Sunday, February 15, 2009

Pointless fiscal expansions that swamp us with debt but don't reduce unemployment

I liked this article by Henry Ergas on the likely ineffectiveness of the $42 billion fiscal package introduced by the Rudd Government as it accords with my own macroeconomic priors. For a small open economy like Australia fiscal actions make sense in a standard Mundell-Fleming macroeconomic context if the exchange rate is fixed.  Monetary policy then does not work in this setting since, with international capital mobility, a monetary expansion to drive down interest rates is simply offset by an outflow of capital seeking higher international interest rates abroad.  This was one reason for seeking exchange rate flexibility - to give local monetary policy more bite. Monetary policy effects were then 'bottled up' in an economy and had a real impact.

With flexible exchange rates however, while monetary policy has bite, fiscal actions do not. The only effect of fiscal expansion is to drive up local interest rates attracting capital from abroad and hence appreciating the exchange rate.  This reduces our exports and nullifies the effects of the fiscal expansion.  The standard theory result is that fiscal actions have zero effect if exchange rates care flexible.  Monetary policy is a preferred means of expanding an economy.

As Ergas states:
"Australia is a small, open economy with a flexible exchange rate. There is consequently a real possibility that any increase in demand caused by fiscal easing will merely raise interest rates, induce capital inflow from abroad, appreciate the currency and reduce net exports.
With growth in China and Japan slowing significantly, why implement measures that could exacerbate Australia's expected export downturn?
 In the Keynesian framework, monetary policy, on the other hand, is actually more effective in an open economy. A monetary policy-induced reduction in interest rates boosts aggregate demand and induces capital outflow, leading to a depreciation of the exchange rate and a reduction imports.
As a result, even if we take the Keynesian approach seriously, fiscal stimulus may not only be ineffective, but by impeding or slowing further reductions in interest rates may stand in the way of a more effective response. As Treasury concluded in 2002, "higher budget deficits (or lower surpluses) can have a significant effect on interest rates in Australia", with the result that the "automatic stabilisers are likely to be relatively more effective than discretionary changes in policy". The federal Government must explain why those findings no longer apply".
Of course the Mundell-Fleming approach is only a simplified model - macroeconomists have a range of Ripley-Believe-It-or-Not macroeconomic models that justify any sort of policy action - but I would like to know why the effects the MF model stresses are not appropriate here. The argument that we might be in a liquidity trap where monetary policy is ineffective does not improve the case for expansionary fiscal actions. IMoreover if monetary expansions are taken to keep the exchange rate low then it is these expansions not the debt-incurring fiscal actions that are providing the stimulus. If the fiscal actions are funded using debt as they will be then although huge debts will be imposed on future generations there will be no immediate stimulus now as would be the case were they money-financed.

The standard objection to the Mundell-Fleming model that it is designed for settings where inflation and inflationary expectations don't have a role is not relevant. We have close to zero inflation.  Another objection might be that capital mobility is low because of the financial crisis itself.  That might be true right now but not so if and when the international economy recovers - and the effects of these fiscal measures will operate with long lags.

Those on the left seem to me to support the Rudd fiscal expansion because they support Rudd not because they understanding Keynesian macroeconomics in an open economy. I wish to know why expenditures of $42 billion by Rudd and his mates - which look likely to leave Australia with a huge eventual debt - were so self-evidently correct. Debate itself - and even quibbling about the size of the package - was portrayed by Rudd as something unpatriotic and unreasonable.  Was it that the $42 billion package was primarily designed to threatrically demonstrate that the government was 'doing something'? Indeed to save one job - Kevin Rudd's? The handouts of course appealed to those who simply like handouts.

We will live with the consequences of ineffective fiscal actions which raise debt, but which do not stimulate economic activity, for decades. If the current measures fail there will be inevitable calls by the Keynesian slobs for more and more and our debts will mount.  I am pessimistic.


Anonymous said...

Henry's right, They could cause the exchange rate to appreciate.

They really have no clue.

Anonymous said...
This comment has been removed by a blog administrator.
hc said...

Anonymous, you anonymously insult Henry Ergas (and by inference me since I agree with him) but add nothing to the discussion because you do not deal with the subject of the post.

Now gone.

Anonymous said...

The handouts of course appealed to those who simply like handouts.

It's unaustralian to not like, want and vote for handouts.

Anonymous said...


there are several reasons why the MF model is not a good guide to the effects of fiscal expansions, especially this one. Here are just a few.

1. It assumes the country doing the expansion is the only one doing it. On this occasion, fiscal actions are occurring all over the world. The exchange rate is a relative price which changes when country's fundamentals change relative to other countries'. But if every country is expanding fiscally, there will be no change in relativities, and so no change in the exchange rate.

