How you measure inflation in rents matters a lot to those on the Living Wage

Today there are several influences which will combine. Later we will receive the employment report from the United States which will tell us whether average hourly earnings have picked up from an annual growth rate of 2.1%. Half an hour later I will be at the Royal Statistical Society discussing the UK inflation measurement landscape. Whilst they are for different countries the principle that you need a good inflation measure to deflate wages and tell you what they can buy is vital. From it we can get the concept of real wages. At the moment the UK has a proliferation of inflation measures from the official CPI (Consumer Price Index) to the unofficial or not a National Statistic RPI or Retail Price Index. Here we have an immediate issue as CPI is running at an annual rate of 0.1% and RPI at an annual rate of 1%. So UK real wages are what exactly and that is before we get to the new planned headline measure called CPIH (where H is for housing costs) at 0.4%, or the new version of the RPIJ also at 0.4%? With wage growth having been so low for so long such variations matter quite a bit and that is before we try to look back and decide how much real wages have fallen since the credit crunch.

The National Living Wage

In a move which of course also had political ramifications the new UK government announced plans for a national living wage in its Budget. The Office for National Statistics reported on it yesterday.

The National Living Wage (NLW) was announced in the 8th July Budget, with the aspiration that it should provide a higher wage floor in the labour market, lifting the earnings of the low paid. It will apply to non-apprentices aged 25 and above from April 2016. In contrast to the National Minimum Wage (NMW) – which is set by the Low Pay Commission the intention is for the NLW to be pegged at 60% of median hourly earnings for those aged 25 and over by April 20201.

What would it be now?

In April 2014, the NLW counterfactual was £7.36 while the NMW was £6.31, 86% of the NLW.

Sorry for the word counterfactual which gets everywhere these days and its use is certainly inflated. But the good news is that the wages of the poorest would rise. Of course we do not know how much it would influence employment although was have had some recent hints with firms looking to employ under-25s who do not qualify.

But the flaw here is that 60% of median earnings has the implication that such earnings will rise and of course we have been through a period where they have often stagnated.

The Living Wage

This is defined differently as this from the Living Wage Foundation indicates.

The Living Wage is calculated according to the basic cost of living in the UK

This gives us the figures shown below.

The current UK Living Wage is £7.85 an hour….The current London Living Wage is £9.15 an hour

So we see a 49 pence per hour difference across the country and a £1.79 per hour difference in London where the cost of living is much higher.

How is it calculated?

The uprating of the Living Wage figure each year takes account of rises in living costs and any changes in what people define as a ‘minimum’.

it rose by 20 pence per hour in November which was a 2.6% increase which was considerably higher than any of our inflation measures so let us take a closer look. We see that some of the factors at play here are rising quite a bit faster than headline inflation measures as shown below.

The results are then uprated by 5.1 per cent in 2013 and 3.1 per cent plus £2 a week in 2014 to take account of standardised increases in council rents, set nationally,

One of the themes of this blog is inflation in the housing market and we are seeing some here in the official sector. I regularly cover both house prices and private-sector rents but we see that the pain is spreading. Another cost has pushed higher too.

The following childcare costs were calculated. In 2014, the calculation was based on a simple uprating of the 2014 by 3.3 per cent.

Actually a cap to general increases in wages is applied to the Living Wage such that it cannot rise by more than an extra 2% leading to this situation.

Based on the above calculations, the ‘reference’ level of the Living Wage, reflecting actual minimum living costs, is £9.20 in 2014, but the applied Living Wage, resulting from the capped increase, is £7.85.

There is a theoretical issue here which is that perceptions of the situation below are likely to change.

The Minimum Income Standard the UK shows how much money people need, so that they can buy things that members of the public think that everyone in the UK should be able to afford.

What about private-rents?

The ONS has published some new research today because its numbers for rents differ considerably from those calculated by the VOA ( Valuation Office Agency) and those calculated by the private-sector. The difference is large.

