The Return-on-Investment Calculator: User's Guide

The User's Guide is divided into two parts. Part I, Inputs, describes each of the elements used to calculate the state's return on investment and provides the user with instructions for entering state-specific data. In this section, on the left side of the screen, the user enters data that is then used to calculate outputs. Part II, Outputs, explains the calculations performed to determine the varied costs and benefits to the state. These outputs appear in the orange and gray boxes on the right side of the calculator. The user can also get a more "detailed financial analysis" by clicking on the link in the Cost Summary Box.

PART I: INPUTS AND ASSUMPTIONS

The calculator allows the user to choose either to enter "Basic Inputs" or "Expanded Inputs." For Basic Inputs, the user chooses a state from the pull-down menu, enters the current number of workers in the state's personal assistance services program, and the proposed wage increase. The federal match rate and current wage are entered by the calculator.

The numbers below correspond to the numbered elements on the Expanded Inputs screen (to reach this screen, click on "see Expanded Inputs" in the Gray Arrow).

1. State: Must be entered by the user. Choose a state to analyze from the pull-down menu.

The calculator uses state specific data for many of its computations. This data includes the state's federal Medicaid match rate, average monthly nursing home cost (used in calculating the savings from institutional diversion), income and sales tax rates, and public assistance eligibility and benefits.

2. Medicaid Match Rates

Federal Medicaid Match Rate: May be entered by the user.
The model enters the state's federal match rate for the Medicaid program. If you would like to consider alternative funding arrangements for wage increases, such as through a state-financed program, enter the appropriate value here.

State Medicaid Match Rate: May be entered by the user.
The calculator computes the percentage of the Medicaid rate paid by the state. You may override this value.

County Match Rate: Must be entered by the user.
Enter a county match rate for Medicaid spending. You may enter zero, if there is no county match rate.

3. Current Wage: May be entered by the user.
The ROI calculator computes an hourly wage for each state by using the state average hourly wage taken from Bureau of Labor Statistics figures and adjusting the national data on personal care wages found in the Current Population Survey. (For more information on how this is done see the About the Calculator tab.)

You may override the default hourly wage and enter your own value here. In states where worker wages diverge widely, for example, due to regional differences or program funding source, users may wish to use the calculator to analyze the impact of wage changes on segments of the worker pool rather than an amalgam of the workforce. (For more information, on how the starting wage affects the outcome of the increase see "Projected Retention Rate" below.)

4. Proposed Wage Increase: Must be entered by the user.
Enter a dollar per hour amount here.

5. Administrative Costs

Current Workers: May be entered by the user.
The ROI model assumes additional administrative costs of 12.65 percent of the proposed wage increase for current workers. This default value represents 7.65 percent for the employer's payroll tax expenses for Social Security/Medicare and 5 percent for internal administrative costs. You may enter a different value.

New Workers: May be entered by the user.
The tool's default value for the administrative cost of new workers is 17.65 percent of the projected wage (or 0.1765), representing 7.65 percent for the employer's payroll tax expenses for Social Security/Medicare and 10 percent for internal administrative costs. You may override this value.

The ROI model assumes a slightly higher administrative cost for new workers than for current workers to reflect the additional administrative burden involved in hiring a new employee as compared to increasing the wage of a current employee.

6. Current Number of Personal Assistance Workers: Must be entered by the user. Enter the number of personal assistance workers in your state.

7. Number of New Workers Due to Induced Demand: May be defined by the user. Based on the assumption that at higher wages new workers would choose direct-care work and fill some or all of the state's unmet consumer need, the model calculates a default value by multiplying the following three variables:

Current Number of Personal Assistance Workers: Entered by the calculator from the data the user enters in Line 6 above.

Percentage Increase in Wages: Calculated by the program from the proposed wage increase entered in line 4.

Wage Elasticity: May be entered by the user. Wage elasticity is a measure of the increase rise in demand for personal assistance services jobs due to an increase in wages. A study in San Francisco found a wage elasticity of 0.54; that is, a 54 percent rise in demand for jobs with a doubling of the wage. The model assumes a more modest wage elasticity of 0.33-that is, for each 100 percent increase in the wage, there are 33 percent more personal assistance workers. You can input a different value based on local conditions, including the extensiveness and maturity of the state's HCBS program and the degree to which there are waiting lists for services and/or vacancies in direct-care positions.

You can override the default value by entering a value in Line 7. For example, if program spending is capped and the state cannot hire any more personal assistance workers, this can be set to zero.

