Andrew Lilico: Two economic policies

Are Gordon Brown's monetary policy arrangements beyond criticism? If not, how could they be improved? Could any criticisms of them be politically useful, or would they just be technical matters best left to academic and practitioner debate?

There are many ways the Brown's arrangements could be criticized, and many politically useful reforms that could be proposed. I shall here mention just two, trying to focus on the politically-attractive criticisms and leaving aside many technical critiques and nuances.

First, the inflation target that Brown gives to the Bank of England is flawed, is widely accepted to be flawed, and is only not reformed because of Brown's notional commitment eventually to join the euro. We should openly criticize this and propose an alternative. The current target, a 2% annual rise in the Consumer Prices Index (CPI), was introduced when Gordon Brown announced that his five tests for euro membership had not been met. It was explicitly introduced in order to promote convergence between the UK and Eurozone economies - they use the same consumer prices index (usually called the "Harmonised Index of Consumer Prices" (HICP)).

About one fifth of household costs in the UK are housing costs, but the CPI has no housing costs component. This is not because there is some argument against having a housing costs component. It is simply because there are technical statistical difficulties in formulating a common way to account for housing costs that will work across the whole Eurozone. These would not apply to a UK-only measure. The only reason it isn't done is because of Brown's notional commitment eventually to joing the euro. This commitment is damaging UK monetary policy, and everyone knows it.

ANote that this is a separate debate from one that some readers may be familiar with - the question of whether the Bank of England should include house prices or or other asset prices in its target. I don't believe that we should argue for that (at least at this stage). Instead, we should argue that the target measure should be a (let's say) "UK Household Prices Index" (UKHPI) - i.e. the CPI with a user cost of housing (something like the rental cost) component. Then the Bank would include in its consideration the single largest item of household expenditure, and apart from anything else, there might be a better chance that people would believe that the government index reflected true inflation.

Second, I propose that we argue for a switch from the current regime - targeting of annual inflation (so-called "inflation targeting") - to the targeting, instead, of average inflation over the lifetime of a parliament (so-called "price-level targeting" or "average inflation targeting"). Monetary theory has moved on since Gordon Brown introduced his arrangements, and they are no longer state of the art. The Bank itself, along with various other central banks around the world, is known to be interested in targeting average inflation over a period of more than a year. That way, if inflation undershot the target one year (say, it were 1% instead of 2%), then in later years of the Parliament the effective annual target would be slightly higher (e.g. say 3% in the following year) so as to correct for past errors. Many people mistakenly believe that that is what the Bank does anyway. It does not - under inflation targeting bygones are bygones, and past policy errors are not corrected for. One consequence of this is that the very long-term average inflation rate is highly uncertain under inflation targeting, but it would be much less uncertain under an average inflation targeting regime in which target misses are caught-up in later years. We should propose a regime under which they are.

Again, although regimes of this sort are now considered attractive by many central banks, Gordon Brown has not been able to modify the Bank's target to reflect this. Why not? Well, average inflation targeting is a long-term policy commitment, and over the long-term Gordon Brown's notional policy commitment is for the UK not to have its own monetary policy at all - we are supposed one day to join the euro (when we are "ready").

By arguing for average inflation targeting and for changing from the European-defined CPI to a UK-based measure that included housing costs, Conservatives would be able to place a wedge between themselves and Brown, and get on the right side of two monetary policy debates in ways he would find problematic to follow and which would play to the Conservative Party's Eurosceptic strong-suit.

Comments

[William Norton]

Andrew - you've written an article without once mentioning the Whigs. Are you feeling well?

This is a convincing case for a change in inflation-targetting (although "user cost of housing" sounds like it needs a bit more work). If you're going to reform monetary controls, would you go the whole hog and reform the way in which the National Debt is calculated? Not just the PFI/Enron accounting argument but, e.g. recognise some current cost of paying future pensions.

[Jonathan Powell]

I agree that inflation targeting leaves something to be desired, but one of its strengths is the way it allows us to hold the central bank to account for inflation, month by month. I worry that with price level targeting this accountability would be weakened since there is only one "day of reckoning" every 4 years, when we can evaluate the performance of the bank. I wonder, for example, what sort of errors would be allowable (since it's unlikely the target would be hit exactly) and what happens if the target is not met -- sack the governor?

