Careful how we treat the “economy”

By Matt Nolan 04/06/2013


I’m always glad to see people discussing the New Zealand economy with data and discussion.  So good on Lowell Manning for doing that here.

Of course, this doesn’t mean I have to agree – but I will try to disagree with some substance ;)

There are a number of the specifics I fundamentally disagree with, I will discuss that more later in the post.  But I wanted to touch a methodological line to start with.  The “economy” is the aggregate of individual actions, we cannot reach a conclusion on policy by placing value on outcomes without discussing what individual choices are involved. Four things come out of this:

  1. Without knowing why, we can’t really describe what is going on – this isn’t like a business balance sheet.
  2. As a matter of course, without a “why” we can’t figure out causation – there is a risk of getting things the wrong way around! (For the wonkish, we need behavioural relationships as well as identities to get anywhere ;) )
  3. When we talk about “debt” someone is borrowing for some purpose.  Someone has to borrow – it is not created from the ether.
  4. We should not want to treat the economy like a business balance sheet even if we could – as we want a society that “maximises happiness” or some derivative of … not economic output.  In honesty, the driver of many structural shifts we see is government policy, which is set GIVEN the fact we’ll take lower output to meet some social needs.  Let us keep that in mind please.

This piece ignores this, where are the prices of non-housing goods, where are the determinants of foreigners willingness to loan to us, and our willingness to borrow?  The links related to these things are weak to non-existent, and it makes the other concerns I have even more stark.  Let us move on with those:

Issue one:  Relative prices

As a starting point, we are looking at the relative price of one asset, the relative price of the stream of services from one product in the economy.  Applying an analysis based on aggregates and money supply is fundamentally misplaced here – such analysis is based on the general price level, the price of all goods and services.

The housing market issue is a “relative price” issue.  Housing is seen as being a lot more expensive than other goods and services.  Housing is seen to have provided risk adjusted returns well above asset classes.  This points to something so so different to monetary aggregates!

Issue two:  Building costs have risen significantly

Between March 2002 and March 2007 building costs according to the same indicator (Stats CGPI dwelling costs) climbed 37.6% (6.6%pa).  Looking at the last few years, when we haven’t built anything, and when house prices in half of NZ’s TA are down 20% in real terms, involves looking past the fact that the period where we saw house prices hit a high level DID involve significant building cost growth.  I agree with Lowell that this is hardly the only driver of what we have seen – and that land prices are very important – but we can’t building costs have been irrelevant.

If we are talking about “housing affordability”, focusing on the last three years is like catching the final 20 minutes of a play – we haven’t allowed for any of the factors to “build up” which lead to the “affordability concern”.  For the later M3 data, Manning does provide a lot more historical context – which makes me think he may have been a bit too loose trying to exclude building costs here!

On that note, I would point out that there are other reasons why land prices may rise … but we’ll put that to the side.

Issue three:  M3 is endogenous and subject to shocks

The level of economic activity, the willingness to invest and borrow, the velocity of transactions, and the price (interest rate) of doing so – these are all factors that can adjust M3.  Activist monetary policy will, in a sense, change the endogeniety and measured relationship of things like M3 and NGDP, as well as dealing with “shocks” … all through the way they adjust the opportunity cost of funding in the banking system (or more importantly the rules they set with regards to monetary policy!).

Now I believe the article accepts this for the most part – but ignores the fact that the movements involved are due to real economy shocks!  M3 is a useful measure for indicating that something may be amiss – but it doesn’t tell us what is amiss, and “targeting M3″ as a policy measure to deal with things going on in the economy is strange.

When people used to talk about “targeting M3″ it was because of a presumed relationship with inflation – a relationship that then quickly broke down!

Issue three and a half:  We can’t talk about borrowing with talking about why people borrow

This is an amazingly important issue, one that seems to keep getting ignored.  When we talk about “banks creating debt” and the “money supply climbing” we are communicating in a way that makes it sound like there is some money gushing lever we pull with no consequences.  This is not the case.

In truth, we need to ask why people are borrowing, and what restricting or loosening of the conditions they face for borrowing means for them.  Banks are only “creating debt” because people are willing to borrow at the prevailing rate of interest – tis the nature of demand by households and firms that is important here.

In terms of causation, it appears we need to ask why people are so determined to borrow to invest in housing (if that is seen as the driver), making the issues in the housing market one of the causes of what is going on in debt markets and potentially a driver of the current account deficits – rather than a product of it!

Given this, I strongly view this article as having the causality backwards – it is talking about indicators that are symptomatic of what is going on, and as a result indicators that are correlated with the outcome.  But the growth in M3 given current policy rules is not the cause of a lift in relative house prices – it is a product of the real underlying causes!  Something the productivity commission did make a fair go at, even if I don’t think they really hit the idea of “affordability on the head” (to be fair, no-one has, and they used the same definition most people have).

Issue four:  Try to avoid regressing a level on level (unless you have cointegration)

Generally we try to avoid regressing levels on levels (unless we have a cointegrating relationship).  There is a graph regressing GDP on M3 which straight OLS.  No mater what your view is – the fact it is levels will at least exaggerate our confidence in the result (by inflating the R2) and at worst will provide an entirely spurious regression.