8. Annual Savings from Institutional Diversion:

The model assumes that the state will realize some savings from the fact that the new workers brought in by the higher wage will allow some consumers currently being served in nursing homes to be served in the community. These savings, which appear under Outputs, are computed by multiplying the following variables:

Savings from Institutional Diversion = (Number of New Personal Assistance Workers to Serve Unmet Need) x (Clients Served in the Community per New Worker) x (Percent Diverted from Institutions) x (Average Monthly Nursing Home Cost) x (12 months)

The variables that are used to make this calculation are discussed below.

Number of New Personal Assistance Workers: Entered by the calculator from Line 7 above.

Average Number of Clients Served in the Community Per New Worker: Entered by the calculator.

The calculator assumes that each new worker is employed on average 29 hours per week. This would enable each new personal assistance worker, on average, to serve 1.45 clients eligible for fours hours of service per day five times per week, based on a minimum level of service necessary to adequately serve a consumer with disabilities who would otherwise require nursing facility care.

Percentage of Clients Diverted from Institutions: May be entered by the user.

This variable estimates the percentage of consumers who would shift from nursing facility placement to home- or community-based setting given new home- and community-based capacity. The default value assumes a diversion rate of 20%, based on a study in Washington State, which found that each new home care client served in the community reduced nursing home use by 10 to 30%. You may alter this assumption based on the extensiveness and maturity of your state's home- and community-based infrastructure.

Average Monthly Nursing Home Cost: May be entered by the user.

The default value is based on the state's 2001 Medicaid average daily nursing home rate as reported by Wichita State University Department of Public Health Sciences' Nursing Facility Reimbursement Study (2002). You may adjust this number.

9. Savings from Reduced Turnover
To compute savings from reduced turnover, the calculator multiplies the turnover costs per worker by the number of persons by which turnover is reduced. To determine the number of persons by which turnover is reduced, the calculator uses the current and projected retention rates (see formula below under "Part II: Outputs").

To make the calculation, the model uses the following information:

Turnover Cost Per Worker: May be entered by the user.
The research on turnover costs shows a significant variation in estimates depending on how these costs are defined and accounted for. The default value estimates turnover costs of $1500 per worker, which is at the low end of the reported range.4 You may enter a different rate based on estimates from state providers.

Current Retention Rate: May be entered by the user.
There is a large literature on personal assistance worker turnover rates, with estimates ranging from 25 percent to over 100 percent per year. As a baseline retention rate, the model assumes 50 percent retention based on a review of this literature. You may adjust this rate if more accurate data is available for the state.

Projected Retention Rate: Calculated by the program (see explanation under Outputs, below).

Turnover Elasticity: May be entered by the user.
An estimate of elasticity is used to calculate projected retention rates and turnover costs. A study in Wyoming found that a 50 percent wage increase lead to a 33 percent reduction in turnover, but a study in San Francisco found that a 100 percent or doubling of wages leads to only a 30 percent decline in turnover. For its default value, the calculator assumes a middle ground -- that is, each doubling of wages reduces turnover by 35 percent. You may enter a different value.

Current Number of Workers: Entered by calculator from Line 6 above.

CALCULATE
After you have completed entering all the necessary data, click calculate at the bottom left side of the screen.

PART II: OUTPUTS

The calculator computes the "outputs," or costs and benefits associated with a wage increase for personal assistance service workers, and displays them on the right side of the screen. A Cost Summary appears in red box at the top right corner; a breakdown of benefits to the state appears in the gray box below.

Cost Summary
This box shows the overall cost of a wage increase, the net cost to the state, and the effective federal match. For a more detailed analysis of state costs, click on "See detailed financial analysis." (These outputs are explained below, under the section Detailed Financial Analysis.)

Total Cost: The total cost of increased wages for existing workers plus wages for new workers serving clients not previously served in the community.

State Cost: The net cost of a wage increase for the state, after taking into account the federal Medicaid match as well as additional savings in institutional diversion, reduced use of TANF and Medicaid, and increases in state revenues.

Effective Match: The effective match rate is the percentage of the state's cost covered by the federal match once all the savings are taken into account. It is calculated as follows:

Effective Match Rate after Economic Benefits = [(Total New Wage Costs for Existing and New Workers - Total Net Costs to the State Medicaid Program)] / (Total New Wage Costs for Existing and New Workers)

Benefits
This section details benefits to the state in three areas: workforce impacts, benefits to the state economy overall, and social benefits such as reductions in poverty.

