Labour Market Programs 1995-96: Interpreting the Outcomes
Steve O'Neill
Economics, Commerce and Industrial Relations Group
Contents
Introduction
Costs and commencements, 1995-96
Table 1: 1994-95/1995-96 Comparison of Commencements,
Outcomes, the Cost of Key Labour Market Program and Unit Costs
Additional Information on cost per placement
Table 2: Estimated Costs of Major DEETYA Labour Market
Programs
SkillShare unit costs: Access Economics and DEETYA
costings
Table 3: Comparisons of Access Economics Calculations
with DEETYA's Revised Estimates
Overview
Endnotes
APPENDIX A - Employment programs
APPENDIX B - Training and Assistance Programs
List of Tables
Table 1: 1994-95/1995-96 Comparison of Commencements,
Outcomes, the Cost of Key Labour Market Program and Unit Costs
Table 2: Estimated Costs of Major DEETYA Labour Market
Programs
Table 3: Comparisons of Access Economic Calculations
with DEETYA's revised estimates
This paper updates and supplements an earlier Current Issues Brief, Labour
Market Programs in 1995: Overview and Assessment (Current Issues Brief
No.4 1996-97). That paper contains additional information on labour
market programs and the two might be usefully read in conjunction with
each other.
The purpose of this paper is to examine and interpret analyses of the
reported costs and usage of labour market programs (LMPs) and to evaluate
their effectiveness in assisting participants to find work opportunities.
At the time of the publication of the earlier paper, LMP data for the
1995-96 financial year had not been made available. The Annual Report
(1995-96) of the Department of Employment, Education, Training and Youth
Affairs (DEETYA) now provides a better picture of the use and costs of
such programs for 1995-96 which, in the main, are delivered under DEETYA's
Program 4.
The question of whether LMPs offer value for money in being able to place
participants in unsubsidised employment has been developed in a number
of studies. In July 1996, the consultancy organisation, Access Economics,
reviewed the costs and returns to the Budget of selected LMP in a commissioned
study. It found that over time, using certain assumptions, two LMP (SkillShare
and JobClubs), provided a gain to the Budget.(1) This paper reviews the
data and assumptions used in the calculations of DEETYA and Access Economics,
over the net gain or cost to the Budget of LMP.
The Minister for Employment Education and Training, Senator the Hon.
A. Vanstone, refuted the assumptions used by Access Economics, arguing
instead that even those two LMP incur a cost to the Budget.(2) Further
information on the method used to establish these costs has been provided
by DEETYA to the Senate's Employment, Education and Training Legislation
Committee (the Senate Committee) and this information is reviewed in this
paper.
It is also important to note that the (August) 1996 Budget proposed a
radical review of both programs and their delivery. These developments
are not covered here. An earlier PRS publication, Budget Review 1996-97
contains a concise overview of the Budget's impact on DEETYA's programs
(as well as discussing other portfolios). Some of the LMP reviewed below
are to terminate; others will continue; new employment placement enterprises
will seek employment opportunities for those of the unemployed referred
to them and the major public employment service, the Commonwealth Employment
Service, is to be operated as a corporatised agency. A transition period
to the end of 1997 has been allowed for the total transformation to take
place
DEETYA reports that LMP commencements increased from 555 000 to 699 000
(1994-95 to 1995-96). As well, program costs (ie, the sums received by
participants excluding costs for their delivery) increased from $1.62
billion to $2.154 billion. Full titles and a brief explanation of each
LMP included in Table 1 are provided in the appendix at the end of the
paper.
As can be seen from Table 1 there has been an increase in costs for the
main LMPs. Data for the New Work Opportunities (NWO) program shows costs
for this program to be the highest of any of the LMP - $498 million allowing
the funding of 49 000 commencements. The increase in the number of NWO
commencements appears to account for about 80 per cent of the increase
in DEETYA's program costs.
