Big Change To Imputed Rentals Calculation In Gdp

[Liquid Goldfish]

I've just come across this ONS paper on changing how rentals are dealt with in the GDP figures from next year.

It seems to be saying they've got it quite badly wrong and have greatly overestimated the increase in the cost of rentals - both imputed and actual rentals.

It's way beyond me - so I'm hoping that FreeTrader and others can explain.

But it does look like a big change and might explain why imputed rentals have been shooting up so much as a percentage of GDP (or it might not! - as I say it's far too hard for me)

The move to basing HHFCE estimates of actual and imputed rentals on CPIH has highlighted some notable differences between the series. The new method for calculating rental data has reduced growth in the current price estimates for household expenditure on actual rental and imputed rental.

This link opens the paper:

http://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&uact=8&ved=0CDkQFjAD&url=http%3A%2F%2Fwww.ons.gov.uk%2Fons%2Fguide-method%2Fmethod-quality%2Fspecific%2Feconomy%2Fnational-accounts%2Farticles%2F2011-present%2Fchanges-to-national-accounts--revisions-to-household-expenditure-on-rentals.pdf&ei=STI5VJHABI2S7AbZs4DIDQ&usg=AFQjCNGVv0OZF0hMVUSkg04IHyNiNEO0CA&bvm=bv.77161500,d.ZGU

Edited October 11, 2014 by oldsport

[Bloo Loo]

not sure what is so hard, actual rental is available from a few dozen phone calls, and imputed rental is a figure they can adjust to flatter the GDP.

Neither need a model.

[XswampyX]

A picture paints a thousand words.

It's painting the words LIES LIES LIES. Talk about locking in your gains!

[Eddie_George]

A picture paints a thousand words.

It's painting the words LIES LIES LIES. Talk about locking in your gains!

That's shocking. I wouldn't put it past them to include house prices in CPI when they start falling.

[Sancho Panza]

Too late in the night to even contemplate reading this.

However,Shaun Richards blogs here

http://notayesmanseconomics.wordpress.com/

So ask away there.He has aone or two commenters who are uber geeks whne it comes to this stuff eg Andrew Baldwin.

[richc]

I wonder if they got this so wrong because they weren't differentiating between the private and non-private rented sectors. If I read this correctly, they were only surveying 1200 households per quarter, so no more than a couple of hundred people in the privately rented sector, and a couple hundred in the social landlord sector. If the two sectors are mixed together and housing associations have been raising rents up to market rates, then it wouldn't seem difficult to show that rental increases have been growing at 10%, even though that number seems completely unrealistic in the privately rented sector.

Theoretically, it doesn't make any sense to base home owner equivalent rents on a survey of social sector rents. If they were excluding that data, which they should have done, and only included private rents, then that means that they were basing 16% of their estimate of total GDP on a survey of about 150 households. Huh? The ONS spends about £350 million a year and that's the best they can do?

[Sancho Panza]

I wonder if they got this so wrong because they weren't differentiating between the private and non-private rented sectors. If I read this correctly, they were only surveying 1200 households per quarter, so no more than a couple of hundred people in the privately rented sector, and a couple hundred in the social landlord sector. If the two sectors are mixed together and housing associations have been raising rents up to market rates, then it wouldn't seem difficult to show that rental increases have been growing at 10%, even though that number seems completely unrealistic in the privately rented sector.

Theoretically, it doesn't make any sense to base home owner equivalent rents on a survey of social sector rents. If they were excluding that data, which they should have done, and only included private rents, then that means that they were basing 16% of their estimate of total GDP on a survey of about 150 households. Huh? The ONS spends about £350 million a year and that's the best they can do?

I don't think they're saying they got anything wrong more that they're trying to improve the data.

The rates of growth for actual/imputed rents are listed in tables 3+4 and then chained volume stats are to the side.These account for changes in consumptive behaviour(I believe)so that as prices rise,consumptive behaviour changes.They would reflect my experience of rents over the last fifteen years.

The wider worry for me would be that they're been using imputed rents in GDP calcs and it'd be interesting to get to the bottom of that particular data set.

Here to learn.

[FreeTrader]

I've been out of the country for the past ten days with no web access, so I missed this thread when it was originally posted. However during my absence a HPC member PM'd me, asking me to throw some light on the subject.

These revisions to the actual and imputed rent calculations, which affect both the income and expenditure estimates of GDP, seek to resolve an anomaly that I've been beating the drum about for the past three years, most notably in this March 2013 post.

However it's a pretty arcane subject, and there are two reasons why I didn't highlight this particular explanatory note when it was published by ONS last month:

1) These latest revisions only apply from 2010 onwards and are thus only a partial implementation of the revised methodology that ONS will be using for household rent expenditure in future. This has resulted in an artificial 'fix' to bridge the gap between the old and new methodologies.

2) Recently the new methodology that ONS is using to compute rents has itself found to be flawed, and this has resulted in the CPIH measure of inflation (the measure that includes OO housing costs) temporarily losing its designation as a National Statistic until the underlying issues have been resolved.

Late next year, when Blue Book 2015 (the UK National Accounts) is published, the above drawbacks should no longer apply. The new methodology will be fully implemented, leading to historical revisions that give consistent and more meaningful imputed and actual rent data series. Additionally the imperfections of the CPIH methodology will likely have been addressed.

Therefore it seems more sensible to come back to this subject when those changes have been made rather than discuss them now.

[richc]

I wonder if they got this so wrong because they weren't differentiating between the private and non-private rented sectors. If I read this correctly, they were only surveying 1200 households per quarter, so no more than a couple of hundred people in the privately rented sector, and a couple hundred in the social landlord sector. If the two sectors are mixed together and housing associations have been raising rents up to market rates, then it wouldn't seem difficult to show that rental increases have been growing at 10%, even though that number seems completely unrealistic in the privately rented sector.

Theoretically, it doesn't make any sense to base home owner equivalent rents on a survey of social sector rents. If they were excluding that data, which they should have done, and only included private rents, then that means that they were basing 16% of their estimate of total GDP on a survey of about 150 households. Huh? The ONS spends about £350 million a year and that's the best they can do?

With the release of the IPRHP (Index of Private Rental Housing Prices) data today, the ONS have included a comparison of Actual Rents in CPI to the IPRHP. Essentially, the CPI rents figure is completely distorted by including social sector rents. I really do think that this is how the ONS has come up with the whopping increase in imputed rents, even though including social rents in the calculation of imputed rents makes absolutely no sense.

[The Knimbies who say No]

Missed this thread first time around- thanks for highlighting and very interesting. The chart says it all really, a complete farce and perhaps better to revisit this in a year as FreeTrader says, assuming the measure of housing costs is sorted out properly in the meantime.

I note Peter Gittins is acknowledged in the ONS paper linked above. He was the guy who responded (I say responded rather than answered) my FoI request to justify the yawning gap between nominal private rents and nominal imputed rents.

£40Bn added to nominal GDP between 2007 and 2012 iirc, used to backstop further 'prudent' borrowing via debt:GDP ratios, backed by nothing much at all.

Interesting that if a change of Govt happens in May then it will be them who inherit any fallout.

And a continuing doff of the cap to FreeTrader for taking the time to identify and explain this to the rest of us.

Edited October 28, 2014 by Joan of The Tower