Saturday, March 16, 2013

Why tax revenue is falling short of budget

Try this quick quiz: which matters more, the growth in ''nominal'' gross domestic product or the growth in ''real'' GDP? Sorry, it was a trick question. The right answer is a favourite reply of economists: it depends.

If your interest is in how fast the economy's growing (or not growing), the answer is real GDP - GDP after allowing for the effect of inflation. But if your interest is in how fast the federal government's tax receipts are growing the answer is nominal GDP - GDP before allowing for inflation.

Why is nominal the right answer for tax receipts? Because, as Treasurer Wayne Swan keeps saying, ''we live in the nominal economy through the prices we pay and the incomes we earn''. As part of this, the income tax we pay is based on our nominal income and the indirect taxes we pay are based on our nominal spending.

Fine. If you didn't know the growth in nominal GDP is the best guide to the growth in tax revenue before, you do now. But why has Swan been making so much of this in recent days? Because it's the main reason why, despite all his savings measures (and creative accounting), the government won't be able to keep the promise it made in the 2010 election campaign to get the budget back to surplus this financial year.

That promise was based on a Treasury projection for 2012-13 included in the 2010-11 budget. But tax collections simply haven't grown as strongly as Treasury projected they would, and the main reason they haven't is that nominal GDP has been behaving strangely.

We're used to assuming that, if the economy's growing in real terms (which it has been), the government's tax revenue will be growing at least as fast and probably faster. (Why faster? Because almost half the feds' tax collections come from personal income tax, which grows extra strongly because, in the absence of the indexation of tax brackets, it's subject to ''bracket creep''.)

When economic events are proceeding normally, the distinction between nominal and real GDP doesn't matter much. Obviously, the difference between nominal and real GDP is the inflation rate, and if inflation is running within the Reserve Bank's 2 per cent to 3 per cent target range, the two totals should be moving pretty much in parallel.

To put it another way, nominal GDP should be growing at a reasonably steady 2.5 percentage points or so faster than real GDP. But we learnt from last week's national accounts for the December quarter that, for the first time on record, the past three consecutive quarters have seen nominal grow by less than real, not more. Since real GDP grew by 1.9 per cent, nominal GDP should have grown by about 4 per cent. Instead it grew by a pathetic 1.6 per cent.

Swan noted in a recent speech to business economists that nominal has grown by less than real for only four short periods in the 53 years since the Bureau of Statistics began producing quarterly national accounts.

The last time nominal was really weak was in the global financial crisis. Before that it was the Asian financial crisis of 1997-98 and, before that, the Menzies government's credit squeeze in 1961. In all but the credit squeeze episode, the explanation was the same: a sharp fall in global commodity prices led to a sharp deterioration in our ''terms of trade'' - the prices we receive for our exports relative to the prices we pay for our imports.

Ah. Whenever we talk of inflation, people think automatically of the main measure of inflation we use, the consumer price index. But in fact there are many measures of inflation, most of the others being derived from the national accounts.

The difference between nominal and real GDP is measured not by the CPI but by the ''implicit price deflator'' for GDP. When the economy's travelling normally, there shouldn't be much difference between the GDP deflator and the CPI and other measures of the change in the price of domestic spending.

But ''normal'' means when our terms of trade aren't changing much. When they're improving or deteriorating sharply, the GDP deflator and measures of domestic-spending inflation really part company.

Why? Because domestic spending includes the prices of imports but excludes the prices of exports, whereas GDP and its deflator exclude the prices of imports but include the prices of exports.

It works out that nominal GDP will grow very much faster than real GDP when our terms of trade are improving sharply, but nominal may even grow more slowly than real when our terms of trade are deteriorating sharply - as they were last year. But why wasn't Treasury expecting

the terms of trade to deteriorate and allowing for this in its projections of tax revenue? It was, and it has been - for most of the past decade, in fact. But it wasn't budgeting for the deterioration to be as fast as it's been, particularly in the September quarter.

That was the first problem with its revenue forecasts. The second, less obvious, one was this: on the basis of past behaviour, Treasury (and everyone else) was expecting any deterioration in the terms of trade to be accompanied by a similar fall in the exchange rate.

To everyone's surprise, the dollar has stayed up. This means the prices of imports haven't risen in the way you'd have expected, causing domestic inflation to be lower than expected. This, in turn, has meant nominal incomes haven't risen as fast as could have been expected.

So this factor, too, helps explain why tax collections haven't risen as fast as forecast. The latest estimate is that tax collections will fall about $10 billion short of what was forecast in the budget last May.

The last thing to say is that by no means all federal taxes are closely aligned with nominal GDP. The strongest relationship is with taxes on profits - company tax, income tax on unincorporated businesses and the two resource rent taxes. These account for about a third of total tax revenue.