The latest “closing the gap” report is full of the usual glossy pictures of happy people and propaganda on how much the government is doing for indigenous Australians — a list of programs and dollar inputs, little about outcomes.

Much of the media noise last week has been about “rivers of grog”, to divert attention from the unfortunately slow, even negative, progress in closing the six headline gaps nominated by the Rudd government in 2008. Census data that allow measurement of change between 2006 and 2011 are hard to find in the report that was ostensibly designed to hold the government of the day accountable.

The gap that one would think is easiest to measure with any certainty is the employment gap. The policy aim here is to reduce the difference in the employment to population ratio of those aged 15 to 64 years (the number employed as a percentage of those of working age) by 50% in 10 years. Employment data from the 2011 census were released on October 31. Between 2006 and 2011 the indigenous employment/population ratio for those aged 15–64 declined from 48.0 to 46.2.

What is especially concerning is that the ratio indigenous to non-indigenous has also declined from 0.67 to 0.64. This is acknowledged but well hidden in the closing the gap report: the employment gap increased by 2.2 percentage points since 2006 rather than declining. Other measures of employment outcome have also been disappointing: the indigenous unemployment rate has increased from 15.6% to 17.1% — imagine such an outcome for the general population.

None of this was mentioned in Julia Gillard’s statement to Parliament. Instead the PM notes: “The number of indigenous Australians in mainstream employment has risen from 42.2% to 44.7%, a small but encouraging increase.” But nothing to do with closing the overall employment gap. When challenged on this discrepancy on ABC’s Lateline by Steve Cannane, Minister Jenny Macklin quickly shifted the conversation to issues of income, not employment.

A campaign of denial and obfuscation on the worsening employment situation has been brewing for several weeks. A lead story on the front page of The Australian last month declared: “Indigenous work rises: Private sector closing jobs gap“. And on the morning of the PM’s statement we again heard from The Oz that following the release of the earlier story, “new data has revealed employment growth for Aborigines increased dramatically” based on an Australian National University major analysis of the nation’s closing the gap strategy by Matthew Gray, Boyd Hunter and Monica Howlett.

How could this be if the employment gap is growing? The answer comes from the emergence of a new definition of “non-CDEP employment”. This is a term that I too have used in the past to refer to mainstream employment but it can be misleading when some CDEP participants work full time, so I stopped using it.

“… seeking indigenous employment truths from official facts is open to diverse interpretations with potential for dire policy consequences …”

Here the term is used differently as employment minus CDEP participation and the unemployed. In other words, at the national level administrative data on indigenous participation in the Community Development Employment Program have been deducted from employment figures in the census on the basis that CDEP work is not “real” or “proper” employment.

While it is clear why a government searching for good news and The Australian for evidence to support its “real jobs” position on indigenous employment might adopt such a view, it is less clear why reputable labour economists take this step. In their view a failure to focus on “non-CDEP employment leads to very misleading conclusions about employment growth and the effectiveness of government policy aimed at increasing non-CDEP employment” — a different goal than closing the employment gap.

To partly justify their focus on “non CDEP employed”, Gray, Hunter and Howlett accept an erroneous definition of CDEP as “an indigenous-specific program that enables an indigenous community or organisation to pool the unemployment benefit entitlements of individuals into direct wages for those people who choose to participate in local employment in various community development or organisation programs as an alternative to receiving individual income support payments”. However, since CDEP began in 1977, unemployment benefit entitlements have never been “pooled” — CDEP has always received a budgetary allocation that has included wages, administrative and project allocations; the only direct link with welfare is that those participating in CDEP cannot get welfare because CDEP participation is regarded as employment, including by the ABS.

Indeed, the new focus on “non-CDEP employment” also contradicts the International Labor Organisation’s definition of employment as “persons in employment comprise all persons above a specified age who during a specified brief period, either one week or one day, were in the following categories: paid employment and/or self-employment”. And by suggesting CDEP is directly linked to welfare the new definition “non-CDEP employment” buys into the emerging populist notion — that all CDEP is welfare poison — divorced from any reality.

There is considerable published research based on official statistics that clearly demonstrates CDEP participants are quite different from the unemployed: they work, earn wages, participate in more non-market production, and are more enabled to meet cultural obligations. To suggest to CDEP workers their efforts are different because of a different form of funding makes little sense and demeans their efforts in a range of jobs.Part of the good news story reported by Gray and others is just a statistical artefact — 32,800 CDEP participants are deducted from 2006 estimates of the employed but only 10,692 from 2011 employed owing to the dramatic shrinkage in the popular CDEP program. If you exclude CDEP employment in 2006 and 2011, then the movement of some CDEP employees in 2006 to mainstream employment in 2011 will automatically escalate an apparent improvement in the rate of “non-CDEP employment”.

But by cutting CDEP from the equation a most significant shift in indigenous labour market activity, the dramatic shrinkage of CDEP, is overlooked. And no one wants to ask how much of employment growth is dependent on government funding in the private (e.g. Australian Employment Covenant and Generation One) and public (e.g. part-time government jobs to replace part-time CDEP jobs) sectors. The new approach renders over 10,000 CDEP participants invisible — they are neither employed nor unemployed, and so get a little lost in a statistical void.

Importantly, when Gray and others look at indigenous employment rates at the sub-national level they find 13 areas where indigenous employment (now including CDEP) has declined by between 20% and 47% — local labour markets have collapsed alongside “reform of CDEP” in remote places. This is not reported by The Australian or the government.

Given welfare dependency and inactivity are a major problem especially for remote indigenous communities, this is extremely worrying — clearly declining CDEP participation has only been partially replaced by employment opportunity and take-up. When confronted by the loss of 22,000 CDEP jobs on the Lateline, Macklin ducked for cover —  instead raising concerns about award wages that those working on CDEP are paid.

There is clearly a duality of views emerging on the indigenous employment situation nationally and regionally.

One view favoured by the government, The Australian and some ANU researchers is that the long boom is continuing and indigenous employment, especially in the private sector, is growing and closing the jobs gap. This view is greatly strengthened by a definition of Indigenous employment as ‘non CDEP employment’.

Another interpretation that I favour is that if we take the Government’s stated employment goal seriously, then since 2008 there is evidence from the five-yearly census that the employment gap is growing. This is especially the case in remote areas because CDEP is being cut in places where there are few alternate forms of employment.

Interestingly, as the long election campaign gets under way, the opposition has made no effort to hold the government to account for what, in my view, are abysmal efforts, both in CDEP reform that has condemned thousands of Indigenous people to enhanced welfare dependence and diminished livelihood; and in meeting its self-inflicted techno-bureaucratic statistical targets.

“Seek truth from facts” is an ancient Han expression promoted by the late Deng Xiaoping and applied to modern Chinese economic and political reforms. Clearly in Australia seeking indigenous employment truths from official facts is open to diverse interpretations with potential for dire policy consequences for the very people whose prospects are to be improved.

*Jon Altman is a research professor in economics/anthropology at the Centre for Aboriginal Economic Policy at the Australian National University in Canberra. An earlier version of this article appeared in Tracker.