The Calculation of Costs and Benefits

For the analysis the project identified an 'average participant' who joined the programme in 1998. The 'average participant' is a statistical model whereby total figures were divided by the size of the survey population to find the average. The information on the 'average participant' was determined using previous research on the Mainstream Supported Employment Programme²¹. The total costs and benefits for the next six years were then calculated using the 'average participant' model. Costs and benefits for years 7-21 were calculated using a residual value. Calculations for the economic analysis are based on figures derived from the financial analysis.

Significant monetary costs and benefits related to the client group were identified. These were calculated using information on average payments; for example, the average payment that is made to a person on the invalid's benefit. These figures are based on the concept that they are the 'most likely value' (except where otherwise stated).

The cost-benefit analysis implicitly compares the programme to the 'null case' of not placing the average participant on the programme.

The marginal cost of taxation was based on the research of Diewert and Lawrence²² who found the marginal cost of capital taxation to be around 18%, and this figure has been used in the economic analysis of costs and benefits. It is assumed that there has been no major movement in this figure since the research.

Figures

The most up-to-date figures were used where possible. The figures for Social Welfare and ACC payment levels are based on the 1996-1997 financial year, the exception being the emergency unemployment benefit which is based on the summer of 1997-1998. Figures from the Ministry of Education are based on the projected costs for the 1998 academic year, the exception being the level of student loan write-off which is based on a 1995 figure. The cost of Mainstream administration is based on the budgeted costs for the 1997-1998 financial year. The training costs are based on the actual costs in the 1996-7 financial year and the salaries are based on the average payments made in March 1998.

Total costs include the fixed administration costs plus the variable placement costs per 'average' individual. Because of technical difficulties it is assumed in the total costs that the number of people on the programme is fixed. This is not the real situation because the number of people on the programme fluctuates for a range of reasons - for example, early absorption of the participant into the employing government agency. For this analysis, the number of participants on the programme is assumed to be 75, based on the number in the programme on one day in March 1998.

Personal income tax calculations assumed that the income earned was the only income being earned by the participants. No allowance has been made for the gains the Government would make from the reduced use of Community Service Cards or for any student loan repayments that would be made as a result of the programme. This has the effect of making the analysis conservative, and therefore understating the benefits.

Dead-weight

One of the key variables in the analysis is the number of people who would have gained work even without Mainstream. This has been termed the dead-weight. Two alternative scenarios have been used:

• The first assumes that none of the participants would have gained employment without the help of the programme. This is almost certainly inaccurate, as it is likely that some participants would have gained employment independently.
• The second more conservatively assumes that some participants would have gained employment without the help of Mainstream. This scenario contains a number of assumptions:
  
  • It is assumed that the participants are no different from the average person living on government support. That is, that the Mainstream Manager is not selective about who comes on to the programme. This is clearly not the case, because Mainstream participants must have a significant disability and be at considerable disadvantage in gaining employment.
  • It is assumed that the rate at which people would leave employment does not change for the first six years. However, in the NPV calculations for the residual value from year 7 no further increase in dead-weight is allowed for. That is, if participants would not have gained employment independently by year 7, it is assumed that they would never gain employment independently. In reality, the rate at which people gain employment independently decreases over time.
  • It is assumed in the dead-weight calculations that participants who independently gain employment never leave employment. This is very conservative.

The vast majority of participants in Mainstream have not been in employment for several years. The dead-weight figures are based on the rate at which people who have been on a form of income support (either Social Welfare or ACC) for one year leave to go into employment in the next 12 months. Where exact values are not known, the best estimate available from the government agencies concerned was used. There was no information for two categories:

• participants who had previously been in tertiary education, and
• participants who had not been on any form of government support (for example, they had a working spouse).

For these examples the calculation conservatively uses the highest dead-weight figure from the other categories. This was the unemployment benefit with 19.3% dead-weight.

As shown in Table 3, calculation of the level of dead-weight was achieved by first identifying the participants' source or sources of support before starting Mainstream. These were then multiplied by the average rate at which people (in the general population), who have been receiving these supports for a period of at least one year, leave to enter employment in the next year. This resulted in a weighted value of 12.87% for the dead-weight calculation.

Table 3: Dead-weight Calculations

For the financial analysis, different dead-weight levels were tested. This was done to determine the level of dead-weight which resulted in the NPV result no longer being positive.

Future Employment

Future employment levels were based on the research on the programme carried out in 1997²³, which found that 13.33% of former participants were not in employment but were on benefits. This result needs to be treated with caution, as not all of the questionnaires were returned. Although the study included participants who had been on the programme over many years, those who had been on the programme more recently were over-represented in the analysis because they were easier to contact.

It is assumed that the numbers both in and out of employment remain static. That is, that the number leaving the workforce is balanced by the number returning. This is likely to be an over-estimation of the number of former participants in employment, because those who did not gain employment with the programme probably responded to the questionnaire at a lower rate.

For the financial analysis, different future levels of unemployment were tested. This was done to determine the level of unemployment, after participation in Mainstream, which would result in the NPV result no longer being positive. Of participants who completed the programme in the 1997/98 year, 73% gained permanent employment.

Duration of Subsidy Payment

It was assumed in the analysis that all participants remained on the subsidy for the full two years. However, some participants leave the programme early without gaining employment, and other participants gain (unsubsidised) employment before the two years are completed.

Net Present Value (NPV)

The discount rate for the NPV was 11%. This was based on the 1997/98 capital charge rate. For the NPV, a residual value for the years 7 to 21 was used. Twenty-one years is the average length of time that present Mainstream participants have until turning 60.

Internal Rate of Return

This was only worked out for the financial analyses as it cannot be calculated without negative values. It includes the residual value calculated for the NPV.

Break-even Point

Also called the pay-back period, this states the number of years before the programme pays for itself. It does not include the discount rate.

Inflation

The effects of inflation have been ignored. The nominal discount rate, which has an inflation component, may have been overstated. However, around 11% is not inconsistent with the 1972 Cabinet-approved real discount rates of 10%, or with Wilkinson's discount rate estimates (Treasury, 1982).

Wage Increases

It is conservatively assumed that the participants will not receive either promotion or any wage increases after the end of the two-year wage subsidy.

Value of Workforce Participation

In the economic analysis the value of workforce participation is included. It is assumed that labour is scarce, so increased employment of Mainstream participants represents an increase in aggregate labour supply. However, it can be argued that the productivity of Mainstream participants is less than the wage paid during the subsidised period, due to factors such as increased training requirements. To test the sensitivity of this assumption the workforce participation has been valued at different levels. For Spreadsheets 1 and 2 the workforce participation in the economic analysis was valued as being equal to the salary. In Spreadsheet 3 the workforce participation was valued as being equal to the salary minus 50% of the salary subsidy.

²¹ State Services Commission (1997) op. cit.

²² Diewert and Lawrence (1994) op. cit.

²³ State Services Commission (1997) op. cit.