The endogeniety between the two series is also not a particularly helpful property for drawing such a graph.  And actually, it makes this problem fairly fraught unless we start to define the structure around other things (eg monetary policy rules, banking policies).

Issue five:  Non-productive is not a nice term

Housing is investment in the non-tradable sector of the economy, it “produces” a stream of “housing services”.  Calling it non-productive and explicitly trying to pull resources away from it is not nice, it is the sort of actions that leads to people living in housing that they don’t like solely so we can support our favoured industry of choice.

There is a problem is society at the moment.  We have this 1950sesque bias towards “productive manufacturing” in our mindset.  But this does not make investment in non-tradable production, and investment in primary and service areas, a bad thing.  Instead we need to ask ourselves about the relative prices of these things – and try to figure out what they represent (changes in technologies, reactions to tax laws, biases in financial markets or financial market policy).

This comes back to asking “why is the relative price of a house so high” – really why!

Issue six:  Foreign ownership isn’t “the primary source of a bubble”

I would like this explained at some point.  People keep saying that “foreign ownership of thingys is creating bubble thingys”.  I’ll be honest, from where I’m sitting this makes absolutely no sense to me.  If foreign owners overpay for an asset then we can buy it back later for less … this is called a transfer to New Zealander’s.

If we are selling assets overseas and borrowing, shouldn’t the main question be “why are we doing this” and “are we making worthwhile investment with the capital we are getting in”.

We ARE NOT causing this in aggregate by buying houses off each other – that capital stays in NZ.  We ARE NOT causing this by buying land off each other – that capital stays in NZ.  Supposedly we aren’t buying assets off foriegners … and as a result, we must be building something here.  This is where we should be thinking, not in terms of aggregates themselves, but in terms of incentives for investment!

Bubble is not a catch phrase for when something is wrong.  A “bubble” is when people are willing to overpay relative to fundamentals – interest rates don’t reduce this bubble component, our policy actions don’t influence it at all.  If it exists, it exists as a transfer between people – and again, if it is foreign people overpaying, they are just sending us free goods and services.  I am uncertain why we are against this.

Issue seven:  The policy conclusions do not follow from the argument

Here is the conclusion:

Either the proportion of the available M3 investment funding ($253.b*0.945) now actively used to purchase and exchange existing non-productive residential property and other non-productive assets is reduced, or the investment pool M3 itself needs to be reduced through progressive debt retirement, starting with foreign debt.

The second option makes more sense because it removes a primary source of the structural asset “bubble”, namely the persistent growth of foreign ownership through the current account.

Reversing that growth means managing the exchange rate, but the government and it advisors like the Treasury and the Reserve Bank of New Zealand claim there is no viable policy tool available to do so. They are wrong.

The government could easily set up a variable and tax neutral Foreign Transactions Surcharge (FTS)[14]. The FTS is an automatically collected levy on all outward domestic currency transactions passing through the foreign exchange interface. Its effect is to increase the aggregate cost of repatriating domestic currency offshore and lower the aggregate cost of using domestic currency locally.

The revenue obtained from the levy would be used only to reduce domestic taxation and to begin repayments of foreign debt, in effect repurchasing those assets already ceded to foreigners. While an FTS will correct the exchange rate and current account it will have little immediate effect on investment prices or the property market. It would instead lead to a gradual easing of the rate of investment price increases over time.

Manage the dollar, tax foreign transactions, and “lowering debt” (which must mean the government cutting spending or increasing taxes – as we haven’t clearly indicated how we are “changing the incentives of private agents”) in order to deal with the housing market … no no no no.

Stripping it down to clarify the way I’ve read it, here is how I’ve read your argument:

  • House prices have gone up due to land prices
  • Land prices have gone up due to M3 growth
  • M3 growth has been “excessive” due to CA deficits

Therefore

  • Tax foreign transactions in order to lower the CA deficit

Here is the kicker – without any description of “why” there is no reason why taxing foreign transactions will lead to a lower current account deficit.  If we were to dig into why, we might start to see some of the perverse places this tax falls upon (taxing export receipts and exacerbating credit constraints are two of the areas that come to mind).  In this way, the tax solution really isn’t related to the premises – so even if I accepted those (note that in reality I rejected them in the prior issues sections, so this is for the sake of argument) the conclusion does not follow.

We can discuss this policy conclusion as a way of “rebalancing” – and in that case I will likely disagree as well, but it will be a debate on points!

But if the relative price of the housing market is the concern, this comes out of nowhere.  You’ve just said “NZ borrows, and foreigners cause bubbles, so we’ll tax all tax transactions related to currency conversion” (Note, this is what I’m reading that tax as – if it is a tax on something else feel free to correct me :) ).

These things are not free, there are people that will be hurt.  This is why instead of focusing on monetary aggregates and trying to rope the RBNZ in, we should be focusing on these issues as fiscal policy issues, and directly asking “who are we hurting and helping with this”.  In that way this isn’t about the housing market, it isnt’ about M3, it is about the costs and benefits of a Foreign Transactions Surcharge – that is the argument that needs to be made for this conclusion, and it is not the one that is made here :)