You might argue that the amount of fiscal expansion differs by country, so relativities will change. But if anything, our fiscal expansion is less than is going on elsewhere - certainly less than the United States, on any measure. So by the logic of the MF model our exchange rate will depreciate, strengthening the output effects of the fiscal policy.

2. Even if you don't accept 1., the MF model gives a very poor explanation of exchange rates. Exchange rates in general and ours in particular are not explained by differences in bond rates. Our exchange rate is driven by commodity prices. This is why in the past 6 months the exchange rate has fallen despite our bond rates being high by world standards.

3. Even if you don't accept 1. or 2., there's very little reason to think our fiscal expansion will drive our bond rates up. The run of budget surpluses over the years created a shortage of Australian government bonds, for which there is great demand by super funds and others. The government was asked just last year by the financial markets to keep the bond market going when it looked like it was going pay back 100% of its debt. (Net debt has been zeto for a while; this was gross debt.)

4. Even if you don't accept any of the above, and take the MF model at face value, your and Ergas' conclusions still don't follow. Behind the scenes in the MF model is an assumption of fixed nominal wages. If instead you assume fixed real wages, then fiscal policy becomes effective and monetary policy does not. There was quite a literature on this in the 1970s and early 80s. So it comes down to what you think wage setters (including unions) will do. For your conclusions about the fiscal program to be correct, you would have to assume that they will passively sit back and let their real wages become eroded by inflation (which no one is forecasting to be zero). I don't believe this would be consistent with your beliefs about union behaviour ...

The moral is, be careful about making declamatory policy conclusions on the basis of simple models.

Anonymous said...

claptrap Harry.

Ergas has interest rates rising whilst the economy is not growing.

If fiscal policy leads to exchange rate appreciation surely that means we are growing more than the rest of the world hence the stimulus has worked.

If Henry had produced some comparable figures it is easy to observe that other counties such as the USA and UK, Japan, China have much larger programs as a % of GDP.

He claims we are near the NAIRU.

Wages were leading inflation higher.

Not here.

He claims the package goes on and on when it finishes in three years.

all this was discussed previously.

either Ergas knows this and so is paddling claptrap or he doesn't and is ignorant.

You were not mentioned either directly or by inference

hc said...

Uncle Milton,

Thanks for taking the trouble to set things out.

On point (1) I cannot believe fiscal expanssions elsewhere will have a bigger effect on the Aussi dollar than local ones. But maybe.

On point (2) the MF model is fixed price. Exogenous shifts in commodity prices will impact on the exchange rate. Given these interest rates will have an impact if capital is mobile.

(3) This suggests there will be no effect on interest rates in selling more bonds. So there must be a latent excess demand. Maybe.

(4) Again in the MF model prices are fixed - this pretty well reflects the current situation.

I think it is better to make policy statements on the basis of simple models rather than no models. I'd be interested to know the sort of analytical setting that you think would be useful here.

Again thanks for the comments which add to the picture of what is happening.

Anonymous said...

Harry, the MF model is only fixed price in the sense you describe in its textbook form. If you add a supply side, which the literature did, then it matters which price, the nominal wage or the real wage, is fixed.

I agree on the need for analytical frameworks. My preferred framework is CGE models which give you a rich range of real outcomes. There are plenty to choose from in Australia.

You can get crowding out in these models from real exchange rate appreciation, especially if government spending occurs when the economy is at full capacity. Of course the whole point of the government's fiscal program is that we are heading for a bad downturn and we will be far from full capacity, and that spending up will things not as bad as they would be otherwise.

The MF model obscures the main point, indeed the only real point, which is that fiscal policy will work when you're at the bottom of the cycle and won't work when at you are at full capacity.

And CGE models, unlike the MF model, are dynamic, so you can get a sense of how long it takes for a fiscal program to work, when any crowding out effects kick in, and so on.

Of course CGE models have their own shortcomings which are well documented. No analytical device is perfect.

Anonymous said...


a sudden demand for funds from the government when the real economy is out for the count will place upward pressure on the exchange rate.

We run a capital account surplus after all, which seems to be driven by the c/a deficit.

It means that a sudden demand for funds will of course cause the exchange rate to either appreciate of not fall has much. The net effect is that it is distorting the external flows of the country.

Anonymous said...

JC, there's even bigger demand for funds from other governments - one trillion dollars this year from the US alone. Not everyone's exchange rate can appreciate.

The current account question is interesting. We might just have some trouble financing it as investors put their money onto US treasuries, Japanese and European bonds and so on. If so the exchange rate will depreciate.

Anonymous said...

True, but the other point is that exchange rate may not actually appreciate, Spiros. It may not fall as much as it needs to, which is what i also alluded to.