Some thoughts on housing

I see that the latest Barfoot figures are out, and they are pointing to fairly strong sales figures in the Auckland region.  That’s nice. There are also suggestions that this is indicative of a bubble or boom coming into the housing market.  However, it is important to look at a broader view of what is going on, not just house sales for one agency in Auckland, in order to get an understanding of the what is really going on.

When looking at the housing market, there are a number of little points we need to keep an eye on.

  • The regional split: We have commonly been told that Auckland and Canterbury have a shortage of property – all else equal this pushes up house prices in these regions.  This is something we have seen happen.
  • Borrowing to invest?: In 2003, households borrowed heavily and developers started building heavily.  Now we have the opposite.  Yes, in the year to April households borrowed 30% more (in gross terms) to buy housing.  Yes, this was $45.4bn.  However, the stock of mortgage debt rose by a more modest $2.4bn more – or by 1.4%, below the rate of inflation.  Even as house sales and prices have climbed, households have taken the funds from sales to pay back mortgages – not to spend or build more houses.  SO, while we could use the rising prices and investment in the mid-2000’s to point towards a bubble, this time we have limited investment and an underlying shortage of property driving up prices – this is not a bubble, this points to a failure somewhere in the building or credit markets.
  • Credit conditions:  Mortgage conditions have eased, and competition between banks (along with low wholesale funding rates) has driven down the cost for households.  This sounds similar to what was going on during the boom time.  But even so – we have pointed out that increases in NET borrowing have been low (or negative in real terms).  On top of this, for various reasons credit conditions are tight when it comes to building houses – increasing the value of existing housing.
  • Quality, income, and constraints:  As the productivity commission noted, there are significant issues currently impeding activity in the building industry – and thereby pushing up prices.

When this combination of factors is taken together, it feels like we have a situation which is “supply” driven, with a shortage or property driving up values.  This compares to any perceived “bubble” which would be “demand” driven, with expectations of capital gains leading to excessive investment and excessively high prices.

This distinction is important when it comes to the actions of the Reserve Bank.  There are three questions they need to ask when looking where housing appears within the context of their goals of inflation targeting and financial stability:

  1. Are current interest rates consistent with the level of general demand in the economy? If the lift in housing market activity is supply driven, this suggests that interest rates should be lower relative to a situation where it is demand driven, with the increasing borrowing that entails.
  2. Is the current stock of debt, and banks attitude to risk in general, taking into account the full social risk associated with these elements?  The RBNZ has introduced a range of policies to help ensure this is the case.
  3. Is the regulatory regime that the RBNZ implemented possibility having a greater impact on domestic demand, or the housing/construction market than was previously expected?

Weekend reading

This post from mainly macro combines my favourite (although getting dated) book on macroeconomics, my preferred methodological statement for economic modeling, and tackles the idea of microfoundations in macroeconomics.

One of my favourite blog posts in a while – I suggest you have a peek at it.  One thing I would note is that, for applied macro, one day we will realise that microfoundations provide many potential paths to a single macro relationship – and in order to get anything policy relevant we will need to come up with an idea of what is “plausible” in order to narrow this set.

This involves not just working with data, but having to make subjective calls – and that is the area where things get hairy fast.

I’d also note that microfoundations are a huge part of academic economics as they are currently doing the technical work of creating micro models that provide potential explanations of macro phenomenon.  This is an essential task, even if much of the work ends up with potential explanations we view as “not plausible” in the end.

Is that an inflation target …

Or are you just flirting with me Bank of Japan?

In truth the BOJ has tried to hold back from an explicit 1% inflation target, and is just discussing it as a “near term goal”.  While this isn’t as positive as the Fed move to an explicit inflation target, and Australia and New Zealand’s long-term policy of having an explicit inflation target and printed rate track, it is an improvement.

With Fed and BOJ policy improving, credit markets in Europe consistently settling since mid-December, implied volatility on markets way down (the VIX), and the cost of credit down significantly in the past 6 weeks we could be seeing a real improvement on financial markets.

What does that mean in little old New Zealand?  Well our higher exchange rate is tempering part of any stimulus coming from offshore, while its up to the RBNZ to keep an eye on the rate track.  If financial conditions look like they are going to improve in the near future the Bank may suggest that they will be lifting rates in larger chunks when they do get around to it.  It will be interesting to see what happens when we get to the March meeting.

Thinking about the US output gap

Via Marginal Revolution I noticed the argument that a drop in lifetime wealth may have reduced potential output, thereby implying that there is a smaller output gap (permanent loss in productive capacity).

Now, I share Scott Sumners concern about this view. It is true that a negative permanent wealth shock will in turn lead to lower consumption – but in of itself this does not imply that it leads to lower output, which is what GDP and potential GDP are measures of.

Tyler Cowen put up the best defence of it when he stated “Simplest response to Sumner and Yglesias is that we may have had a biased estimate of the previous trend, for bubble and TGS-related reasons.” [note, he improved the defence further in response to Krugman here], but I think we need to go a step further and ask “how could we have been past some long term potential output before”?  In truth we need an explanation that works for why potential rose and why it fell that uses the idea of wealth.

In order to understand why potential may have risen then fallen we need to ask what factors were influencing the expectations of individuals so that they supplied too much labour/invested in too much capital.  We can’t just say “they consumed more” because without the ability to produce we consume more by borrowing and importing – which leads to the increase in consumption and imports canceling out in GDP.  We need a reason why production, output, GDP, was higher.

For this we need to rely on expectations.  Start with the drop.  Suddenly wealth is lower – wealth is the stream of returns on an asset, in the aggregate sense it is the discounted sum of expected income/output that is expected in the economy.  A drop in wealth here suggests that peoples expectations of future potential output have fallen – for better or worse.  As a result, your expected return on investing is lower – whether that be in skills for work, or whether it be capital in your job.