Commencements Positive Outcome:%(a) Program Costs:$M UnitCost:$
EMPLOYMENT 1995-96 1994-95 1995-96 1994-95 1995-96 1994-95 (1993-94)(b)
JobStart 101 000 95 212 56.3 58.7 236 217.6 1 124
NTW 33 000 7 953 n/a n/a 63 6.7 --
JobSkills 27 000 20 456 37.8 45.9 273 191.9 6 126
NWO 49 000 10 868 24.8 n/a 498 89.1 10 009 c
LEAP 13 000 14 930 33.7 40.0 89.0 93.7 6 241
TRAINING
JobTrain 93 000 90 519 42.7 41.0 165 169.9 776
Spec 90 000 69 948 39.4 40.5 194 164.6 1 096
Intvntn
ATY 1 700 4 692 41.2 37.9 9.2 25.9 3 209
Job Clubs 45 000 44 817 43.8 45.8 30.0 30.6 669
SkillShare 165 000 139 482 47.0 45.4 182 209.5 961
MAS 49 000 13 667 84.9 83.8 20.0 10.6 1 357
CAP ceased 998 -- -- n/a 0.37 --
TAP 11 900 13 667 50.5 45 37.0 44.9 6 150
NEIS 12 000 5 914 84.0 84.0 104 80.7 --
Other 8 400 19 500 36.9 34.8 -- -- 1 518
Sources: DEETYA , Annual Report 1995-96. Unit cost data
can be found in Additional Information Received by the Senate's
Employment, Education and Training Committee (Vol.2, October 1996) and
in Hansard (House of Representatives) 25 August 1995 in answer
to Question No. 2044. Where a unit cost for a program has been revised
since August 1995, the revised figure has been used. Other
programs include contracted placements, apprenticeships and work experience
for people with disabilities.
(a) Positive Outcomes. This is the proportion of participants
who having gone through the program were in unsubsidised jobs or, while
not employed were participating in a non-DEETYA education or training
place
(b) Unit Costs. This data has been slightly revised from that
provided by the Minister in answer to Question No. 2044 (Hansard,
29 August 1995). According to DEETYA, net unit costs are equal to the
average gross unit cost of assistance minus the average savings in JobSearch
and NewStart Allowances (JSA/NSA) payments accruing to the Department
of Social Security (DSS) as a result of jobseekers participating in
a program.
(c) NWO. Unit costs are for 1994-95
Under the former Government's Working Nation strategy, NWO was
to be offered as a public employment facility to 'Job Compact' clients
where the prospect of alternative private sector employment was extremely
limited and therefore clients would be hard to place in employment. DEETYA's
latest annual report has confirmed these assumptions by showing that the
proportion of NWO places taken by long-term unemployed to be 99.8 per
cent, with such people being unemployed for more than 18 months.
The increase in NWO commencements in 1995-96 is earlier than the forecast
of Working Nation which expected this number of commencements in
1996-97. Otherwise, program costs appear to be marginally above last year's
outcome allowing for the higher number of commencements. Positive outcomes
(ie, an unsubsidised job or training placement 3 months after DEETYA assistance
terminated) appear to have declined for the employment programs, while
positive outcomes for the training programs appear to have improved slightly.
As noted in Labour Market Programs in 1995: Overview and Assessment,
a DEETYA report on LMP earlier this year gave mixed blessings to the Working
Nation effort to reduce long-term unemployment.(3) Information supplied
by the Minister during the Budget provided more detailed analysis of cost
of securing an LMP participant in employment.(4) There has been further
explanation provided by DEETYA to the Senate Committee on the concepts
of 'net impact' and 'cost per net impact'.(5)
LTU places(a) Net Impact(b) Positive Cost per Cost per net
% 1991 - 1994 % Outcome % positive impact (e)
outcome (d) $ $
JobStart 72.1 23 59.3 1 895 4 887
JobSkills 93.4 8 40.8 15 015 76 575
JobTrain 71.8 11 41.2 1 883 7 055
JobClubs 82.9 11 47.3 1 414 6 082
SkillShare 59.8 12 42.2 2 277 8 008
Source: Information supplied by DEETYA to questions asked by the
Senate's Employment, Education, Training Legislation Committee in Additional
Information Received (vol.2 October 1996).
