The ABS's environmental statistics reporting fails the basic test of rigour

The Australian Bureau of Statistics’ (ABS) 2002 report Measuring Australia’s Progress found that five of the six "headline indicators" for the environment suggest deterioration over the last decade. Land clearance rates, dryland salinity levels and greenhouse gas emissions have all apparently increased, our inland waters are overused, and Australia’s biodiversity is in decline. The only indicator that showed improvement was air quality with fine-particle pollution dropping in major cities despite an increase in motor vehicle use.

During the recent visit by Associate Professor Bjørn Lomborg, Danish statistician and author of the controversial book The Skeptical Environmentalist, an article in The Courier-Mail by renowned Australian environmentalist Professor Ian Lowe cited the ABS findings as evidence that economic growth in Australia is running down our natural capital.

The ABS’s conclusions run contrary to Lomborg’s treatise that, in developed countries, trends with respect to important environmental indicators show improvement.

In his book, Lomborg emphasizes the relevance of environmental statistics as a measure of the state of our environment, including the magnitude of the problems we face. If the statistical trend is one of improvement, we can be confident we are "on the right track". If environmental statistics indicate a deteriorating situation, however, as appears to be the case in Australia, there is reason for concern and a need to re-look at "our political aim". Indeed, if the situation is as bad in Australia as the ABS statistics suggest, our billions of dollars worth of investment in environmental programs and the plethora of new environmental legislation and regulations introduced over the past two decades are clearly inadequate.

The ABS report begins by acknowledging that, "it is difficult to obtain national time-series data that encapsulate the changes in Australia’s natural capital". However, the report then proceeds to present graphs with trend lines suggesting that relevant facts have been systematically collected. Yet with reference to the first indicator, biodiversity, the report states that its use of numbers of species listed as vulnerable and endangered under federal legislation as a measure of biodiversity may not be reliable.

The report then concludes, however, that declining biodiversity is nevertheless a reasonable conclusion because, "many experts … believe that total Australian biodiversity declined during the 1990s". Apparently the ABS will use opinion as a substitute for measured statistics.

An increase in land-clearance rates is provided as further support for the claim that biodiversity is declining.

But what if the area of new native forests is greater than the area cleared? Statistics can be misleading if an analysis appears more scientific than it really is, or if an indicator provides a spurious correlation or the comparison is not really valid.

Biodiverstity

The ABS conclusion that Australia’s biodiversity is declining is based on the total number of birds and mammals listed as extinct, endangered and vulnerable in schedules to the Commonwealth’s Endangered Species Protection Act 1993 and Environment Protection and Biodiversity Conservation Act 1999 (EPBC) for each year from 1993 to 2001. In fact, no new species were listed as extinct. Almost 80 per cent of the increase can be attributable to species being listed as vulnerable under the EPBC Act.

According to the Act, species can be listed as vulnerable if there has been a substantial reduction in its numbers and its geographic distribution is limited. The reduction in population numbers, however, could have occurred decades ago with population numbers now stable or potentially increasing. As a consequence, new listings do not necessarily give an indication of current trends with respect to biodiversity.

A significant scientific literature recognises that threatened-species listings are not a good measure of biodiversity change and advises against the use of such lists in state of the environment reporting.

Any person may nominate a native plant or animal species for listing under the EPBC Act, with over 311 species nominated over the last two years. Indeed, nominating a species is an integral part of many environmental campaigns. Successful nominations normally secure significant State and Federal government funding for the development of associated Recovery Plans.