Also I wonder if, given the long and variable lags of monetary policy, the length of a parliament would be long enough. If not, wouldn't a longer period weaken accountability further? And what about the fact that the length of a parliament is not fixed, wouldn't that mean that price-level targeting could only operate for the 1st 3 years, say, then revert to inflation targeting until the election is called?

On including housing in the CPI -- meh. I don't think it's much of a vote-winner, especially since we already have the RPI which is used to measure the cost of living. I think the evidence is pretty strong that the CPI is on average 0.5 points below RPIX, so with the lower target I don't see it as a big deal.

[Tony Makara]

What concerns me most about Gordon Browns economy is the underlying inflation that is attached to debt in the form of interest. Remember that a demand for interest creates a demand for money, just like a wage demand. This creates an inflationary pressure. At the moment the strong pound is masking the true rate of inflation. However as and when the pound takes sustained falls this inflation will come to the fore. The Bank of England will then have to raise interest rates to counter the inflation, however because the inflation has been created by a debt fuelled economy, any hike in interest rates is only going to add to the demand for money. Business in particular will have to raise the cost of goods and services to cover the extra cost of interest. That will lead to higer prices in the high street. The Bank of England will not be able to get out of the loop. Gordon Brown's debt fuelled economy has created severe economic problems that have still to surface.

[Andrew Lilico]

William,

The user cost of housing is very precisely defined, and indeed there even used to be an official government series for it. I just didn't want to lose even more readers by describing it technically.

Jonathan,

Price-level targeting would be subject to constant accountability, as is inflation targeting. The target would consist of a target average inflation rate (i.e. a target path for the price-level) and a permitted tolerance (rather like the +/-1% for the inflation target). Exceeding the permitted tolerance band would trigger consequences. Thus there would be accountability month-by-month. I would not term deviations within the band as "errors", though excessively sustained periods one side or other of the target path might attract a Treasury response.

If consensus developed around the use of average inflation targeting, and around the particular rate chosen, then the effective policy horizon would eventually extend well beyond one parliament, and the average inflation targeting regime would converge towards a price-level targeting regime proper. This would have considerable additional advantages. In particular, if average inflation targeting became sufficiently embedded as a policy, we could lower the average inflation rate to something like -0.5% (i.e. target slight trend deflation). I don't propose that we argue for this in the short-term (probably for at least one parliament, and perhaps two), because we would have to see how average inflation targeting worked in practice, first, and it would be a bad idea, from a policy perspective as well as politically, to have a negative year-on-year inflation target (for very technical reasons I'm not going to try to explain today, relating to the possibility of a liquidity trap and why that does not arise under a price-level targeting regime).

[Andrew Lilico]

Jonathan,

RE: the inclusion of housing costs in CPI, I think you miss the point with your comment that you don't think it is a vote-winner. Gordon Brown's unchallenged position as the master of monetary policy management is definitely one of his main vote-winners. We must challenge him if he is not to garner those votes by default. On missing out housing costs from the Bank's target he is wrong, and all key City journalists, MPC members, and other financial commentators will agree with us. We will therefore be seen to be offering competent thoughts on monetary policy, and to have the advantage on something in this area over Brown. Even if the public is only dimly aware of the precise issues, they will recognise the balance of expert comment, and we will gain political advantage from that - from being right.

[William Norton]

Andrew: please forgive me. I thought the user cost of housing was one of those things which everyone agrees would be a jolly good thing to measure, but no one can agree how to measure it - isn't that why it isn't included in HICCUP?

[Andrew Lilico]

William,

I don't think there is any particular problem with producing an estimate for the UK. We could even use an updated version fo the DCLG experimental rents series. If you think it would be better to use a "shadow rental cost" series, then people have done that, also (e.g. in the reports of the Barker Review of Housing Supply). You might find it interesting to consult former MPC member Willem Buiter's recent blog on this: http://maverecon.blogspot.com/2007/07/plugging-hole-in-uk-and-eurozone.html

The reason it isn't done is just because of difficulties in agreeing one way to do it for the whole Eurozone. There would be no particular difficulty in coming up with a reasonable proxy for the UK.