(a) LTU (Long term unemployment). The proportion of places taken
up by participants registered as unemployed for 12 months or more.
(b) Net Impact. These estimates account for the differences
in the employment levels of program participants and similar people
who were not assisted, 6 to 8 months after participants had left their
program places. Measurements of net impact control for criteria such
as duration of unemployment, age, gender, educational attainment and
geographical location. They represent the average net impact for all
program participants but do not control for differences in client characteristics
between programs, such as the proportion of places taken up by long
term unemployed
(c) Positive Outcome. This is the proportion of participants
who, 3 months after leaving assistance, were in unsubsidised jobs or,
while not employed, were participating in a non- DEETYA education/training
place.
(d) Cost per positive outcome. This is the cost to government
for each former program participant who is in an unsubsidised job or,
while not employed, is participating in a non-DEETYA education/training
place.
(e) Cost per net impact. This is described as the cost to government
for each person who got an unsubsidised job specifically as a result
of having participated in a program. Using the JobStart data as an example,
the cost per net impact of $4 887 can be derived by taking the cost
per positive outcome ($1 895), multiplying this by the positive outcome
(59.3%) divided by the net impact (23%) = $4 887.
In its advice to the Senate Committee, DEETYA confirmed that the cost
per net impact is an estimate of the cost to government incurred for each
person who got a job as a result of the program. Net impact is the difference
between the employment outcome levels of program participants and those
of similar jobseekers who have not been assisted.
Measurements of net impact allow for criteria such as duration of unemployment,
age, gender, educational attainment and geographical location (so that
the groups being compared are, from an employment perspective, similar).
They represent, according to DEETYA's advice to the Senate Committee,
the average net impact for all program participants but do not control
for differences in client characteristics between programs, such as the
proportions of places taken up by the long term unemployed (emphasis
added).(6) The lower the positive outcome rate for any program, the higher
the cost per positive outcome. It is always important to note that the
object of deriving these costs is to estimate a success rate vis a vis
a similar unassisted group.
Relying in part on the information provided to answer Question 2044 (cited),
information provided in DEETYA's Post Program Monitoring reports and further
information provided by DEETYA, Access Economics used the following parameters
to make a case for LMP outcomes.(7) These were:
(a) initial net unit cost
(b) percentage of program clients who achieve unsubsidised employment
relative to a control sample of matched CES registrants
Period after the end of program
(months) 1 to 12 13 to 24 25 to 36
SkillShare 12% 8% 4%
JobStart 30% 8% 4%
JobSkills 8% 5% 2%
JobClubs 12% 8% 4%
(c) The length of time over which this improvement persists.
(This reflects) the average 40 per cent annual turnover in the labour
market, the net improvement from participation in labour market programs
is assumed to diminish to zero over 36 months following the end of participation
in the program.
(d) The proportion of those placed in unsubsidised employment that
represent additional overall employment, rather than the displacement
of existing employed. The central assumption is that 90 per cent of
unsubsidised jobs gained by program clients reflect the displacement
of existing employees rather than an expansion of total employment.
(e) The proportion of those displaced who end up receiving Commonwealth
allowances. The central assumption is that 20 per cent of displaced
employees receive a Commonwealth allowance. This compares with 70 per
cent of CES registrants who seek allowances.
Using these parameters and making other assumptions, the net Budget impact
of the LMP were calculated by Access Economics. The results, together
with DEETYA's revised calculations can be seen in Table 3. In short, while
Access Economics reports a 'net gain to the Government Budget for the
SkillShare and JobClubs programs of $82 million and $12 million respectively,
DEETYA now finds that these two programs result in losses to the Government
of $136 million and $24 million respectively (over a nominative time period).
