When do we want it? After a meticulous analysis of the relative efficiency of various taxes!

And they say economics is boring.

I'm having placards made up.

Economists talk about tax reform like it's a universally accepted good thing. But what does tax reform mean?

Broadly, tax reform means raising more government revenue through taxes that do not distort behaviour in the economy and less from highly distorting taxes.

All taxes distort. They prevent mutually agreeable transactions that might otherwise have taken place from going ahead. Companies invest less than otherwise. Workers work less. Consumers spend less. This reduces economic output.

This is particularly true in a globalised world, where investment can go overseas and workers can relocate in reaction to taxes.

Of course, taxes are the price we pay for all the public services we demand of government. They are a necessary evil.

So how do we decide which taxes are more or less evil?

A very simple rule of thumb is that we should tax things that cannot move.

You wouldn't poke a tiger, would you? But you might poke a tortoise: he's much less likely to react.

Taxing companies and individuals is a bit like poking a tiger. They react, either by working less or running away overseas.

Poking things like consumption and land are easier. People still need to buy bread and land is the ultimate immoveable object.

Economists get a little more technical, using complex models of the economy to estimate the "marginal welfare loss" or "marginal economic burden" of taxes. For each extra dollar raised through a particular tax, how much economic value is lost?

Modelling for the Ken Henry tax review by KPMG established a clear hierarchy of taxes.

All taxes are bad. But of the taxes levied by the federal government, company tax emerged as the worst, reducing community welfare about 40¢ for every extra dollar raised.

Income taxes came second on the league table of bad federal taxes and the GST ranked as the least-bad federal tax.

For every extra dollar raised through the GST, only 8¢ of economic value was destroyed. Each extra dollar raised through income tax destroyed 24¢.

The implication was clear. Although Henry was forbidden by the Rudd government from mentioning the acronym "GST", he did advocate a shift away from taxing incomes and towards taxing consumption – that's the GST to you and me.

But subsequent modelling by Treasury this year for the Hockey tax review has thrown a hand grenade into this comfortable consensus.

When they crunched the numbers, Treasury found income taxes destroyed 21¢ of economic value while the GST destroyed 19¢ – not much of a difference there. Why bother with a politically damaging campaign to cut income taxes and raise the GST?

It's hard to explain exactly why the modellers get such different results. Treasury had a stab, concluding that "subtle variations in model calibration" can affect the outcome. That sounds like a nice way of saying "we're not sure".

Basically, models produce varying results depending on the assumptions you feed into them.

Assumptions about future growth, inflation, employment and even global trade can and do swing results wildly.

Modellers typically build their model based on one household and extrapolate results for all households, regardless of the Technicolour  variations among households.

Do they do this because they're evil? No. They do it because modelling is hard and necessarily involves simplification.

So who to believe? Deloitte also models these things and gets a result closer to KPMG.

Most economists continue to believe that GST – particularly a broad-based one that applies to all goods and services – is less distortive than taxes on incomes.

For high-income earners, who derive prestige and self-esteem from their work, higher taxes might not discourage them from work. But they may be more likely to hire fancy accountants to minimise their tax or move to Hong Kong.

But for middle and lower-income earners, particularly working mothers, income taxes do heavily influence the decision to work or not. Many mothers lose more in tax and benefits foregone than they would earn by returning to work. This is a real economic loss.

Treasurer Joe Hockey has tied his colours to the mast of lower income taxes and most economists would agree that lower income taxes would be a good thing.

Of course, there's the question of how to fund them given the budget is in the red.

Reading between the lines, and the unspoken consensus of this week's National Reform Summit, is that the only way to raise the revenue required to fill the gap is by raising the GST.

Of course, so far we've talked only about efficiency. Fairness is another question entirely. The GST is clearly more regressive than income tax because it hits all households, including retirees, regardless of their ability to pay.

Does that mean we shouldn't increase it?

That comes down to whether you think the potential economic gain from a more efficient tax base is worth the trouble and the potential unfairness.

On the latter, the regressive nature of increasing the GST can be ameliorated through pension and benefit boosts and income tax cuts that compensate vulnerable households fully.

Our priority must be to grow the economic pie first, then distribute it.

Whether the economic boost from such a reform would be sizeable enough to make much of a difference, we just don't know for sure.

But with the mining boom over and living standards on the slide, the Australian economy needs all the help it can get.