Six years after mining billionaire Andrew Forrest’s Australian Employment Covenant was announced in a blaze of publicity with then-prime minister Kevin Rudd, questions are being asked about the authenticity of the more than 64,000 jobs pledged by hundreds of employers that have signed up, and the almost 19,000 commencements claimed so far under the program.

Crikey understands a number of the job pledges are unrealistic — in one case a pledge by the prospective builder of a new hospital at Minto in western Sydney, which went out of business, remained on the books — and many of the jobs commenced proved temporary, with workers churning through the system, creating the risk of systematic over-counting. GrainCorp, for example, is a signatory and employs casual Aboriginal workers at harvest time, although the company assured Crikey it does not re-count those positions.

When launched in October 2008, the covenant aimed to provide 50,000 indigenous jobs within two years and, with wealthy backers like James Packer, Kerry Stokes and Lindsay Fox, the program quickly attracted open-ended promises to hire appropriately trained Aboriginal candidates from hundreds of companies. As at July 31 this year, some 349 companies self-report twice a year under the system, and there is no transparent reporting of pledges, starts or retention data by signatory or year pledges were made.

The covenant is managed by Forrest’s non-profit organisation GenerationOne, wholly funded by his family charity, the Minderoo Foundation. The covenant is now at the centre of the federal government’s strategy, courtesy of Forrest’s appointment to review indigenous employment, which resulted in the “Creating Parity” report released for consultation at the beginning of August. The government has committed an extra $45 million to support up to 5000 extra jobs placed through GenerationOne over the next two years.

That money will be paid progressively to a growing network of vocational training and employment centres (VTECS), which arrange training for Aboriginal candidates to meet the specific requirements of employers who have signed the covenant — a demand-led model pioneered by Forrest’s Fortescue Metals. VTECs get cash payments per candidate once an Aboriginal employee lasts 26 weeks.

“Submissions following the Forrest report by a number of academics … are critical of the departure from evidence-based policy-making …”

GenerationOne claims 75% of jobs in the program last that long — much higher than the comparable 20% retention rate of indigenous workers placed by the government’s Job Services Australia under the Remote Jobs and Communities Program — but has no data on whether jobs are retained beyond that.

Last year the Australian National Audit Office was critical of the covenant scheme, noting that some $150 million of public funding had been committed as direct financial support for signatory employers by March 2013, and while “the initiative has been successful in generating aspirational commitments from employers … [these] do not necessarily lead directly or immediately to identified jobs that can be filled by Indigenous Australians”.

GenerationOne said company-by-company reporting data could not be released publicly because it was confidential and the organisation was not privy to details of public funding provided to signatories or VTECs. Ten of the biggest, most engaged covenant signatories, which have high retention rates, are Programmed, Leighton Contractors, John Holland, Downer EDI, Crown, Transfield, ISS, Compass Group, Rio Tinto and Fortescue. Given the size and profitability of such companies, it is debatable whether they should be receiving taxpayer subsidies to employ Aboriginal people.

Late last year GenerationOne provided an unusual submission to Andrew Forrest’s Creating Parity review — undated, unsigned, not on letterhead, it ran to just two pages — but acknowledged that “for every one job filled, two will go unfilled due to the lack of connection between job-ready Indigenous candidates and employers with positions to fill”. The submission noted inadequate recognition of pre-employment training and post-placement support left marginalised indigenous jobseekers “churning through the system”.

Despite this, the Forrest Review recommends persistence with the GenerationOne/VTEC model as the key to adding some 188,000 indigenous jobs this decade, ending employment disparity, and brands the former government’s $1.5 billion Remote Jobs and Communities Program a “failure”.

Submissions following the Forrest report by a number of academics from the Australian National University and University of Western Sydney, not yet online but seen by Crikey, are critical of the departure from evidence-based policy-making, including this submission by Dr Kirrily Jordan:

“There is no recognition of the potential conflict of interest in the Review being led by the chairman of a major corporation who has publicly lobbied for a very particular kind of employment and training service (the ‘VTEC’ model) for several years prior to the Review being undertaken. While I make no suggestion of impropriety, it is concerning to me in these circumstances that the Review presents the VTEC model as the only model worthy of consideration. For this argument to be convincing it would need to be supported by a careful analysis of existing evidence on the outcomes of various employment programs, as well as a considered reflection on the potential implications of replacing other employment and training services with the VTEC model nationwide.”