Workforce Impacts

Projected Number of Personal Assistance Workers: The ROI calculator adds the current number of workers plus the number of new workers serving previously unmet demand.

Projected Wage: The ROI calculator adds the current wage plus the hourly wage increase.

Percentage Increase in Wages: The calculator computes the hourly wage increase as a percentage of the baseline average hourly wage.

Projected Retention Rate: The ROI model assumes that an increase in wages will improve retention rates. To calculate a projected retention rate, the calculator uses a "percentage increase" in wages so that the impact of a given increase is not the dollar amount alone but also how it compares to the current wage. That is, a $2 per hour wage increase is likely to have a greater impact on workers making $6 per hour (a 33 percent increase) than workers making $16 per hour (a 12.5 percent increase). This percentage is multiplied by a "turnover elasticity" variable (defined above in Part I: Inputs and Assumptions). Thus the Projected Retention Rate is computed as follows:

Projected Retention Rate = [(Percentage Increase in Wages) x (Turnover Elasticity)] + (Current Retention Rate)

Economic Multiplier Benefits

The ROI model assumes that there will be a net increase in federal dollars coming into the state, even with the reductions in federal match payments for institutional care and public assistance. As a result of these new federal dollars coming into the state, there will be an increase in state Gross Domestic Product (GDP), created by the increase in purchases of goods and services from local businesses.

To calculate this impact, the tool uses an economic multiplier of 2.00-that is, for each $1 that comes into the state economy from outside the state's borders $2 in increased state GDP is produced. To compute the increase in state GDP, the calculator uses the following formula:

Increase to State Gross Domestic Product = [(Federal Match Payments Offsetting New Wage Costs) - (Match Payments Lost by Institutional Diversion Savings) - (Match Payments Lost by Reduced TANF Expenditures)] x (Economic Multiplier)

Social Benefits

A number of social benefits are likely to accrue from increased wages that do not directly affect state expenditures. These social benefits not only strengthen the state's HCBS infrastructure, but also offer significant improvements in quality of life for personal assistance workers, people with disabilities, and the families of both workers and consumers. For example, reduced turnover benefits consumers and their families by improving continuity of care; workers also benefit when dollars now used for recruitment can be invested in health insurance, training, or other job enhancements. The tool estimates these anticipated social benefits as follows:

Savings from Reduced Turnover Costs: As described above, savings from reduced turnover costs is calculated by multiplying the turnover costs per worker by the number of persons by which turnover is reduced. The number of persons by which turnover is reduced is estimated by multiplying the baseline number of personal assistance workers by the change in the turnover rate. The change in the turnover rate is found by subtracting the projected turnover rate from the current turnover rate. The turnover rates are the flip side of the retention rates and are calculated by subtracting these rates from 100 percent.

Savings from Reduced Turnover Costs = (Turnover Costs Per Worker) x (Current number of Personal Assistance Workers) x [(100% - Current Retention Rate) - (100% - Projected Retention Rate)]

Number of Families Leaving Poverty: The tool compares an estimated annual income of personal assistance workers before and after the wage increase and determines the number that would move from below to above the federal poverty line ($18,400 for a family of four), assuming a single parent working a 29-hour workweek (national mean number of hours worked by community-based direct service workers) at the increased wage.

Number of Children Leaving Poverty: Using the model structure described in About the Calculator , the ROI tool estimates the number of children that are likely to live in families that move from below to above the poverty line, assuming a single parent working a 29-hour workweek at the increased wage. According to the worker sample used to construct the model, the average number of children per direct-care worker family is 0.824.

Number of Facility Placements Avoided: This figure is the number of nursing facility placements that are avoided through increased home care staffing.

Number of Facility Placements Avoided = (Number of New Workers to Serve Current Unmet Need) x (Number of Clients Served in Community per New Worker) x (Percent Diverted from Institutions)

Number of New Clients Not Previously Receiving Services: The model calculates the number of new clients served in the community through increased home care staffing by multiplying the number of new workers to serve current unmet need by the average number of clients served in the community per new worker by the percentage of clients not previously served. These new clients do not include those diverted from institutions. The percentage of new clients not previously served is calculated by subtracting the percentage of clients diverted from institutions from 100 percent.

Number of New Clients Not Previously Receiving Services = (Number of New Workers to Serve Current Unmet Need) x [(Average Number of Clients Served in Community per New Worker) x (100% - Percentage of Clients Diverted from Institutions)]

Detailed State Financial Analysis
This section details the direct costs and benefits of the wage increase for the state budget.