[Andrew Lilico]

Other references for interested readers might include:
- My article criticizing the switch to CPI in the Parliamentary Monitor for December 2003
- Mervyn King's recent call for inclusion of housing costs in his target: http://news.bbc.co.uk/1/hi/programmes/inside_money/6906592.stm
- Any of my various papers on price-level targeting, e.g. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=233437
- My chapter in the IEA book here: http://www.iea.org.uk/record.jsp?ID=305&type=recommendedBook
- A Bank of Canada working paper here: http://www.bank-banque-canada.ca/en/res/wp/2000/barnett-final.pdf

[William Norton]

Andrew: thank you for that. I will study with interest at greater leisure. But I'm actually more interested in the implications for national/governmental accounting.

If you're going to try to provide a more honest monetary targetting then in principle you ought to do so on the basis of more honest economic figures. The only problem with that is that if HMG were required to report its financial statements on the same basis as a quoted company, we might discover that the UK is actually bankrupt. This information could well be considered to be a mixed blessing....

[Jonathan Powell]

"Price-level targeting would be subject to constant accountability, as is inflation targeting. The target would consist of a target average inflation rate (i.e. a target path for the price-level) and a permitted tolerance (rather like the +/-1% for the inflation target). Exceeding the permitted tolerance band would trigger consequences. Thus there would be accountability month-by-month. I would not term deviations within the band as "errors", though excessively sustained periods one side or other of the target path might attract a Treasury response."

That does seem sensible, actually. It's more like inflation targeting than I first thought. Personally, I'd like to see it work in practice before adopting it, though. I mean, the ERM seemed like a good idea in theory...

"In particular, if average inflation targeting became sufficiently embedded as a policy, we could lower the average inflation rate to something like -0.5% (i.e. target slight trend deflation). I don't propose that we argue for this in the short-term (probably for at least one parliament, and perhaps two), because we would have to see how average inflation targeting worked in practice, first, and it would be a bad idea, from a policy perspective as well as politically, to have a negative year-on-year inflation target (for very technical reasons I'm not going to try to explain today, relating to the possibility of a liquidity trap and why that does not arise under a price-level targeting regime)."

I'm all for lowering the inflation rate, although I'd prefer a rate of zero for the sake of simplicity (e.g., you could easily compare prices over time, and there would be no distinction between real and nominal interest rates). However, I think this could readily be achieved with inflation targeting. I think the liquidity trap idea is a bit of a red-herring, since the central bank can always increase inflation by printing more money.

[Andrew Lilico]

Jonathan@02:08

Actually, in an economy with cost-reducing innovations (basically, a growing economy), you get more comparability of prices over time with slight trend deflation. The reason is that cost-reducing innovations tend to be sector specific, so if the overall inflation rate has to be zero, then prices would have to rise in less innovative sectors to balance the cost reductions in the innovating sectors. This means that there are uninformative price rises. The number of price changes is minimized with slight deflation. (In fact, they would be minimized under something called a "productivity norm", rather than price-level targeting with slight trend deflation, but the constraints in discretion from price-level targeting and various other expectations effects mean that it is a better policy overall.)

Anyway, this is all much more technical than is really useful for a politics blog. My two proposed policies have political sellability as well as technical virtue, so I suggest we stick to them.

[William Norton]

Andrew: technical virtue, yes: but still a complex sell. Remember, we're operating:

(a) in a country in which a commitment to a growth rule (2001 election) and a milder growth rule (2005 election) were rebutted with some success by Labour as either economically illiterate or as requiring savage expenditure cuts;

(b) with a Party which doesn't understand the 'sharing the proceeds of growth' formula.

The average-inflation targetting you've discussed is a better policy, but still for most people an obscure technical adjustment. "Lower price rises for longer" would be a good starting point for selling the policy. I'm not sure John Voter would need to follow the intricacy of the argument - he'd just see Labour refusing to follow and be led towards the right inference from that. What would happen, though, if Brown/Darling did adopt your suggestion? (Arguably it fits quite well with golden rules being applied over remarkably adjustable business cycles.) Presumably we all cheer.

There is less likelihood of Labour curing its HICCUP, because that's an innovation of Brown's own devising. Here presumably the line is "An End To Fiddled Figures" (probably more attractive than "We Don't Want No European Numbers Here").