Net cost per Savings per Net gain per Total gain to
participant $ participant $ participant $ government $m
(a) (b) (c) (d)
Access Economics
calculations
JobStart 1130 971 -183 -17
JobSkills 5460 471 -4989 -145
SkillShare 440 971 507 82
JobClubs 680 971 267 12
JobTrain n/a n/a n/a n/a
Revised DEETYA
calculations
JobStart 1803 302 -1501 -139
JobSkills 6748 99 -6649 -193
SkillShare 999 157 -842 -136
JobClubs 680 144 -536 -24
JobTrain 776 144 -625 -57
(a) Initial net cost per program client
(b) Savings in the 3 years following program participation
(c) Net gains/losses over 3 years per program client
(d) Net gains/losses to the Commonwealth Budget over 3 years
Note: It is not clear as to why savings (column 2) less net cost (column
1) do not always result in the figures shown in column 3 (net gain per
participant)
The points made by DEETYA to the Senate Committee in its rebuttal of
the Access Economics calculations were:
(a) the assumption that income support reductions apply to all people
assisted through SkillShare was incorrect. Income support reductions
apply to around one-third of those assisted.
(b) It was not correct to assume that former SkillShare participants
are in full-time employment after participation. DEETYA finds that 50
per cent of such participants are in part-time employment with many
receiving an allowance (or part allowance).
(c) The assumption that 20 per cent of displaced employees will get
income support is doubtful. DEETYA believes that it is most unlikely
that program participants displace employed persons. In any case DEETYA
concludes that 65 per cent of displaced employees will seek work and
allowances. (These arguments are difficult to rationalise, unless Access
Economics is taking 20 per cent of a 'large' group of displaced employees,
while DEETYA is referring to 65 per cent of a 'small' group of displaced
employees).
(d) DEETYA claims that the assumption of the positive impact of participation
(over 36 months) is open to question, since DEETYA surveys generally
are limited to 8 months after participation.
The main conclusion to be drawn from the exercise is that in assessing
labour market programs, it is important to make allowances for the 'market'
for which a particular LMP was designed to address. There are variations
in the proportion of LTU places in each of the major DEETYA labour market
programs; this is evident from the data presented in Table 2. For example,
the share of LTU places in JobStart is considerably lower than for JobSkills
(let alone NWO). This means that in terms of the costs presented in Table
3, the 'cost per net impact' is going to be higher where a higher proportion
of the LMP placements are from the LTU.
Reliance on this and other information has allowed DEETYA to correct
the calculations made by Access Economics. However Access Economics, in
its analysis, acknowledged that it did not have access to detailed administrative
records which DEETYA has relied on in its rebuttal, and therefore 'were
not able to explore the detailed phasing of costs and savings during and
immediately following clients' involvement in the programs'.(8)
DEETYA's costings and assumptions hinge on a comparable, non-assisted
group being available for the purposes of comparison with those assisted.
It might be expected that after July 1994, all Job Compact clients would
receive some assistance; that all registered LTU would be expected to
participate in a program (or forfeit allowances) and therefore the availability
of a non-assisted comparative group post 1994 should diminish. Eighteen
months or more would appear to be a long time to be unemployed and
not assisted, but the existence of such a group is essential for DEETYA's
comparison purposes. It is noted that DEETYA used net impact surveys taken
over 1990-1994 (and therefore the Job Compact obligation would not apply
over all this period). However, the issue of comparability of assisted
and non assisted unemployed groups might be clarified by DEETYA providing
additional information.
DEETYA's response to the Access Economics' calculations of net gain to
the Budget from LMP, inevitably give rise to the question of whether LMP
spending bears sufficient return to warrant the costs. In a recent post
program survey of participants, the (then) DEET concluded that all programs
had high positive outcome levels, (refer to Table 1) with large proportions
of clients in unsubsidised employment or in non-DEET education and training
places, and that for most programs, unsubsidised employment outcome levels
were higher at six months after the conclusion of program participation,
than at three months after the conclusion of the program. The evidence
therefore seems to suggest that participation in a program improves the
chances of an unemployed person winning employment.