Total Cost of Wage Increase for Existing Workers = (Current Number of Personal Assistance Workers) x (Hours Per Year Worked by Each Worker) x (Hourly wage increase)

Total Cost of Wages for New Workers = (Number of New Workers Serving Unmet Needs x (Hours Per Year Worked by Each Worker) x (Projected Wage)

Total Additional Cost of Wages for Existing and New Workers = (Total Cost of Wage Increase for Existing Workers) + (Total Cost of Wages for New Workers)

Match Payments Offsetting Increased Wages: [(Federal Medicaid Match Rate) + (County Match Rate, if any)] x (Total Cost of Wages for Existing and New Workers)

These three elements are subtotaled to produce the Net New Wage Cost to the State Medicaid Program.

Net Savings from Institutional Diversion
The calculator computes the anticipated savings from reduced use of institutional care that may result from the addition of new community-based personal assistance workers, and then subtracts any loss of federal matching funds that may result.

Total Savings from Institutional Diversion = (Number of New Workers to Serve Unmet Need) x (Clients Served in the Community per New Worker) x (Percent Diverted) x (Average Monthly Nursing Home Cost) x (12 months)

Match Payments Lost by Institutional Diversion Savings = [(Federal Medicaid Match Rate) + (County Match Rate)] x (Total Savings from Institutional Diversion)

These two elements are subtotaled to produce the Net Savings from Institutional Diversion realized by the state.

Other Medicaid Savings

Total Savings from Reduction in the Use of Medicaid for Health Insurance: Many personal assistance workers rely on Medicaid for their health insurance. As their incomes rise above income eligibility thresholds, workers and their families will leave the Medicaid health insurance program, resulting in savings to the state's Medicaid program. The tool uses a representative worker sample (see "Model Structure" in About the Calculator ) to estimate these anticipated savings. The model estimates $2,000 for adults leaving Medicaid and $1,000 for children.

Match Payments Lost by Reductions in Use of Medicaid: Should there be a decreased reliance on Medicaid for health care among personal assistance workers, states will reduce their Medicaid expenditures, resulting in a corresponding decrease in federal (and county) match payments. The tool calculates this estimated decrease in match payments:

Match Payments Lost by Reductions in Use of Medicaid = [(Federal Medicaid Match Rate) + (County Match Rate)] x (Total Savings from Reduction in Use of Medicaid)

The above figures are subtotaled to calculate Other Medicaid Savings.

Total Net Cost to State Medicaid Program: To compute the Total Net Cost to State Medicaid Program, the ROI calculator adds the above figures as follows:

Total Net Cost to State Medicaid Program = (Net New Wage Costs) + (Net Savings from Institutional Diversion) + (Other Medicaid Savings)

Other Savings to the State

Increase in State Tax Revenue: The tool estimates an anticipated increase in income and sales tax revenue that may result from workers' higher wages.

As noted in the "Model Structure" (see About the Calculator), each personal assistance worker is assigned a tax rate based on the state selected by the user, which the tool uses to compute increases in state income tax revenue from increases to personal assistance worker incomes. To calculate the sales tax revenue, the ROI model assumes that individuals in this economic class consume all of their income, so the tool multiplies the sales tax rate times the income increase.

Reduced TANF Expenditures: A number of personal assistance workers participate in TANF as do many new workers entering the workforce to serve the current unmet need for personal assistance services. As their incomes rise above income eligibility thresholds, personal assistance workers and their families will leave the TANF program, resulting in savings to the state. The tool uses a representative worker sample (as described in About the Calculator ) to estimate the impact of wage increases on state TANF expenditures:

• The tool determines whether TANF participants are still eligible to participate if they receive the new higher wage rate. If not, the model assumes that the state saves those benefits payments.
  
  
• For all new personal assistance workers the model assumes that half were previously TANF participants. The model determines whether these workers are still eligible to participate given their anticipated annual income. If not, the model assumes that the state saves those benefits payments.

Federal Match Payments Lost by Reduced TANF Expenditures: Should there be a decreased reliance on TANF among personal assistance workers, there will be a corresponding reduction in federal match payments associated with this reduction in use. The calculator computes this estimated decrease in match payments as follows:

Federal Match Payments Lost by Reduced TANF Expenditures = (Federal Match Rate) x (Savings from Reduced TANF Expenditures)

The above figures three figures -- Increase in State Tax Revenue, Reduced TANF Expenditures, and Lost Federal Match Payments -- are subtotaled to produce the estimated Other Savings to the State.