Wednesday, October 15, 2008

The Australian economic stimulus package - distributional and timing issues

This post continues the discussion I began in The Rudd economic stimulus package -a missed opportunity, looking at distributional and timing issues. There is nothing profound in the material that follows. Think of it as brief notes to myself to try to increase my own understanding.

Economic Linkages

The various Government initiatives are meant to fit together in this way:

  • the various elements in the finance guarantee package are meant to ensure domestic liquidity and to provide a measure of insulation for Australia from the previous global financial turmoil.
  • then there are the various longer term measures already announced that are intended to provide a long term growth base. I am not sure that these, and especially the integration between them, have yet been properly analysed.
  • and now there is the fiscal stimulation package intended to provide a short term fillip to domestic demand (the one-off payments) that should in theory then phase into increased housing construction.

The One-off Welfare Payments

The one-off welfare payments are equivalent to around .9 per cent of Australian GDP. However, those to receive the payments are not distributed equally across Australia, with higher concentrations in particular areas. This means that the impact in local terms will vary and in many cases may be far higher than .9 per cent of local GDP.

The one-off pension payments ($A4.8 billion) will be especially important in those areas with above average concentrations of age pensioners. I have not attempted to chart this, although ABS census data could be used to chart distributional impacts using a combination of demographic structures with data on welfare recipients.

As a broad generalisation taking NSW as an example, the greatest relative impacts are likely to be in regional NSW simply because older people make up a higher proportion of the total population. The overall impact could be quite high in individual localities, equivalent to a one-off increase in total weekly incomes in some areas (the income in that area in that week) that could, on some rough back of envelope calculations, be over 50 per cent, in some cases well over 50 per cent.

Conversely, the one-off child support payment ($A3.9 billion) will have its greatest relative impact in areas where there is a substantial proportion of young from lower and middle income families eligible to receive the Family Tax Benefit such as the outer suburban areas. The area impact here is likely to be less simply because the payment is smaller, while the relative proportion of the young is also smaller.

I am not sure that I draw any substantive conclusions from this in terms of its aggregate (average) effects. I just find it interesting.

All these one-off payments will be made from 8 December, sort of a Christmas bonus. However, we should not assume that the total impact will be an increase in aggregate spend hitting at one point.

A reasonably significant proportion of benefit recipients live in social housing and pay income based rents. If the one-off payments are counted as income, and it is unclear as to what extent this might happen, then around 25 per cent will be absorbed by rent increases. To this degree, the overall spending stimulus will be reduced.

We also need to take debt into account. A number of recipients have small debt amounts and will use the money to reduce that debt, thus reducing the aggregate spend effects. Others may borrow against the future benefit payment, this bringing spend forward.

The net result of all this is that the aggregate spend impact may be lower, its timing a little more distributed, than would be suggested by the raw numbers.

In measuring impact, we also need to take into account where the money will be spent.

Based on my limited knowledge, I would expect the majority of the money to be spent on groceries and beverages including alcohol, entertainment including gambling, car repairs (many pensioners are struggling the keep old cars running), some smaller consumer durables, things for the kids, meeting medical and dental backlogs etc.

So the the spend will translate to an immediate increase in consumption for certain types of goods and services. However, the flow-on stimulatory effects, the multiplier effect, depends upon leakages including especially imports. Because we import so much, a significant proportion of the spend is likely to flow though to a one-off increase in imports, reducing the subsequent stimulatory effect.

The First Home Buyer Grant

The time limited funding for first home buyers ($A1.5 billion) applies to existing and new homes, with a deliberate skew to new homes.

Increased payments for existing homes should immediately flow over increased property prices as compared to what would otherwise have been the case.

I find the economics here a little complicated.

The lower the house price, the greater the impact simply because of the ratio between the fixed amount of the grant and the house price. We saw this when the original scheme was introduced - some of the greatest percentage increases in house prices occurred at the low end in regional market places. One outcome was a reduction in the margin between highest and lowest house prices.

So regional house prices should increase at the lower end.

The effect in the outer metro suburbs is a little more complicated. House prices have come down in some of these areas as people struggled with higher repayments. This decline will be reversed to some extent as new buyers come into the marketplace. This will actually have the useful side effect of making it easier for the over-extended to exit.

The effects of the change on the rental marketplace are a little unclear, at least to me. An increased number of new home buyers implies less rental houses. However, by definition, as first home buyers they area already renting. So that frees up the house they were living in for rental. On the surface, the net effect on rental availability is likely to be zero.

I have heard some commentary that the rise in house prices flowing from this move will of itself encourage construction. I cannot see it.

The only thing that I can think of is that the move will encourage transfer of people between areas, soaking up unused and presently over-priced inventory in some areas. This may encourage further construction, especially of flats.

The grant for new homes should encourage new construction, subject to the key constraint that land is available. Again, the ratio between the size of the grant and the development price is important in determining actual outcomes.

Development price combines land costs, all the multifarious state and local government charges, construction costs plus a profit margin. On the surface, the greatest relative impact is likely to be in regional areas and in the construction of flats.

There is an interesting linkage here between new and existing homes, one that was not clear to me when I started writing. New homes compete with existing homes for buyers. In theory, a rise in price for existing homes may make construction of new properties more attractive.


I have run out of time. As I said, this post is really a note to myself to help me work through the issues.

Afternoon Update

As you might expect, the airwaves have been full of the package.

Just to check on the regional reaction I looked at Tamworth's Northern Daily Leader. One headline in the paper reads Home buyer incentive set to ignite market. Another story suggested that $4 billion of the funds would go to the country. I could not find the story again, but North Coast Voices carries the full press release on which the story was based.

I see that Neil is going to use his money to pay December accounts. He promises not to waste it!


Joel said...

One thing concerning me about the increased payment for first home buyers of new housing is the short lifetime of the grant.

On my understanding, it would be a longer term trend of increased buyers (therefore prices) that would send out a signal to build more properties. If the grant only lasts 9 months(ish), then the length of this trend would be shorter than the lag between receiving the price signal, planning the project, gaining Council approval, construction, etc.

Understood that some latent housing supply might be bought on, meaning shorter construction time. However, just using up latent supply just changes the timing of the problem instead of solving it.

Surely an 18 month scheme would better achieve the aims of solving rental affordability through increased supply?

And give me more time to scratch a deposit together?

Jim Belshaw said...

Interesting point, Joel, one that I had not focused on. Nine months is short. I am not sure about eighteen months, but twelve months? Presumably this could be modelled.

Good luck with the deposit!