The Coalition proposal to dam rivers in northern Australia to create a massive food bowl has been floated before, as have various schemes for harvesting water from what is perceived to be the over-watered and under-populated north and redirecting it to the under-watered and populated southern regions of the country.

Both of these ideas are superficially attractive, especially when you sit in an office in water-restricted Canberra, Sydney, Melbourne, Adelaide or Perth, look at a map of Australia and see that 60% of the continent’s runoff is in the north.

Yet neither of them withstands any serious economic analysis, which may be why private investment in such schemes has been so modest and ephemeral.

Before any new dams in the north are proposed, it is crucial to understand the lessons from the Ord River. Public investment in Ord Stage One from 1958-91 incurred a net loss of $511 million (1991 dollars) against a net private benefit of $14 million, and only 4400 hectares of a potential 70,000 hectares was being cropped at that stage.

This unviability had been predicted by the economist Bruce Davidson in 1965 in his seminal (and much ignored) book The northern myth: A study of the physical and economic limits to agricultural and pastoral development in tropical Australia. In the 20 years since, the area under crop has grown to more than 15,000 hectares and the gross value of agricultural production to around $100 million.

Salinity problems had already emerged by the early 2000s. Taxpayers are continuing to underwrite the expansion of the scheme: the WA government is investing $220 million and the Commonwealth $195 million in Ord Stage Two, but the level of private investment is uncertain and no serious economic analysis has been done.

If and when the Ord Irrigation District is in the black — let alone consistent with COAG water reform principles that water pricing should reflect the full cost of infrastructure provision and maintenance — we may be able to apply the lessons learnt to other tropical schemes.

Joe Ross, chair of the Northern Australia Land and Water Task Force, observes that most of the value currently being generated from the Ord today is from sandalwood, not food crops. “Now, you tell me, can you eat a bowl of sandalwood?” asked Ross.

That taskforce, which delivered its final report in 2009, was the latest of many studies since 1912 to document the formidable constraints in northern Australia to conventional notions of irrigated agriculture — whereby you dam a river, to supply a large contiguous irrigation district downstream of the dam.

First, for sustainable food production at any scale, reliable water supply is essential but not sufficient — you also need good fertile soils. The soils of northern Australia have been leached by monsoonal rains over thousands of years. They are low in nutrients and organic matter, they can’t hold much water, they erode easily, and they have low infiltration rates. These traits, combined with the very high amounts of solar radiation and evaporation rates of the north, mean that crops have low water-use efficiency and evaporative losses from dams are extreme.

Prospective dam sites are limited, and tend not to be located near reasonable areas of arable soils. A megalitre of water weighs one thousand tonnes, so moving it any distance over flat lands requires lots of energy, which is why the calls to pipe or channel or carry water from the north to the south are so misconceived, failing the most basic tests of physics or economics.

Extreme monsoonal rainfall events are problematic for large-scale irrigation systems. Input costs — fertiliser, chemicals, diesel — are much higher in the north, as is the cost of getting produce to market. Labour costs also tend to be higher, and it is difficult to attract skilled workers to live remotely in difficult climatic conditions, especially in competition with the wages offered by the minerals sector.

I am concerned here primarily with physical and economic arguments, but suffice to say that the environmental, social and cultural impacts of a network of dams and irrigation schemes across one of the largest regions of free-flowing rivers left in the world, would be profound.

My colleague Professor Michael Douglas has noted the crucial connections between wet season flows and coastal fish stocks. Many tropical fish species need to move between salt water and fresh water to complete their life cycles, so the connectivity of these monsoonal systems, both between upper catchments and the coast, and between rivers and floodplains, is crucial for recreational, commercial and indigenous fishers.

The Northern Australian Irrigation Futures research project (2003-2007), led by CSIRO, examined what sustainable irrigation systems in northern Australia might look like.  That project concluded that a patchwork mosaic of smaller-scale irrigation based on groundwater, located in areas with better soils and better transport options, possibly integrated with the pastoral industry for fattening cattle and other livestock options, would be more prospective than large dams.  But this concept has yet to be tested at any significant scale in the north, nor subject to economic analysis.

The Australian editorial of September, while critiquing what it called “the dam buster mentality” concluded with a sensible caution toavoid projects that are not viable, either economically or environmentally”.  Proponents of large-scale, conventional irrigation schemes based on dams in the north need to be able to disprove Davidson. No one has yet done so.

Yet for all the reservations about dam-based irrigation schemes, there is something inherently tantalising about the idea of big, nation-building projects in the north.