- Access Economics, Net Budget Impact of SkillShare, July 1996
- See Senator Vanstone's media release SkillShare, the real cost
to the Budget (v76/96 17 August 1996).
- Working Nation: Evaluation of the Employment, Education and Training
Elements Department of Employment, Education , Training and Youth
Affairs (July 1996).
- See papers Budget Initiatives, Budget 1996 by the Department
of Employment, Education, Training and Youth Affairs (1996): 11.
- DEETYA response to Question No.225 in Additional Information Received
by the Senate's Employment, Education and Training Committee (Vol.2,
October 1996): 8. Post program monitoring studies conducted by DEETYA's
Evaluation and Monitoring Branch have used concepts such as net impact
and cost per net impact; see Net Impact Study of Job Clubs (EMB
Report 5/94).
- ibid: 8.
- Access Economics, Net Budget Impact of SkillShare, July 1996.
- Ibid: 6
- DEET 6 Month Post Program Monitoring Survey Report (EMB Report
3/95) October 1995: 4.
These programs pay an allowance to the jobseeker, or a subsidy to an
employer. When a jobseeker is placed under an employment program, the
allowance/subsidy replaces the individual's benefit received from the
Department of Social Security (DSS) thus DEETYA bears the cost of employment
program placements.
JobStart A wage subsidy program paid to employers, allowing them
to employ a jobseeker under a subsidy (commencing at $100 per week for
an adult) and generally available for between 13 to 26 weeks.
NTW, National Training Wage Also a subsidy scheme available to
employers, but the employer and jobseeker need to sign a training agreement
to be eligible for this subsidy.
NWO, New Work Opportunities Allows public funding of employment
projects in situations/regions where it is unlikely that a private employer
would be able to offer a placement (under a wage subsidy).
LEAP, Landcare and Environment Action Plan Provides training
and experience to people, aged 15 to 20 years, on environmental projects.
JobSkills Provides adults with training and work experience through
community organisations, whom DEETYA terms brokers. Brokers submit a costed
project (including the numbers of trainees to be employed) to DEETYA for
funding.
NEIS, New Enterprise Incentive Scheme A program which provides
financial assistance for an unemployed person to start a small business.
In most cases, these programs allow the jobseeker to retain payment of
an unemployment benefit paid by DSS while the individual undertakes training
under a DEETYA program.
JobTrain Allows jobseekers to receive a benefit (the Formal Training
Allowance) while participating in training courses, usually of 8 to 10
weeks duration.
Special Intervention Provides for professional assessment of individual
jobseekers.
ATY, Accredited Training for Youth Provides young jobseekers with
special access to training places in courses such as TAFE courses.
Job Clubs These are support and personal development programs
allowing jobseekers to access skills and facilities (fax, telephones)
in order to approach employers.
SkillShare Allows community organisations to receive funding to
run training programs, but they must secure 15% of costs from the local
community. Services include personal referral and enterprise activities.
There were 402 SkillShare projects in 1995.
MAS, Mobility Assistance Scheme Provides relocation assistance
to jobseekers.
CAP, Community Activity Program Allows jobseekers and spouses
to undertake voluntary work. CAP functions have been replaced by the NWO
program.
TAPS, Training for Aboriginal and Torres Straits Islanders
Program Is designed to provide broad ranging employment assistance
to indigenous peoples, both employed through career development and to
the unemployed through public and private employment providers. TAPS is
not the same program as the CDEP, Community Development Employment Program,
which also provides employment placements. It was decided in 1992 that
the Aboriginal and Torres Straits Islander Commission would administer
the CDEP program.
OLMA, Office of Labour Market Adjustment (DEETYA) Allows
retrenched workers and those threatened with retrenchment access to information
on alternative employment opportunities as well as financial assistance
to seek new work.
|
