life is a controversial issue in cost benefit analysis. But note that by not monetizing the value of life, the approach taken here, a downward bias is imparted to our estimate of the economic value of the program – that is, our estimate of the program’s benefits will be conservative – because extending lifespans 10 years definitely has value. This issue can also be considered qualitatively in your assessment of the program.
Overall, the question is what is the probability that the net economic value of Safer Choices is positive, given the data in Tables 1 and 2? Given the information you have, can you recommend with reasonable confidence that Safer Choices provides positive net-benefits?

Distributional and Financial Analysis
We now consider how Safer Choices might be financed. One option would be to have the taxpayer finance the program. KHT1 below shows the distributional effects of Safer Choices when taxpayers finance the program. The table reflects the fact that insurance companies receive the fraction “a” of medical cost savings on the assumption that insurance companies would otherwise have paid the fraction “a” of these costs. The student beneficiary of the program receives the fraction (1-a) of the benefits – effectively, the value of the savings in copayments they would otherwise have had to have been paid had the program not benefited the student in one of the possible ways it might. On the assumptions that (a) the cost of the program is completely covered by taxpayers and (b) the employers gain in productivity just equal the wages they would pay out, we can see that Safer Choices benefits two groups: (a) the students who use it and benefit from it: they gain a lifetime stream of co-payment savings, plus additional wages from additional work and (b) insurance companies, who get a lifetime stream of savings in reimbursements for medical expenses. To get a sense of the magnitude of the benefits to these two groups, numerically specify KHT1 based on the mean npv estimates for the stakeholders in question. Choose the means from either the college educated or non college educated npv distributions.
A second KHT should be specified for an alternative financing arrangement whereby insurance companies pay schools to adopt “Safer Choices” (see KHT2). Again choose means from either college educated or non-college educated distributions to specify the KHT
Note that if the financial returns to health insurance companies from avoided future reimbursements are higher than the subsidy they would have to pay to school districts to institute safer choices, it would be in their interest to do so. It is for this reason that many employer-provided insurance program, such as the TIAA-CREF programs which cover IU faculty, reimburse preventive care at 100% in the hopes that people will live healthier lifestyles — reducing health insurance payouts later down the road. In the same way, it might make sense for insurance companies to pay schools to encourage students to lead healthier lifestyles.
Note: you can assume that “a”= .8, and therefore that “1-a”= .2 In short, students affected by the program (or their parents) would normally make a 20% copay, with health insurance companies covering 80%.
Case Outputs
The basic output is a no longer than a 3 page double-space memorandum (this page length EXCLUDES ANY TABLES YOU MIGHT INCLUDE) that describes the analysis, presents the results, and then makes a recommendation whether or not Safer Choices is a program local school districts should promote, with also a recommendation about the financing mechanism. As with the pervious case, your memo should be broken down explicitly into sections with bolded headers as follows:

Introduction
Analysis Method and Assumptions
Results
Policy Recommendation

Your assessment should be based on four figures/tables.
Figure 1 shows the NPV distribution of the program for non-college educated students, with the summary statistics @Risk provides (mean, standard deviation, minium, maximum, etc.) Also, do the overlay that shows the kind of distribution generated.
Figure 2 shows the NPV distribution of the program forcollege-educated students, with the summary statistics @Risk provides (mean, standard deviation, minimum, maximum, etc.) Also, do the overlay that shows the kind of distribution generated.
Table 3 is KHT 1 based on the mean estimated from the distributions in either Table 1 or Table 2.
Table 4 is KHT2 based on the means estimated from the distribution you chose for KHT1.That is, do either non-college educated or college-educated for both the KHTs.

Ground Rules
As in The previous case, you are encouraged to form working groups and collaborate on the analysis. But you must write up the memo yourself.

Let me know if you have any questions.

Table 1: Program costs per 1000 students in initial period (period zero)
Cost Category Cost
Teacher training 59000
Teaching 36153
Peer facilitators 53166
Site coordinator training 48913
Site coordination 12121
Curriculum packages 8000
Implementation manuals 1063
Activity kits 1250
Photocopies for students 58
Photocopies for teachers 64
Videos 600
Total 220388
Table 2: Program Benefits

Cost Category Value and Time Period Cases Avoided per 1000 students
Avoided medical cost of STD treatment (a weighted average of bacterial STDs (Chlamydia, Gonorrhea, PID) and viral STDs such as Herpes
Lognormal distribution
Mean $75 per case avoided,
Sigma=$5

realized every year for 51 years, starting in program year (period 0)
PERT
Low=0
Medium=12
High =15
Avoided cost of one unwanted pregnancy (a weighted combination of cost of live birth, prenatal care, or pregnancy termination)
Lognormal distribution
Mean=$5,200 per avoided pregnancy,
Sigma=$900

realized in the program year, and the following year (Periods 0 and 1)
PERT
Low=0,
Medium=2
High=6
Value of work productivity gain from avoiding HIV — non college educated
Normal distribution:
Mean=$14.00 per hour
Sigma=$2.00 per hour

2000 hours per year over 40 years starting in year after program commences (period 1)
Lognormal
Mean=2
Sigma=.4
Value of work productivity gain from avoiding HIV — college bound
Normal distribution
Mean= $18.00/hour
Sigma=$3/hour

2000 hours per year over 40 years starting in the 4th year after program commences, for 36 years (starting in period 4)

Lognormal
Mean=2
Sigma=.4
Cost estimate of value of medical cost savings from avoiding HIV (savings in antiretroviral drugs and other treatments)
Normal distribution
Mean=$15,000 per case avoided per year
Sigma=$1000

Realized each year for 40 years
Lognormal
Mean=2
Sigma=.4
KHT1-Taxpayer financed Students School Administrators Employers Tax Payers Insurance Company Net
Avoided medical costs of STDs (1-a) B1 aB1 B1
Avoided medical costs of unintended pregnancy (1-a) B2 aB2 B2
Avoided Medical costs of HIV (1-a) B3 aB3 B3
Benefits of improved quality of life/ length of life
Worker Productivity Gain from reduced HIV incidence B5 B5
Program Cost -C1 -C1
T1 -T1 0
T2 -T2 0
Net (1-a)[B1+B2+B3]
+T2 T1-C1=0 B5-T2=0 -T1 a[B1+B2+B3] B1+B2+B3+B5-C1

KHT2-Insurance Company Financed Students School Administrators Employers Insurance Company Net
Avoided medical costs of STDs (1-a) B1 aB1 B1
Avoided medical costs of unintended pregnancy (1-a) B2 aB2 B2
Avoided Medical costs of HIV (1-a) B3 aB3 B3
Benefits of improved quality of life/ length of life
Worker Productivity Gain from reduced HIV incidence B5 B5
Program Cost -C1 -C1
T1 -T1 0
T2 -T2 0
Net (1-a)[B1+B2+B3]
+T2 T1-C1=0 B5-T2=0 a[B1+B2+B3]-T1 B1+B2+B3+B5-C1