However, given the cynical world in which we live, if I were a Treasury minister having to respond to your attack I'd try to muddy the waters and spread the idea that Lilico wants to include housing costs in the inflation target because the Tories want to force down the value of people's houses - then remind John Voter of early 1990s negative equity, ERM etc etc. Cue patronising remark welcoming the Conservative change after 15 yrs from an incompetent and inadvertent attack on house prices to a deliberate and snobbish one. Newsnight cuts to picture of Cameron's house plus windmill etc etc.

Also - if we are on a campaign for less-fiddled figures, what do you say when Paxman asks you if that means you'll put the PFI numbers inside the National Debt?

[Jonathan Powell]

Andrew:

You make an interesting point, which I hadn't considered. However, I would stick with a target of zero since it's widely believed that measures of inflation such as the CPI overstate inflation by 0.5%-1.5%, so a zero target would probably imply de facto deflation of at least 0.5%, with the advantage of simplicity (I think the man on the street would struggle to understand your rationale for negative inflation!)

[Andrew Lilico]

Jonathan,

Yes, there is an over-estimation of something like 0.5% or perhaps even 1.0%, because of under-adjustment for quality improvements. That's why I proposed 0.5% (or perhaps 0.75%) deflation instead of 1.8%. If you are particularly keen, you can read more of this here: http://www.blackwell-synergy.com/links/doi/10.1111/1468-0270.00400

[Andrew Lilico]

William,

The thing about average inflation targeting is that it will be very easy to find central bankers and academics to support your position and hard to find anyone to oppose it. Hybrid inflation/price-level targets are close to the mainstream thought about where policy will go next. City journalists will enjoy having something to write about to explain clearly. And I think that the idea of targeting the long-term inflation rate so that policy mistakes are reversed is pretty intuitive and easy for the interested layman to get.

What will be even easier to get will be the thought that Brown hasn't gone for this yet because of his commitment to joining the euro.

If Brown and Darling do copy the policy, all the better - then the Conservatives are seen to be proposing highly credible reforms (so credible that they are enacted) in monetary policy, supposedly Brown's unassailable strong suit, and we're back in the game!

[Jonathan Powell]

Andrew:
Thanks for the link. I have downloaded the article and I shall read it later.

One final thought re your comment " What will be even easier to get will be the thought that Brown hasn't gone for this yet because of his commitment to joining the euro."

Is it really credible to suggest Brown wants to join the Euro, given that most commentary suggests he prevented Blair from doing so? Also, why would he set a target of 2% when the Euro target is 0%-2%? Surely a target of 1%--i.e., within this range--would be more consistent with Euro membership?

[William Norton]

I'm with Jonathan here. I see no traction in dragging in Europe. It will just remind everyone that it's down to Gordon Brown that we're not in the euro already, and that undermines our attack line on the latest treaty.

To be clear: I would support both of Andrew's suggestions - and I'm not too bothered if Labour do adopt them. I'm less convinced that the Tory party would get much credit if that happens.

If things are running your way as an opposition party, and the Govmt steals one of your policies, then it underlines that they have lost authority and you get the credit: but you're already making the running, there's no change of direction or initiative - you were never out of the game.

If things are running for the Govmt, and the Govmt steals one of your policies, then either nobody notices or the PM gets the credit for tacking towards the centre ground (cf border police force, drugs reclassification etc) - you're not back in the game.

[Andrew Lilico]

Jonathan,

I didn't say that Brown wants to join the euro. I said that he is committed to eventually joining the euro - a different thing entirely. And there is no question of credibility concerning whether the CPI target is there because he is committed to us joining the euro eventually. It is what he said himself - check his speech on the five economic tests: http://www.hm-treasury.gov.uk/documents/international_issues/the_euro/assessment/euro_assess03_speech.cfm

I do not suggest that Brown would find it impossible to bring housing costs into the inflation target or to adopt average inflation targeting. But I do believe he would find it problematic to follow such a course - by doing so he would be changing from a "not yet" to a "not in my political lifetime" stance on the euro, which would give him difficulties with EU partners and with his backbenchers. Forcing him to pay that price would be (a) right; and (b) to our political advantage.