Overall, the sample average rent grows by around 35% between 2010 and 2015, 11% of this can be explained by index growth while an additional 17% is explained by changes in the composition of the sample between years.

There is a missing 7% here! But let us move on to what the ONS is driving at.

This is because more rental properties in more affluent areas are now included in the sample.

Okay so what happens?

In total, rents rose from around £810 per month at the start of 2010 to around £930 per month by March 2015. Around £45 of the £120 per month (or 40%) increase in rent can be attributed to changes between the samples, rather than growth within years.

In case you are wondering this is what is called chain-linking where the basket is adjusted each year. But I am sure that most of not all readers can see the problem with this. If you are one of those paying the extra £120 you may be wondering why £45 per month effectively disappears?

What we are discussing in the difference between an inflation measure and a cost of living index and it is nice of the ONS to produce research which puts it into context in an important area. It is most evident in London as shown below.

The impact of this is that between 2010 and 2015, the average rent for unfurnished flats in London has increased by more than half, but no individual category has grown by more than a third.

As so often we come back to the issue of quality of the statistics. Are the collectors that good at telling us the composition of the market? It is all very well to tell us that people are renting in more affluent areas but do we really know the conditions. My part of town (South London) has a very cheek by jowl situation as regarding low and high quality housing stock. Wandsworth has gone from  1.8% of the London sample to 6.7%.

Let me put this in another form, here is the official measure.

The OOH component annual rate is 1.8%, down from 1.9% last month.

Now here is a private-sector estimate from Home Let

The Index reveals that the average rent on new tenancies agreed in the three months to July 2015 across the UK was £977 a month, (£761 a month excluding Greater London). The average UK rent price was 11.8 per cent higher than in the same period of 2014.

Quality must have shot up must it not?


I wanted to bring together today some different concepts which are much more interrelated than the mainstream media will have you think. It is also true that official pressure heads in that direction to as in the Johnson Report which recommended that CPIH be the main UK inflation measure. But as you can see there are a lot of doubts about measurement in this area and I have omitted so far the fact that the data is from 2010 because 2008/09 are a shambles! Oh and that the whole series has been rewritten once already as it was in an alternate universe.

If we return to linking this to a minimum wage we need to consider that we appear to be in a phase of time where inflation hits the poorest the most.

On our preferred measure, among the lowest-spending households experienced average annual inflation of 3.3% between 2003 and 2013, compared with 2.3% for among the highest spending households. These differences compound over this period, and consequently the prices of products purchased by the former group have risen by 45.5%, compared with just 31.2% for the latter.

That’s quite a lot when you are in a position where you have not got much. Lets hope that the lower oil price has helped redress a little of this. But as you can see rents and housing costs seem to be on the march.


15 thoughts on “How you measure inflation in rents matters a lot to those on the Living Wage

  1. Shaun, I can give you three actual rent increases from properties that were rented out for a year and have just gone back on the market in August. These increases apply to rents that were already previously increased at the end of 2013 or early 2014 so it’s not as if they had been low for a long time. A flat in Bristol +17.6%, a two bed terrace in central Cheltenham +13.4% and a three bed terrace just outside Cheltenham +14.7%. Two of these are rented by my son and daughter and another by their friend. Their friends in Bristol and Cheltenham are reporting increases of a similar magnitude so I don’t think these are unusual. I recall that two of these properties were increased by around 10-12% at the end of 2013 so it appears to be a trend. Despite these increases and the high rents paid the demand is such that they are snapped up usually within a day. These increases and the high value of property can’t go on – I wonder what the trigger for a major correction will be?

    • “This can’t go on” is an emotionally persuasive cry, yet it does go on. And on. Anybody got any serious historical charts that might help us get a better idea of whether it does “go on”, or some reason to think it can’t, won’t, or more likely, ‘might not’.