On the other side, suddenly wealth expectations are higher.  Income now has a greater expected rate of return in the future, you are more willing to invest now.

There is a case to be made that, if the rate of return is higher now, you will be willing to invest in order to reap the benefit.  Furthermore, you would be willing to supply more labour in order to achieve the capital gain (a return) associated with those “higher house prices” in the future.

If your wealth expectations suddenly fall, you are not willing to invest as much in the future, as the expected real rate of return is lower.  You are not willing to work as much given that the return on savings will be lower.  As a result, “potential output” would have declined.

Note:  You could in turn read these the other way around, it depends on the magnitude of “income” and “substitution” effects from the change in the expected real rate of return in the economy.

Note 2:  This is an entirely supply based argument, as it is about potential output.  Potential output is the “supply” notion of the economy, while many of the other cyclical issues we discuss are “demand” based.

Sidenote

These shocks exist for any view of “potential output”.  And this doesn’t mean potential isn’t a useless concept – it just means that maybe there is a more solid variable we can use to tell us the same thing without the confusion.

Conveniently we measure the UNEMPLOYMENT RATE, and we have a relatively clear and fixed idea of what the natural rate of unemployment is.  As a result, the gap between these two is a lot more useful to look at when trying to ascertain whether we are below or above potential IMHO.

UpdateScott Sumner discusses why this doesn’t make sense for the US.  However, I think it is a partially workable argument for NZ given the inflationary pressures we were experiencing, the high participation rate, and the amazingly low unemployment rate all prior to the crisis.

Automatic smoothing with VAT/GST?

Via the Money Illusion I see that there is a suggestion to make consumption taxes an automatic stabiliser for a given economy.

Russ Abbott, who is a computer science professor at Cal State LA, sent me an ingenious plan for having the Fed use fiscal policy to stabilize the economy.  It involves sales tax rebates when times are bad and tax surcharges when times are good.  It would be easiest to implement in an economy that already had a VAT, and/or state sales taxes.  I see it as analogous to my proposal to makes cyclical adjustments to the employer-side payroll tax rate.  These plans tend to work best when the central bank is targeting inflation.  Of course an even better policy is to directly target NGDP expectations.

The entire paper is only 2 pages, a model of clarity and concision.

I can not access the article, so just have to discuss it in terms of a concept.  It also remember that the RBNZ has discussed the issue before, but I said this on the blog before I got into the habit of hyperlinking everything … I will find a link at some point.

The way I see it, expectations regarding the relative price of consumption now to future consumption are incredibly important – which is why we need to think about these very issues in terms of expectations.  This raises four things to keep in mind when thinking about a proposal like this:

  1. How do we determine the cycle,
  2. what timing is there between data on the cycle and when the tax change occurs,
  3. Given information about what the cycle is, and regarding any lags in timing, what does the change in VAT/GST do to relative prices for consumption in current and expected future periods,
  4. Is this necessary with inflation targeting?

So, in answer to point 1 we would determine that we are on the upswing of a cycle when we are a certain % above some trend per capita – it may not be perfect, but it works in an operational sense.  Very good.

In answer to point 2, we would know whether we were above this point with a lag of about a quarter (three months) in New Zealand.  A similar lag would likely exist overseas, as they wait to have sufficient data finalised.

So, when it comes to forming expectations, people in the economy will already know if we are near a point where the consumption tax is going to be hiked (due to this being a some cycle), and they will have partial data from other economic indicators for the three months after the end of this quarter – as a result, if it seem sufficiently likely that taxes are going to be hiked, consumers will foresee it and lift their consumption now.  Similarly, on the downside if people start to expect that a cut in VAT/GST will happen at a near time in the future (due to expectations of a downswing) people will cut consumption now.

In other words, by making the level of VAT/GST depend on the perceived point in the cycle we are at, we make any expectations of a downturn or upturn self reinforcing.

However, this is not the full impact.  When activity is below trend, then consumption taxes will fall, when activity is above trend they will rise.  Although the timing does cause some cyclical elements, the existence of the tax would likely smooth out the magnitude of any underlying cycle that happens to exist in economic activity (if the cycle is sufficiently larger than the deviations from trend that are targeted with the change in VAT) – in many ways it can be justified in the same way as an interest rate target for smoothing economic activity, as both involve changing the relative price of consumption based on where we are in the economic cycle.

Do we need this

Well no.  With inflation targeting, we have a natural built mechanism for dealing with cyclical changes in economic activity.  Furthermore, without having to rely on some arbitrary trend estimate we don’t have to worry about a supply shock knocking the fiscal position into a perpetual deficit/surplus.

Wealth distribution and demographics

A very good point from Stephen Gordon at Worthwhile Canadian Initiative that the growing concentration of wealth may, in part, be to do with changing demographics.

Although I doubt this is the sole factor behind the growing concentration of wealth, it is a factor I’ve been thinking about – and that I’m keen to see someone else quantify.

One thing we have to keep in mind is that standard economic theory does predict a growing concentration of wealth as

  • the idea of  “rational expectations” when some agents are not rational implies that rational agents will tend to suck up wealth from irrational agents – it is a common misconception that “rational expectations” requires any agents to be “rational” in the strictest sense …  however, it does imply that agents that are “more rational” do receive transfers through time from their “irrational” buddies.
  • the fact that individuals have different discount factors suggests that more patient individuals who are more patient will tend to accumulate more wealth.

Now this isn’t necessarily even an issue, after all if people built up wealth due to their own choices they deserve it.  However, trying to understand how much of the recent change is due to the full functioning of financial markets in recent decades, how much is due to demographics, and how much is due to other policy change is important to understand before we really know anything.