If you don't know what your inputs really cost, it's hard to know the real value of your company's product. Whether you build cars or make milk powder, that can cause problems.

When Germany reunified after the fall of the Berlin wall, legend has it that a West German auto executive emerged weeping from a Trabant car plant. Why? The value of the engine-block iron and the other materials going into the plant was higher than the value of the cars coming out the other end.

But nobody in Soviet-era Germany really knew it, because nobody really knew the value of those inputs. Building Trabants destroyed value.

And we might have similar problems out on the Canterbury plains. It's not likely, but it's possible.

There has been a lot of outrage that a water bottling plant is being set up in Ashburton. A similar one was built in Hastings last year.

Demonstrators protest against Nestle bottling water during the California drought, outside a Nestle Arrowhead water bottling plant in Los Angeles last year. Americans drink an average of 129 litres of bottle water per person, according to the International Bottled Water Association.

The plants draw water from the aquifer, put it in bottles, and sell it in Asia. Because New Zealand awards consents to draw water but nobody puts a price on water, critics see this as profiteering on an unpriced resource.

​At the same time, over a thousand Canterbury dairy farms put water into cows. Dr Daniel Collins estimated that it takes about 250 litres of river and aquifer water, through irrigation, to produce a litre of Canterbury milk. That will not be a net measure, as some of the irrigation does flow back into the aquifer.

But it will take many more litres of water to produce a litre of milk than it takes to produce a litre of bottled water. The milk is collected, the water extracted, and the powder is sold in China.

And so we come to what might be the Canterbury Trabant plant. Does anyone really know whether water from Canterbury's aquifers is more valuable when put directly into bottles and sold to Asia, or when it routes through a cow along the way?

How low does the price of milk have to be before it would make more sense to leave out the middle-cow?

It is a tough question to answer, and especially where water allocation is set by consent rather than through markets. East Germany allocated iron by something not that different from consents, rationing scarce resources across various industrial uses, and wound up making cars that were worth less than the inputs that went into them.

New Zealand allocates scarce water by consents, and hopefully does a better job of it. Trabants were ghastly; New Zealand milk is delicious.

Too often, New Zealand dairying is derided for being insufficiently "value-added". Those critiques utterly miss both the technological complexity of a lot of New Zealand milk products, and the potentially large value created by turning water into milk. But unless we know what the water is really worth, it is a bit hard to tell.

Is there a better way? Of course there is.

Former Canterbury University Senior Lecturer John Raffensperger developed a smart market system letting people with water drawing rights on the Canterbury plains trade those rights.

The system cleverly managed things where drawing a litre of water has different effects on the aquifer depending on where that litre is drawn. It also could make sure that the rivers continued flowing, even in dry years.

What happens when farmers, and water bottlers, and towns and cities, can trade litres of water consent with each other? We can find out where the water really has the most value.

If a dairy farm wants to expand and cannot draw more water for irrigation under its consent, it would have to buy drawing rights from other users.

If those users are happy to give up their drawing rights for the amount offered, then the water is more highly valued when it is run through a cow and turned into milk products. If a water bottling plant that wants to expand is able to pay dairy farmers to hand over some of their drawing rights, then those litres of water are more valuable going straight into the bottle.

Public outrage about the bottling plant seems focused on the presumed fact that the company bottling the water is not paying the government or anyone else for that water. That is not quite right, because the drawing consent comes with the section for sale: the value of the water is built into the price of the section that is for sale by the council. And if a dairy farmer bought it and profited from drawing that water to produce milk, nobody would bat an eyelash.

Shifting to a trading mechanism would make sure that New Zealand water were put to its best use. That also makes water more valuable – a gain for those with current consents. It also would make it simpler to manage the system in dry years: the government could buy back drawing rights to keep rivers flowing, then sell them back to farmers when rivers run freely. And the farmers who can most easily scale back irrigation in dry years would be the ones to do so.

Public outrage about the bottling plant is really misdirected. What is really outrageous is that New Zealand's system for allocating water looks so much like East Germany's system for allocating iron. Let's hope we have no hidden Trabant plants.