      • James, there have been a number of major corrections in the past and I believe one is again overdue. Rents are now a major part, or in the case of my kids, taking the total income of one person to pay and are at a level that are really unaffordable. Salaries have lagged far behind rent increases and logically it can’t go on much longer! I’d be a rich man if I could forecast when it will change but something has to give.

        • Hi Paklavi
          I, luckily well-pensioned, also have two children, adults, one in Manchester, on in London. The London one is pursuing a four year parttime professional qualification only available in London, does not qualify for any sort of support, will end up in the underpaid public service, probably NHS. After a hellish year moving from shared dump to dump (expensive enough), to make her life tolerable and safe I now subsidise heavily her in a decent small studio flat, at the astronomic market rate. The Manchester one rents, has her eye on a small £110,000 terrace, bur works in the NHS, needs deposit help, and I can do that too. Lucky me, lucky them. (I did work long and hard and pay fully for my good pension!)

          The reason I responded doubtfully to your ‘it can’t go on’, is because I wonder. The free market and buy-to-let encouragement will push rents until they stabilise at a ceiling? I wonder if we don’t fully realise what the sources are for transferring money into the rent funnel, and how long it may take before the ceiling is reached. We, and hundreds like us able to do it, are similarly, informally and irregularly doing something to address the scandalous generational weath inequity, also a neocapitalist free market development, by shovelling money into the BTL pockets. As our young people get poorer, and have to rent, there may be all sorts of money being moved to them, pensions, equity releases, charities, who knows what, their relative poverty a productive line in blackmail. So long as the current political and economic zeitgeist prevails, I fear rent infaltion could run and run.

          Don’t know about you, when I was a young adullt starting out, even before as a student, the problem was made tolerable, just about, by legislative rent controls and other tenant protection. Seemed to work. But it wouldn’t do, got between the pigs and the trough.

  2. Great column, Shaun, as usual and thank you for devoting all of it to inflation measurement issues. I hope that the meeting at the RSS went well.
    Regarding your scary data on the gap between ONS and private measures of rent inflation, Gareth Jones of our RPI CPI User Group has done a lot of work on this issue. If I understand him correctly, a rental unit will typically be part of the ONS rent sample for about 14 months, when it was replaced by another unit. There is an effort made to match the departing units with entering units and it seems that a lot of the downward-bias in the measure of rent inflation that seems to exist comes from flaws in the matching process.
    I suppose this raises the question whether matching is required at all. You could simply let the old rents exit and the new ones enter, only basing your rent comparisons on matched samples. This is, in fact, what is done to calculate the rent component of the Canadian CPI. All dwellings are part of the sample for only six months, and a sixth of the sample is replaced every month. So there are only five months in which a rent change can be picked up for a dwelling when the average dwelling will only experiencea rent change every 12 months at best.
    If a rent cannot be picked up from a tenant in a given month, it is carried forward from the previous month. If a rent is still being carried forward in the sixth month it is in the sample, the Kovar imputation is used to potentially assign a rent increase to the dwelling. It is well described in the attached document by Jeannine Claveau and her StatCan colleagues.
    Unfortunately, Mme Claveau does not mention that the Kovar imputation has not been consistently applied in the CPI rent calculation since it was introduced. Moreover, ther e is good reason to believe that it does not, in any case, resolve the problem of downward bias in the rent index. Similar to the discrepancy between the ONS measure and private measures in the UK, there is a big gap between the CPI rent increases and those registered by the Canadian Mortgage and Housing Corpn., a landlord-based survey of rent increases.
    So while the ONS rent index seems to be downward biased, the Canadian CPI rent index may not be any better in this regard, and possibly is worse. One thing is for sure, even if you believe that the CPIH is sound in principle as an inflation indicator for the Bank of England, which neither you nor I do, the data problems with ONS rents absolutely prohibit any move in that direction for a long time to come.

    • Hi Andrew and thank you

      Yesterday’s RSS meeting saw in general two main themes. Firstly there was support for the proposed Household Inflation Index although one person felt it did not define what is inflation clearly enough. There was also agreement about the unsuitability of CPIH both in itself and as the new headline inflation measure as proposed by the Johnson Report.

      I pointed out that the research on rents published that morning by the ONS highliighted in practical and numerical terms not only the flaws of CPIH but also the likely differences in practice between it and the proposed HII. Oh and yet another consultation is in progress! It is hard to keep up….

      I note your views on the problems of rental inflation measurement in Canada too. Are there such issues in the United States where rents are a substantial component of the CPI?

      • The brilliant Randy Verbrugge has authored or co-authored a bunch of papers on the US rent CPI and imputed rent series, including this one:
        The US rent index is based on a tenant-survey like its Canadian counterpart, but it certainly doesn’t suffer from a short rotation period. Units were rotated out of the sample about once a decade, and from 2009 forward every six years. To keep the response burden bearable, rents are only collected every six months, with imputation for the months between. If I understand it correctly, as in the Canadian sample, units are simply rotated out, with no attempt to match exiting and entering units as in the ONS survey. So on the face of it, there is not the same reason to believe there is serious downward bias in the US rent survey. My friend Leonard Nakmura, of the US Federal Reserve Bank of Philadelphia, and a pair of his colleagues tried to compare US CPI rent estimates with a hedonic rent index based on American Housing Survey (AHS) data. They concluded that there was very minimal underestimation of rent changes from 1985 to 1999, if any. Perhaps both the ONS and StatCan can learn something from the BLS on how to measure rent change.
        By the way, with regard to the formula effect, Randy’s paper mentions this change that was not picked up in Bethan Evans’ paper on international practice: “From January 1987 through December 1994, the average of the pure rent price ratios of the matched renters was used to move the implicit rent estimates of the owner units. This average of ratios formula was biased, so it was changed to the ratios of the average pure rents of the matched renters.” (p.22) In other words, prior to the January 1995 CPI update, the Carli formula was used and after that the Dutot formula. It was only four years later that the US CPI moved to the Jevons formula for most items, but there was no change in the formula used for the rent CPI. Mr. Evans takes no note of it, but a ratio-of-average-rents formula continued to n be used for the Canadian CPI rent component after the switch to the Jevons formula, although here it just seemed to be overlooked rather than deliberately treated as a an exception.

  3. Hi Shaun,

    Lets hope that the lower oil price has helped redress a little of this

    I am often left wondering how much the unexpected plummet in the oil price has benefited the UK economy and whether it has, in effect, saved young George’s bacon. I realise that there are both ups and downs for the Government in a lower oil price but on balance do you think that Osborne is secretly grateful?

  4. Even though I’m two days late,another good column Shaun.This is the only place on the web that I see any of this stuff mentioned.I always remember you pointing out sometime ago that the rental data used for imputed rents in UK GDP was different to the renatl data in the CPI(I think-confusing,espceialy at this time on a Sunday morning)

    Keep up the good work.

  5. Pingback: How the UK establishment tries its best to mis-measure inflation | Notayesmanseconomics's Blog

  6. Here in Scotland the industry itself is trying to understand/explain private rents in detail and is using portal data to do it. We actually have far bigger samples of new advertised rental values than the Government and are reporting them at far smaller geographies than Broad Rental Market Areas which is basis for calculating Local Housing Allowance. See more here:
    We do recognise that advertised new lettings are not telling whole story – in many respects new lettings capture the moment when most rent increases actually occur (ie between tenancies) and so are likely to exaggerate actual rent inflation. To overcome this we are exploring pooling lettings data of sitting tenants from large agents and landlords – this will give insights that advertisements for lettings can’t.

    I saw some comments on the impact of oil prices. This chart shows how quickly rents have fallen in Aberdeen

    • Hi PRS SCOT and welcome to my corner of the world wide web.

      I am glad someone has a better rental database than the government as it is sorely needed for CPIH! However it has other problems and will never work satisfactorily in my opinion.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s